Soon after, we went together to Athen’s biggest business forum at Panathenea, and there I had the opportunity to set my camera and interview him.

George is the founder of Minuttia, a content and SEO shop built for B2B SaaS.

We have worked together on multiple projects, most saw great success, including the on we run on LearnWorlds, I spoke about it in my presentation in Start-up to Scale-up here.

Use coupon code NICK10 to get 10% off Minuttia’s AEO Course.

In this episode, you will hear us discuss

• Why George stopped calling AI search a new channel

• The answer 40% of marketers gave when asked about their biggest AEO problem

• The gap between doing AEO and knowing what you’re doing

• Where backlinks fell to, and what sits above them now

• The math for defending a 5K per month AEO budget to your CFO

• How to check whether your buyers are even on AI search

• What Minuttia actually reports back to clients

AEO is an extension of search, not a new channel

George’s position is that the overlap between traditional search and AI search runs deep. Deep enough that treating AEO as a fresh start is a mistake.

Picture SEO as a pyramid.

The base holds up everything above it, and most of that base did not move.

What still counts:

• topic breadth, and depth.

• Publishing velocity.

• Topical authority.

• Content monitoring…

…which matters more now that study after study shows AI systems favor fresh content.

The building blocks of content optimization. Crawl efficiency and accessibility, with one nuance.

Do not block the crawlers that feed AI.

CCbot pulls the Common Crawl dataset, and Common Crawl sits inside the training corpus of most of the systems you are trying to show up in.

My metaphor about SEO vs AEO:

SEO and AEO are the same set of glasses. The difference is how full each one is. For SEO you pour more into backlinks. For AEO you pour more into citations and brand mentions.

SEO & AEO are at least at 80% the same work, but different quantities of each tactic to affect it. A few differences exist, but doing great SEO means you are 80% at great AEO.

Author’s note: I keep returning to this framing because it kills the reset instinct. Teams hear “AI search” and want to burn the old playbook and buy new tools. The work is mostly the same work, reweighted. I made the data case for that in AI vs SEO & PPC: What the Data Says. Nothing George said moved me off it.

Measurement is not Solved. Read it Directionally.

Ask George for the biggest surprise in the survey and he goes straight to measurement.

Question 15 asked marketers to name their hardest problem.

Around 40.6% said measurement and attribution.

His read is blunter than the number. With the tools that exist today, measuring AEO properly is not possible.

AI engines mask the sessions they send you.

Referral traffic shows up as something else, or as nothing.

And that is one touchpoint.

Someone reads an answer in ChatGPT, opens Google two days later, lands on your site as direct a week after that because they remembered your name.

You will never draw a clean line through that path.

It behaves like dark social.

Your real AI-influenced traffic is almost certainly higher than what the dashboard shows.

Then there is the board problem. Leadership hears AI search is a thing, hands over budget, and three months later asks for the ROI.

Author’s note: This is the oldest fight in marketing wearing a new jacket. The best work rarely leaves a clean attribution trail. I wrote the long version in Measuring Brand Performance - The Hard Truth, leaning on Rory Sutherland’s line that marketers are opportunists and CFOs are optimizers. AEO just handed both sides a fresh thing to argue about.

80% are Doing it. Almost Nobody is sure.

The survey found close to 80% of companies actively working on AI search, either in early experimentation or as core to their efforts. High number.

Then it asked how confident they felt about that work, and the confidence dropped.

George’s theory, and he is careful to call it a theory, is that people confuse activity with competence.

SEOs are doing things.

Doing things gets mistaken for knowing things.

Some of that motion is real experimentation, which is how you are supposed to operate in a channel this young. Some of it is motion staged to show a manager motion.

Out of that, three types of company keep surfacing in his calls:

• Panic, no change. They have the mandate from the CMO or the board. They take agency calls, spend everyone’s time window-shopping, and change nothing.

• Same playbook. No panic. They run their 2020 SEO motion untouched and assume it transfers.

• Willing to change. The smaller group. They treat the environment as moving and make decisions that account for the movement.

Only the third group compounds.

Author’s note: This maps to something I have argued before. The edge right now is effort and willingness to run experiments, not access to a secret. I made that case in The New Competitive Moat: Effort. The panic companies and the same-playbook companies are both standing still, for opposite reasons.

What ranks now, and where backlinks fell to

Here is where George reorders the furniture.

For years, backlinks were the whole game.

In AI search he drops them to the tip of the pyramid, least important of the things that still matter, though correlated, since pages with more links tend to get pulled into the search features these systems run when they research a query the way you or I would.

His hierarchy, top to bottom:

• Brand experience. What it actually feels like to use your product. If your support response times are bad, no volume of ghostwritten Reddit accounts saves you. The bad experience overflows and spreads. You cannot hide it.

• Owned perception. Your site, your social handles, the directories you manage.

• Multi-channel perception. The mentions you do not engineer. A Reddit thread about your brand, positive or negative.

• Backlinks. Still there. No longer the point.

More brand mentions and citations for AEO, more backlinks for SEO. Same activities, different weightings.

Is $5K a Month for SEO/AEO Worth it?

George opened with “it depends,” which is the correct answer.

Drop traffic as the unit.

Use average selling price times the number of people you influence.

An e-shop at 100 dollars ASP influencing 100 people a month makes 1,000 dollars off a spend that costs more than that.

Thin.

An enterprise product at 100K ASP influencing one or two buyers can justify a new customer a quarter on 5K a month, which is triple ROI.

The economics flip entirely with the price of what you sell.

One condition sits before the math.

Your buyers have to be on AI search in the first place.

A dev tool selling to engineers can safely assume its audience lives in ChatGPT, Claude, and the rest.

The honest part came last. When a CEO or CFO asks for an ROI number on a channel this new, that number is a well-educated guess.

A calculated risk…

Wrapping up

Six things I took from this one.

• AEO is reweighted SEO.

• Read AI search numbers as a floor. The influence is larger than the dashboard shows.

• Audit whether you are still using the same playbooks.

• Check that your ICP is on AI search before you spend a $/€ on it.

• Price the risk with ASP, tell your CFO it is a calculated guess, and mean it.

• Brand experience sits above everything.

George’s closing line held the whole hour in one sentence.

“Take all these things directionally, not at face value.”

That is the job.

Read the signal, size the risk, keep moving.

Google Trends is showing near-vertical spikes to 100 for terms that have no business moving.

I tested 31 keywords around topics I knew pretty well and added some always-relevant high-volume queries as well.

This is happening in the same year that human searches per user fell nearly 20% and AI Overviews are being expanded to cover the whole world.

Three Waves, The Timing

The anomalies do not all start at the same time. They arrive in three waves across 2025, and each wave lines up with a Google AI release.

• Wave one, around June 2025. Music, cmo, remote work, AI tools, digital marketing strategy, fractional cmo services. This lands one month after Google I/O, where AI Mode launched in the US using a technique Google calls query fan-out. One question gets broken into subtopics and fired off as a batch of separate searches.

• Wave two, around August and September 2025. Marketing, social media, travel, food, artificial intelligence. This is when AI Mode went global, and when Google removed the num=100 parameter on September 11, stripping a wave of bot impressions out of Search Console overnight.

• Wave three, October to December 2025, peaking January 2026. Electric car, how to invest, best travel destinations, work from home. AI Mode reaches 180+ countries, Gemini 3 Flash becomes the default model, and then Gemini gets embedded directly into the Trends Explore page.

Now the mechanism. When an AI answers one question, it fires off a batch of background searches to Google to assemble the answer.

A single human question becomes many machine queries.

If Trends is sampling from that pool, it has stopped measuring human curiosity.

Huge disclaimer here, while it matches some Google events, it may come from other AI tools, crawlers, or misconfigurations hitting Google searches. We are not certain what causes this, and there has not been any public acknowledgment around this.

The Proof

I first noticed Fractional CMOs are trending, this is a topic I have been following, because I wrote an article and worked as a Fractional Head of Marketing for a couple SaaS.

Similar trends were showing up for Fractional CROs, and Fractional CFOs.

Google Trends shows “fractional cmo” on a steady climb to an all-time high. Here is what my actual Search Console says for the exact same query, same period.

An average position of 21.5 that barely moves most of the year. And the impressions do the opposite of the Trends line.

They run high through mid-2025, fall off a cliff in the fourth quarter, sit on the floor through January and February 2026, and only partly recover in spring.

The Trends Graph does not agree with the Google Search Console data.

Why is that?

Either the demand was never as real as the Trends line implies, or Google cleaned the bot traffic out of Search Console and left it inside Trends. Neither version is good for anyone making a content bet off that 100.

This is worst for B2B, because Trends is noisiest for low-volume terms, and low volume is where B2B lives. Semrush’s own guide shows a term like “fractional content team” peaking in Trends on 10 monthly searches.

Online Chatter on the Topic

I am not the only one seeing this. James Finlayson, Head of SEO at MPB, started a conversation a couple weeks earlier. Mature categories like air fryers, car insurance and London hotels are all at all-time highs that bear no relation to reality.

The thread underneath is a who’s-who landing on the same conclusion.

• Joshua Squires at Amsive and Peter Jaffray both point at fan-out and LLM trackers, with Jaffray noting Trends climbs while human sessions in GA4 flatten.

• Lily Grozeva sees the same distortion in her own Search Console and traces it to AI bots from her Rankscale projects.

• Malte Landwehr, CPO at Peec AI, posted the same week that the most common fan-out terms all spike in Trends.

• Daniel Pokorny reframes the whole thing. Search data used to be a proxy for human curiosity. Now AI generates the queries on our behalf, so are we measuring people or machines.

Then the honest counterpoint, because it matters. Chris Byrne and Natalie Le both cite Google’s own documentation, which says searches made by AI Mode and AI Overviews are filtered out of Trends.

So if anything is leaking in, it is the other LLMs grounding their answers on Google, the GEO trackers, and the scrapers.

Or, Google’s filter is not holding.

Either way, the clean “Gemini broke Trends” story is wrong, and the real one is messier and more interesting.

Why it Matters: Trends Became a Feedback Loop

Separate the two problems. The data problem started mid-2025. Gemini only entered the Trends interface in January 2026. So Gemini-in-Trends did not cause the spikes. What it does is make the loop worse.

Trends is no longer a passive chart. With Gemini on top, the platform now interprets the data for you and tells you what is “emerging”.

Here is the process:

1. The AI flags a breakout.

2. Content teams chase it.

3. More content gets published.

4. The signal looks validated.

If the original spike was noise, you have spent pipeline and budget against phantom demand, and then manufactured the very trend you thought you spotted.

What to do Instead: Triangulate

Validating every spike against absolute volume and first-party data before it touches your calendar.

Use first party data to provide trends (if possible), else, use more tools.

My tool recommendations:

• Ahrefs or Semrush to pair trend direction with real volume, keyword difficulty and SERP competition. Combine the data you see with their own tools, where work to decouple traffic from multiple sources to the keyword level. But, much of the data is not grounded in reality, it’s just another touch point.

• Exploding Topics for genuine discovery, organized by sector including SaaS, using analyst review instead of a relative index. It combines data from multiple sources including social media trends to provide trends to what’s happening.

• Brandwatch for social listening across millions of sources, with Talkwalker or Brand24 as cheaper entries.

Then cross-check anything interesting against your own Search Console and CRM. Your first-party data is the one source the machines are not inflating for you.

A Few More Examples

I first thought Fractional CMO was trending, and I updated my article dating from 2023 on the topic. This did not work as well as expected, I lost little time… but I started thinking, what if I did the same for 10, 20, or 30 pieces?

That’s actual time & cost sunk into trusting a tool that is broken.

Here are a few more charts going haywire because of query fan-outs, LLMs, and bots…

Established review platforms screen their reviews and are quite trusted by both people who still visit them and LLMs training on their data to provide answers to your customers.

I have run review projects in two B2B SaaS now, utilizing a support or CSM team to collect the reviews.

High-touch, white-glove B2B companies with a low number of users but strong relationships have to use semi-manual methods to collect reviews.

Let’s go through on the process & some tips :)

Get Your Profile Updated

Buyers self-educate through most of the funnel, the buying committee uses top X lists to prepare their short list (if they don’t already have a solution in mind).

You probably want to start with sites like:

• G2,

• Capterra,

• TrustRadius, or

• Clutch for agencies

G2 recently acquired Acquire Capterra, Software Advice, and GetApp from Gartner. Making it one of the biggest consolidation in the market.

Avoid TrustPIlot

Keep away from TrustPilot, I have terrible experiences with SaaS users blackmailing companies using reviews. IT does not review or remove any review for any reason. It’s a place where black hat marketers use to manipulate reviews, hurt competitors, and bad actors run wild.

Your Profile

Be sure to create your profile or take-over your existing profile. In most platforms, you may need to prove your ownership, talk with an account executive and work your way through some burreacracy.

What you need to complete:

• Consistent Profile Name to your product - consistency helps for SEO / AEO / GEO.

• Company / product description - on brand & descriptive to help new users understand your offering.

• Products & features - be sure to complete the products, services, and features as they help define and position your in the category.

• Visuals, screenshots, and videos - showcase your product, prepare videos, and screenshots to enrich your profile

The most complete profiles unlock features and help in-depth readers understand your product.

Pro Tips

Platforms are not always closed systems. You can do a lot with a little bit of negotiation:

• Ask for more categories to be included (if it’s relevant).

• Negotiate to open a new category if it makes sense.

• Price does matter - G2 overcharges for companies with more than 50 employees, but has a startup pricing for under 50 you can take advantage.

Badges help, but they are mostly a growth method of the platforms not you. Earning badges might be a good way to use for press, or an email to showcase category leadership. But, it does more to promote the platforms than you.

Measuring Brand Performance

Brand is hard to measure, reviews are part of what influences brand.

But, a shameless plug here, I wrote a post about Measuring Brand Performance you should read next :)

Reviews are a System, NOT a Campaign

Here’s where most teams get it wrong.

They treat reviews as a one-time push. I did as well, when I have not enough time on my hand to actually work on it.

• A single email blast to the customer base once a year,

• a handful of responses,

• a screenshot for the report to the board, and

• back to the roadmap.

Platforms weight recency.

Ten fresh reviews a month compound.

It has to be repeatable, lightweight, and always-on, or it isn’t a motion at all.

1) Build the Team

First, you need to build the team, internal political capital, and support for collecting reviews. If it’s not a product-led SaaS, you need the Support / CSM team to help.

The split that works:

• Customer Success identifies the right accounts, sends the ask at the right moment, and follows up.

• Marketing builds the rails: the request templates, the tracking, the incentives, and the repurposing of every review into sales and demand assets.

2) Repetitive Motions

This is the process I run for my own team:

1. Bi-weekly CS sync. Quick standing call to keep momentum.

2. Each CSM picks 3 to 5 candidate accounts per cycle. Prioritize larger or more recognizable logos and your core ICP segments. They carry more weight on the page. (positioning matters)

3. Send the ask within 24 hours of a milestone. Timing matters a lot!

4. Marketing validates the review once it’s live. Capture reviews, be sure you have a good positive ratio and review negative ones (you can turn them around sometimes).

5. Have an incentive, update the leaderboard, repurpose the quote, offer recognition to the one who collected it.

3) Getting Reviews - When to Ask

Timing is the whole game. Ask when the goodwill is already there:

• After a key product milestone or a clean onboarding

• After a positive Quarterly Business Review

• Post-renewal or expansion

• Right after a customer sends you unsolicited praise over email (this one converts the best, and almost nobody acts on it)

4) The Asking Process

• Pick one platform your buyers actually use. G2, Capterra, TrustRadius, and Gartner Peer Insights cover most B2B SaaS. Don’t spread thin across six.

• You send the customer a direct review link. Use a templated email your team can copy paste for different cases.

• You incentivize the act of leaving an honest review, not a positive one. This is about compliance, you can’t ask for a positive review or you risk being banned. But, you can control who reviews based on good experiences and relationships.

• Track responses & send the gifts. Some reviews handle the gift giving, if not, use the templates shared below to track reviews and gifting manually.

5) What Incentives Work

Keep the customer gift small.

A maximum of $50 digital gift card for a written review, $75 for video, with a charity donation as an alternative.

Many organizations have policies against bribing, anything above $50 starts to look like a conflict of interest, and plenty of procurement teams won’t let their people accept it.

Download 2 Templates for Review Collecting

Here the templates I use as part of my process.

Download the Templates

1. The program one-pager (Doc).

Purpose, goals, roles, monthly workflow, budget, and the leaderboard model on a single page.

Fill in the bracketed placeholders, drop your tracker and leaderboard links at the top, and share it with CS as the source of truth.

2. The Review Tracker (Spreadsheet).

• Review Collection is one row per opportunity, with dropdowns for Review Type, Rating, Gift Card Sent, and Status, so the whole pipeline stays clean as requests move from Pending to Live. Delete the grey example row before you start.

• Review Score logs your average rating and total review count by month, so you can watch the trend instead of guessing.

• Instructions explains every column for whoever you hand it to.

Happy Review Hunting!

Remember, reviews is a repetitive motion process.

You can’t shoot and forget, at least you can’t do it once and forget.

You need a repeatable framework, a flywheel if you may, that needs to be repeated weekly, monthly, quarterly with heavy ownership.

And… it’s tougher than it looks!

Good luck!

I run growth for B2B software (SaaS) out of Europe, and most of my work has nothing to do with the local market.

It’s with European companies selling into the US, or companies trying to leave a home market that’s too small to scale in.

I’ve watched this play out as both an internal employee, fractional, and a consultant. The main challenge of local businesses is being less bold than their US counterparts.

Based on an Interview

This newsletter is based on an interview I recently gave for a Cyprus-based news organization in Greek.

As a good portion of people reading this newsletter comes from Greece & Cyprus, I thought to share the full interview as well.

I would appreciate a like, comment, and maybe a hype on the video 🙏

How Much Can You Grow Locally?

A profitable company in a small market can make a significant profit. Local businesses can be locally big.

But, do you want to be a big fish in a small lake, or a small fish in a big lake? It depends!

I tend to monitor some high-performing businesses in Greece, our closest market in terms of culture. The ~1 million people in Cyprus with higher buying power, versus the ~11 million people in Greece with lower buying power. Eshops, chains, software, everything tends to get bigger in Greece due to the critical mass that allows them to utilize markets of scale.

Italy, Germany, Spain, all around the ~70 million and up marks, are quite self-sufficient with companies performing most kind of services and software coming from inside the country and speaking their language.

But, still the ~300 million unified market of the U.S. provides both scale and capital to outperform most. The bigger lake has advantages.

Think of it like that (Greece vs Cyprus example):

• Sell at €100 per seat in a small market, win 400 customers, and you’ve built a €40k/m business, that’s a half million ARR.

• Take the same product into a market ten times the size, cut the price to €70 to win on a more competitive field. With 2000 clients you get to €280k/m, €3 million in ARR.

Lower price, bigger volume, more total revenue.

The small market loses on the only number that matters: the total.

For B2B software specifically, the home market is almost never enough.

The economics need volume the local market simply cannot supply.

The Multi-Country vs International Expansion Challenge

It feels easier to expand in another market in the EU or adjacent.

Every country is an island. The One Europe, One Market roadmap is still far, and too little to cover the differences.

Every European country is its own market. Own language, own systems, own incumbents, own buying culture.

Each country defaults to either local solutions or global ones, and “global” in most categories means American.

Crossing a European border means re-entering from scratch:

• A new language and a localized funnel

• A new set of entrenched local competitors

• A buying culture that quietly prefers its own

• And the bigger the country, the harder the entry

Possibly you need local speakers to get each market.

The US is more open to external competition, even when it prefers it’s local solutions, still. One large, English-speaking market that is more willing to try a vendor it has never heard of.

Americans are faster, take more chances, and churn more easily. It’s still better than the long decision cycles and legal blocks of EU.

One little secret: when you say “international” or “global,” it means U.S.-first with Tier 1 and English-speakers.+

• Primary: the Tier 1, English-speaking markets. The US first, then the UK, Canada, Australia, and Ireland. This is what we mean when we talk about targeting international markets.

• Secondary: the larger economies that speak English as a strong second language. The Netherlands, the Nordics, Germany, and similar. You can often sell here in English with light localization, which makes them the natural step after the Anglosphere.

• Everything else is a long tail. Real markets, but later, and usually only once the first two tiers are working.

So going straight at the US, then widening into the rest of the Anglosphere, is often a shorter path than grinding across fragmented European borders.

Build the Base First. Feet on the Ground Soon After

You do not need a US office on day one.

That is the expensive move.

Leading with it means betting the company before you have any proof the market wants you.

Do it in two stages.

First, build the base remotely.

Land your first customers from where you already are.

Work with digital marketing, outreach, a few short trips for B2B & conferences.

Prove the market!

Feet on the Ground

If you are working through B2B sales motions, feet on the ground, local people in events, conferences, dinners, meet-ups are a must-have.

For lighter touch services, you might need a good, clear accent and remote employees. But, a small presence in the U.S. while you grow helps.

A Story

Two years ago, I was working with a healthtech startup. Great, strong team.

On one of our general conversations, I told them one thing “your audience is in conferences, they need you to be there, you need to consider feet on the ground or travelling often to achieve your goals.”

Their answer surprised me:

“It’s already planned. In a couple of months, I (the CEO) am going for at least 3 months to join industry conferences and talk with investors. I will stay as long as I need to”.

He spent almost 6 months, and brought business & investment back.

If you want an international expansion, but not willing to pay the price, you are not going to see a strong return.

Slow & Weak Kills Your Marketing Motions

Entering a new market takes months.

It takes money.

It takes persistence.

Running one campaign and concluding “the US doesn’t work for us,” is a half-measured approach.

You have to commit. Truly try, test, pivot, talk with potential clients, and repeat.

This is the same argument I keep coming back to in The New Competitive Moat: Effort.

The bold move works because almost nobody makes it.

Effort is expensive.

That is exactly why it is a moat.

Now is the Time!

Speed is the other half.

Spot an opening, decide to move on it in three years, and the opening is gone. Someone else walks in and locks it.

• Your product is not ready.

• Your team is not ready.

• Your budget is not ready.

What you are building might not be a moat in 2-3 years.

The market waits no-one.

A Word on MarTech Stacks

I still like Spreadsheets for reporting, sometimes simpler, less complex solutions work at the beginning.

Tech bloat is killing growth.

Complex system need expensive maintenance.

Don’t be a Sniper

Most tech founders think having the TOP tool for every job is great.

But, one tool for each job might be more expensive in the long-term.

What I advocate is one of the following:

a) Use an all-in-one tool like HubSpot and a couple others for where it really falls short. But, keep in mind they work well with HubSpot.

b) Choose your CRM or most important central tool first. EVERYTHING else has to play well (integrate) seemlessly with it - no dev requirements.

Simplify, Consolidate

Avoid adding extra tech headaches.

Keep your stack simple, workable, easy for new employees, the same across departments.

This allows you for a birds’ eye view and quick work.

What you lose in visibility, tracking, niched-down metrics, you gain in speed and ease of use.

Where Europe Loses on its U.S. Counterparts

The thing that separates American go-to-market from European go-to-market often isn’t capability. It’s boldness.

American tech sells with enormous conviction.

The biggest AI companies are the clearest example we have right now.

Their leaders, from Sam Altman to Dario Amodei, and even the more serious Demis Hassabis, project outsized confidence and inflate expectations to a degree that would make most European founders wince.

Discount what they say by 90% and you’d still want to buy.

That confidence isn’t an accident or a personality quirk.

It’s how you win attention in a crowded market, and it works.

European founders tend to do the reverse.

They discount themselves. “We’re not ready.” “We’re too small.” “Let’s wait and see.”

(especially in my home-country, Cyprus)

That caution reads as humility in the boardroom and as weakness in the market.

I’m not telling you to lie about your product.

Forget what you know about the EU way:

• Slow

• Compliant

• Bureaucratic

• Long sales cycles

Now is the Time!

If you’re scaling B2B software out of a small European market, the constraint is mindset, not capability.

It’s how fast and how hard you’re willing to move on a market that will not sit still and wait for you to feel ready.

Are you building a Software targeting the international market? If yes, I would love to hear more in the comments 👇

You can easily ideate & draft, but it always needs lots of editing.

And, the more AI learns about you, the better it writes… or not?

Being Consistent

Marketing is about consistency. You, your team, your product, support, success, everyone needs to follow the same guidelines.

Now, everyone is writing content.

But, how do they keep consistent on-brand?

A Tone of Voice skill shared across the company helps everyone writing in the same tone, use on-brand wording, and follow the same positioning to avoid confusion.

Creating a Tone of Voice for Claude (and not only)

In this newsletter, I will both guide you and provide the tools to create a tone of voice you can upload as a skill to Claude. The same doc can be used as instructions to other LLMs to teach them to speak like you.

This guide is split into two parts: The easy way to build a skill, the advanced way.

a) The Easy Way: An Interview

An interview skill will help Claude to ask questions, and based on your answers deliver the skill.

It will be asking a series of questions until it can fill up the whole skill as it is structured.

b) The Advanced Way: Scraping Your Content

This is intended for more ambitious and technically adept marketers who have a lot of content already written. You can either scrape your LinkedIn or upload your content from other platforms and let the AI do it’s work.

Build a ToV in a 15-minute Interview

A Tone of Voice Builder skill you upload to Claude. Download the skill, get it on Claude, and ask it to run the interview.

Interview Skill

You say “help me build my tone of voice,” and it walks you through about ten short rounds:

• Who you are, or what the brand is

• Who you’re writing for (and just as important, who you’re not)

• The values and beliefs you keep coming back to

• The two to four adjectives that define how you sound

• How your register shifts between a quick note and a formal document

• Your language mechanics (tense, person, sentence rhythm)

• The words that are unmistakably yours, and the ones that make you cringe

• Your hard guardrails. The things the AI should never do in your name

At the end it spits out a finished, installable voice skill.

You upload it, and from then on it triggers automatically whenever you ask Claude to write for you.

Pro Tip: It’s a good idea to have project folders with instructions to call the skill when you are writing company or personal brand posts. This way, the skill is always prioritized when you use that project.

The Advanced Way: If You Already Have Content

If you’ve been publishing for a while, your real writing is a better source of truth than your self-description.

Or, it can be a great starting point to be consistent.

ToV ToolKit

I have already created this toolkit for myself, and it helped me scrape my last 2 years of LinkedIn posts, comments, and newsletter.

What’s in it:

• An instructions guide for exporting your LinkedIn archive and scraping your own posts, comments, and articles along with their performance data. (LinkedIn’s export takes about 24 hours, so request it before you need it.)

• A scraper chrome extension with instructions to be customized to export your content. Right now, it has to identify you by name, so you need to use Claude Code to edit it for how your name is shown on LinkedIn.

• Three channel-specific Voice Skill Builder prompts, one for each medium, because the same voice formats differently per channel:

  • Post Voice. How you open, structure, and close a social post. Hooks, line breaks, the lot.

  • Comment Voice. How you show up in someone else’s comment section, which is a genuinely different register from posting. Shorter, reactive, generous.

  • Article Voice. Long-form structure, section rhythm, how you carry an argument across 1,500 words.

Writing with AI vs Human Writers

AI is our enabler, and can be pretty helpful while at the same time push us to the wrong direction.

I use AI for drafting, expanding my thoughts, and improving the hooks, positioning, and keeping on brand.

There are three major problems (at least) to keep in mind:

Blindly trusting AI:

Letting AI write & post content about you without reviewing it should be a sin!

If you can’t read your own content, how do you expect other people to?

At the very least, read the content before posting, and I recommend reviewing the data, information, and doing a good editing.

Expertise vs silver-tongued liars:

AI is a silver-tongued liar, when you use AI to do something you have no expertise at, it can backfire spectacularly.

Prefer to use AI to deliver more and better content under your own expertise, content you can review, understand, and scrutinize.

You can work to expand your domain of knowledge, or support you in ways you might need another expertise, but you shouldn’t trust it… the further away it goes from your expertise, the more dangerous it can be.

Outsourcing thinking:

There is a fine line between having AI help, and outsourcing our ideas and thoughts to it. Our brain is a muscle and should be battle-tested and trained, not let to atrophy.

They send updates and call it nurturing.

It’s a start, but not your end-goal.

And the difference shows up in pipeline.

In this section, we will go through an example sequence and how it can help you start actually nurturing leads into clients.

Updates vs nurturing

A monthly update is a one-to-many announcement.

It helps engage and stay top-of-mind, but it’s not enough to move the sales cycle quickly enough.

A nurture sequence is lifecycle-aware.

It branches on

• what the contact did,

• when they did it, and

• what stage they’re in.

A contact that downloaded a gated report last week is in a different conversation than a contact that requested a demo two months ago and went silent.

What an Actual Nurture Sequence Looks Like

(diagram above)

This is an example nurturing, I have done before. You can copy it, but consider it more like a guideline to better understand your options rather than a strict “perfect” model.

A few things to discuss about:

• Three entry lanes, based on how the contact joined the list. Demo requests get one motion. Disqualified contacts (people who showed intent but are not ICP-fit) get another. Lead magnet downloads and general subscribers get a third.

• Each lane has its own qualifying logic before it converges. The demo request lane sends a case study, waits 7 days, then checks if the contact is eligible for a an offer. The lead magnet lane checks whether the contact already downloaded the main asset, because there’s no point sending it again.

• A shared track picks everyone up. After lane-specific work, all three columns converge at a free resource / community email. From there the branching continues on actual contact behaviour. Did they download Lead Magnet #2? Did they download the gated report at the end?

• Two exits. A cold exit, with re-enrol after 30 days for contacts that never converted. A hot exit, with handoff to sales the moment they download the gated report.

This sequence does two jobs.

1. It nurtures existing demo requests back into conversation.

2. It moves cold leads into new demos.

Important Checks

The hardest part of nurturing is to be mindful of leads, stages, and conversations happening. For example, sales often does not like marketing sending too many emails to their hot leads.

A real nurturing sequence has to know:

• Is this contact a current customer? If yes, don’t send them prospecting content.

• Are they currently in a sales conversation? If yes, drop the sequence. Sales owns the relationship now.

• Did they book a demo since they entered the sequence? Stop the sequence.

• Are they an active opportunity in CRM? Pause until the opportunity closes (won or lost).

Own the Channel, Own the Communication

By owning the lead’s email you have undisrupted access to their inbox. You have to use it without abusing it to get them to buy.

A nurturing sequence does a lot of heavy lifting, and it’s difficult work.

It also requires ownership, auditing often, and checking whether it’s working or not.

A few more pointers to keep email comms working well:

• Lifecycle stage hygiene. Half the CRMs I’ve audited don’t have clean lifecycle stages or correct notes. A clean CRM, updated stages, and removing dead leads is important.

• Segmentation discipline. “Demo request” and “lead magnets” should be different entries. Understand how to treat different people, without ignoring them. Your cold lead today may be a future client in 2-3 years.

• Marketing-sales handoff. When sales takes a lead into a deal, marketing should know. Most companies don’t have this loop running, so sequences keep firing into active deals, this kills the marketing <> sales relationship.

• Ownership. Clear ownership of a CRM / Email manager is important. This is a responsibility and needs to be re-visited often!

So teams default to the safe option.

A monthly newsletter goes out to everyone.

I love newsletters, it’s your first step.

But, never stop at the first step, no one won a marathon with just one step. You need to go the full way!

A Final Note

A monthly newsletter is a fine baseline.

I’ve written before about why email is the channel you actually own, and that hasn’t changed. Owning the inbox is the right way to go.

But owning the inbox is not the same as nurturing the contact.

A newsletter going to everyone is a brand signal.

A sequence branching on contact state is a pipeline signal.

If your team is sending one thing to everyone, you have brand.

If your team is building nurturing into the pipeline, you got an accelerated pipeline with more ARR coming up faster.

This was my story as the first marketing employee at LearnWorlds, a startup of around $100k ARR when I joined that grew to $1M ARR before the second hire.

In this speech, I talk about growth hacking, and how startups create “luck”.

Because, growing a company has many ingredients:

• Talent,

• Timing,

• Speed, and a bit of

• Luck

And, all those should be wrapped around a great product.

P.S.: Subscribe to the newsletter, and you will get the presentation of the event as well!

Two Growth Spurts in 2017

For more than a year I kept trying to reverse-engineer what actually grew LearnWorlds in 2017. I kept checking our analytics, dashboards, data… for anything to explain our x4 growth that year.

It came down to two things, and they each accounted for roughly a quarter of revenue that year.

Affiliate #1

The first was a listing in Jeff Cobb’s Learning Revolution.

He was one of the top affiliates in the space, ranked first in Google for people searching for an LMS or course platform. He put us fifth on his list.

That single placement drove about 25% of LearnWorlds revenue in 2017, and you could see it in the data. The affiliate clicks, the mirrored lift in direct and organic that you could not attribute cleanly but lined up too perfectly to be anything else.

That looks like luck.

It was not.

Emails, LinkedIn messages, comments on his content.

There is a comment from Panos two years before I even joined, asking to be added to the list.

Investing in Marketing

The year it finally happened was the year we were also investing in the team and in marketing across the board.

The opportunity was created.

The second 25% came from a Forbes article that landed around the same time.

Same effect.

Direct and organic traffic moving in a way attribution could not explain, but the timing was unmistakable.

The COVID Effect: Timing

Then COVID hit, and everything quadrupled. Leads, trials, revenue. Every company in the world suddenly needed to sell courses or run internal training online.

We had spent two years on SEO with Minuttia.

By December we had about 80% of the SEO plan executed and LearnWorlds was ranking first to fifth on most of the keywords that mattered.

When the market quadrupled, that inbound organic traffic just landed on us.

If we had not done those two years of unglamorous SEO work, we would have lost at least half of the industry’s growth, or paid affiliates the same money to recover it.

If you only take one operational idea from this: demand generation and demand capture are separate disciplines, and the second one has to be built before the spike, not during it.

Product Back-end Story

You might think marketing generates growth.

However, up until November 2019, our CTO was creating every single trial by hand. A request would come in, and within 24 hours he would manually create the account, send the email, and walk people through.

Enormous friction.

If that process had still been manual when COVID hit, he would have been burnt on the first month, and we couldn’t have served the clients.

That bottleneck was removed 4-5 months before the lockdowns started.

Speed is a Growth Hack

When you are small, speed is your only structural advantage against larger competitors.

Doing more things, faster, means more opportunities.

More opportunities means more chances that one of them goes big.

Most will fail.

If you are not comfortable with chaos, failure, and experimentation, growth hacking is not for you.

Working with limited data, being fast, and spending budgets on big risks.

This is also why I push back hard on the “great product sells itself” belief, especially with technical founders.

A valuable product with no promotion and no distribution does not reach its potential.

The team and the execution are what investors actually evaluate, and they are what turns a good product into a growing one.

Effort is the moat now, not the product spec. I wrote about why the bar moved from volume to effort here:

Some Growth Hacks are Internal

A quick story from Cyberbit.

It took me five days to onboard a freelance agency I needed for performance.

A colleague told me that normally takes two months because of legal and approvals.

To me, five days was already an eternity.

I was used to two emails and done.

Recently I spent three days fighting finance to approve software we need, specifically so we could start that work on Monday instead of in April or May.

This is actually growth hacking for B2B, being able to move faster.

Side Story from a Friend

A client they work with at a big SaaS, in order to go faster often approved smaller projects and budgets by his personal one.

$1k-$2k add-ons on services per month did not need approvals by the CFO.

This allowed them to move fast.

When the time for the invoice came, the CMO had to defend the decision and explain the invoice. Due to having enough internal political capital, it took a couple of days to greenlight and a personal chat to the CFO.

Move fast first, fix it later.

Bottlenecks, Growth Hacks, and Startups

Luck is real.

Speed is important.

It’s all in the recipe for success.

Building luck through quick actions is important. Just, keep in mind that if you don’t have good decision-making or taste, speed is going to kill your business faster.

The blade cuts on both sides…

P.S.: Subscribe to the newsletter, and you will get the presentation of the event as well!

And… it works!

That’s why pricing hacks are dangerous.

They become addictive.

In this newsletter, I will be talking about one of my most proud moments that I ended to regret…

The Discount Trap - How LearnWorlds Almost Became A Discount Brand

At LearnWorlds, we were running shy discounts.

10% off (monthly).

Sometimes 20% off (monthly).

They didn’t work well.

The yearly plan came with a 20% discount baked in.

That worked.

But the buying decision was heavy.

You’re asking someone mid-trial to commit 12 months upfront based on a feeling.

The Hack

I pitched Panos, our CEO, on a different test.

Take the monthly plan:

Slash it 50% for 3 months.

The customer pays less upfront than two months at full price.

We collect 50% revenue for 3 months, it’s a lower discount than 20% yearly, but looks much bigger.

Emotionally the barrier to entry dropped.

It worked. VERY WELL.

So we ran it again. And again.

Every time the numbers dipped, we ran the promo.

Every other month, like clockwork.

The dashboard looked beautiful. Stable, growing MRR. Investors happy. The board happy. The team happy.

The Downside

What changed was how customers perceived the brand. Users became educated to the lower price, and looked to purchase when the discount was there.

LearnWorlds became, in the eyes of a meaningful chunk of new customers, a discount brand (kind of). In reality, this kept us from increasing the price.

Value for money, is often a discount brand positioning.

Would you do it again?

Yes! Discounts are part of marketing tactics.

But, they should be done carefully, feel special and rare.

A recurring discount becomes your SaaS’ real price into the eyes of the customers, and blocks you from increasing prices over time.

Freemium

Freemium is the most PLG motion you can have.

Most SaaS categories come with Freemium, the loss-maker to create an addiction to the buyer.

But, if you can’t convert those free users into paid, then you are burned 🔥

For a $0M to $10M ARR company copying that playbook without those underlying engines, freemium is a support tax with a vague payoff.

When freemium works:

Freemium works when you:

• have low costs and need to attract huge crowds.

• can build funnels and conversion mechanisms.

• can lock users for the long-term.

• can burn money to grow

Lifetime Deals

The AppSumo-style lifetime deal is the most efficient cash injection in early-stage SaaS.

Run a campaign.

Collect $200k in a week.

Book six figures of “revenue” without a single sales call.

The danger

Then live with the consequences:

• Every LTD customer is on your platform forever.

• Support costs.

• Server costs.

• Gross margin on that cohort approaches zero by year three.

The upside

LTDs can bring money in, but you need to be smart with it and include hard limits.

You need to study your costs and what platforms like AppSumo will ask.

You need to consider how this virality will pay off.

You need to promote well…

Lifetime deals are great to get a quick cash injection and increase customer base to grow faster, but come with many traps along the way.

The Annual Prepay Lock-In

The holy grail of SaaS may be yearly and multi-year deals.

Most SaaS offer a 20%-30% discount for annual prepayments.

The upseides

It’s a value for money for both you and your customers.

• Better retention.

• Better forecasting.

• Better planning.

The downside

Your whole revenue depends on yearly renewals.

And, you have less engaged users that are more likely to churn on renewal.

Price increases after the first year are harder.

Grandfathering, Done Right

Grandfathering is the one pricing hack on this list I love.

Loyalty pays off, both ways

When you raise prices, your loyal customers shouldn’t get hit with a 20% increase the next morning.

They bought the product at a price.

They built a business on it.

They might be loyal, or got the promise of the price.

You have to give users at least one year since they bought for grandfathering.

I recommend up to two years for your first / loyal customers is enough to show some loyalty back.

The challenge

Most SaaS teams make is treating grandfathering as a permanent shield or upgrade everyone regardless.

Both practices are destructive.

On permanent grandfathering, you are potentially losing money.

On upgrading everyone shows lack of loyalty to early customers, and that might create negative word of mouth and high churn.

What to do instead

The clean version of grandfathering looks like this:

• Hold the old price for 12 to 24 months after a price change.

• Tell the customer at the moment of the change, in writing, that the protection has an end date.

• Stage the migration so by month 24 they’re on current pricing without a renegotiation circus.

You’re offering more now.

The pricing should reflect it.

What This Is Really About

When you try to “hack” pricing, you are opening yourself to big challenges.

The question isn’t whether a pricing hack moves the metrics.

It can also backfire wonderfully.

Pricing is positioning.

Every hack repositions you in the market, whether you intended it or not.

A discount once is a test.

A discount every other month is a brand positioning.

The teams that win the pricing game aren’t the ones who never run hacks. They’re the ones who know when to stop.

Final Words: A/B Test

And, remember…

Pricing is fluid not stable.

You should change prices and A/B test them every now and then.

It might sound dangerous, and you are “locked” by competitors.

It’s also essential.

If you test, you might lose some revenue.

If you don’t test, you leave money on the table.

Choose your poison!

This was not a marketing call. I booked time on his calendar to evaluate Slate as a potential buyer.

The conversation that follows is what a senior practitioner sounds like trying to figure out whether a piece of software

• actually solves a real problem at scale,

• what it costs, and

• who it’s actually built for (ok, this is more professional curiosity).

About Shiyam & Slate

Shiyam runs TripleDart, a growth marketing agency for B2B SaaS, and built Slate (slatehq.com) on the back of the content automation problems his agency was already solving for clients.

The product solves problems content teams actually have, first for the team at TripleDart.

It is an automation tool with pre-built workflows on content automation, updates, and content management at scale, with a human-in-the-middle approach to AI.

In this episode, you will hear us discuss:

• What Slate actually is (and what it isn’t)

• The live demo

• The differentiator from tools like N8N and Make

• The credit pricing model, what $500 actually buys, and how Slate set its prices

• Who Slate is built for (Spoiler: 1,000-page minimum sites)

What Slate Actually Is (and What It Isn’t)

The first thing Shiyam wanted to clarify on the call was the positioning, because it gets confused often. Slate is not an AI search tool.

It is not Profound, Peec, or Athena.

It is a content automation platform built around composable AI actions, what Shiyam calls “nodes.”

You build agents.

The agents run sequences of nodes.

• The nodes generate images,

• pull keyword data from Ahrefs and Semrush,

• take screenshots of brand pages,

• write articles,

• do internal linking,

• do external linking, and

• publish to your CMS.

It’s like Clay, but for SEO and content workflows. You build a workflow once, then call that workflow as a single block inside another, bigger workflow.

The Live Demo: From Brief to Published Article

Shiyam walked through a real workflow built for Guideflow, a European SaaS company. The chain runs like this:

• Take an input brief

• Generate a hero image with GPT-image

• Run a fact-checker node to find fresh statistics on the topic

• Extract brand URLs from the brief

• Take screenshots of those brand homepages

• Map screenshots to brand mentions inside the article

• Write the full article

• Convert to clean HTML, flatten to a single file, generate a slug

• Publish directly to Webflow under a new section of the blog

The end output is a published article with original images, embedded screenshots of competitor products and internal links.

How is this different from N8N or Make.

First, the blocks themselves.

Slate has SEO-native nodes that don’t exist in N8N like Ahrefs, Semrush, citation formatting, web search with AI overviews, schema generation, internal linking, external linking, content humanizer, Reddit scraper, G2 reviews scraper.

The general-purpose tools can technically do this, but you’re building it from scratch. Slate already has the functionality built-in.

Second, every block is structured to take a prompt for both the AI agent and the end user. Slate’s blocks have a dual prompt by design with instructions for the system, and instructions for the human reviewer.

That distinction matters more than it sounds. I tend to prefer workflows with the human-in-the-middle-loop.

Bulk Content Refresh and Internal Analysis

The use case that turns Slate from “interesting” into “I would actually buy this” is bulk content refresh.

If you have 10,000 pages and traffic is declining on 800 of them, the manual job is unbearable. Slate’s content refresh agent ingests an existing article, runs the analysis, and produces a refreshed version with modifications highlighted.

1. Yellow for sections it changed.

2. Green for sections it added.

3. You review, approve, and publish.

Razorpay, an Indian fintech directly competing with Stripe, has roughly 10,000 pieces of content and uses Slate to refresh 25 to 30 per week.

Atlas, a global payroll company, uses it to monitor every country’s official visa site and update their content every time visa rules change.

SentinelOne uses it to maintain a daily-updated repository of new cyber attacks.

Agentic Workflow Building

The internal analysis side is where Slate gets ambitious. Shiyam asked the platform a single question on the call: “Where am I losing to competitors on listicles in the last 30 days?”

The system pulled brand-level SEO metrics, gathered competitor listicle pages, ran a diagnosis on why specific pages were underperforming, and recommended refresh-or-rebuild for each one. Then he asked it to fix the problem. It built the refresh workflow on its own.

Pricing, Credits, and Who Slate Is Built For

This is where I pushed hardest, because Slate’s pricing isn’t on their site in a useful way and I wanted the practitioner answer.

Slate uses a credit model.

Pricing

Workflows and agents (the main product)

• Solo: $500 for 50,000 credits → roughly 30 to 40 content creation or refresh tasks, including image generation

• Pro: $1,000 for 100,000 credits → more credits, more users, plus access to Sanity, Contentful, and additional CMSs (Solo only covers WordPress and Webflow)

• Scaling up from there: $2,000 for 200,000 credits, $3,000 for 300,000

Prompt tracking (separate product, for AI search visibility monitoring)

• 100 prompts × 4 platforms × 4 weeks = roughly $199 to $299

I asked Shiyam how they set the price. The answer was unusually direct.

“Purely on our margins.”

Slate isn’t priced based on competitor benchmarks or willingness-to-pay testing. It’s priced as cost-plus on the underlying API and data fetching costs.

Not a Self-Service SaaS

There is no self-service.

I tried to use Slate before the call.

It does not work without setup.

The white-glove model is built-in to the system.

The company sets up your workspace, configures your CMS connections, and helps you build your first agents.

The Target Market: Enterprise

The target customer is mid-market or enterprise, with a minimum of 1,000 pages on their site.

Average ticket size: $10,000 to $25,000 a year. He has roughly 78 independent customers plus 20 from TripleDart’s agency book.

If you have 200 blog posts and traffic is fine, this is not for you. If you have 5,000 product pages and a content refresh problem nobody on the team has time to solve, this is exactly for you.

Takeaways

Three things stayed with me from this conversation.

Slate solves a real problem for the agency.

Pre-built AI workflows to create and refresh content at scale.

This is for big websites.

This is enterprise SEO automation, sold to teams that already have a content engine, and thousands of pages.

The pricing methodology is going to change.

Cost-plus margin is a good way to start, but I expect the pricing to change as it moves forward. Already, the credit limits have been expanded.

I prefer gaming on the couch after a long day at work, which includes reading, writing, and communicating.

That’s why most books I “read” are Audiobooks - rarely physical, sometimes ebooks.

Audiobooks & podcasts are my preferred company while driving, walking, or cycling. Adding on the empty space, and reducing unwan…

Read more

Mostly because I followed them, and my friend talked so well about them, I wanted to meet them.

They invited me on to talk SEO, growth, in-house politics, attribution, and AI.

We went long. Close to an hour.

And, enjoyed a chill talk between long-time professionals.

About the Hosts

Sara Taher is an in-house SEO lead, newsletter writer, and course creator. She teaches Python for SEO and runs her own newsletter where she writes because she wants to, not because she’s trying to sell something.

Nick Vujic runs a growth agency focused on organic marketing for early-stage B2B startups (zero to ten million ARR is his sweet spot). He’s the agency voice in most of our exchanges.

Two of us are in-house. One is agency.

In This Episode, You Will Hear

• Why I moved from SEO to Head of Demand (and why it never was just about SEO)

• How to survive in-house politics when your 30-page audit gets 3% accepted

• Why technical founders are the hardest founders to sell SEO to

• The case for email marketing over almost every other channel for early-stage B2B

• Why I think attribution is dying and why that’s fine

• The side project rule: why in-house marketers need something with no limitations

• An honest AI take from a skeptic (and the Claude moment that changed my mind)

Why I Chose Growth Over SEO

I get bored.

If I’m forced to do one thing, I lose interest. Always, my role was about growth… people knew me because I talked about SEO.

At LearnWorlds, even when my title said Head of SEO, SEO was only about 30% to 50% of my actual job.

I was opening channels. Facebook and remarketing in 2017 before most B2B companies were spending that much there. Affiliates. PPC. Email. I was doing inbound sales with the CEO until we hired for it.

» Then I moved on to the next channel.

Channel agnostic is the right frame.

SEO was a massive lever for me, but it was one lever. If the channel that works for your business this year is paid social, do paid social.

If it’s email, do email. The goal is pipeline. Not proving SEO is the best channel.

🏆 Author’s note: Email marketing is still my favorite. Almost every early-stage B2B SaaS I’ve consulted for never emailed their database once. A basic newsletter activates a base you already paid to acquire.

The Survival Guide for In-House Politics

In-house marketing is 60% doing the work and 40% defending it. I am lucky in my current role that’s less the case, but it’s still part of the job. While, I heard horror cases with marketers spending 90% of their time defending their work.

If you want to survive long enough to ship anything, you have to play with the politics, it’s unavoidable as companies grow.

Pick your fights

I don’t fight the brand team for the homepage H1.

That’s their baby. I ask for one or two dedicated SEO pages where I have full control. They keep their brand positioning. I get my traffic. Everyone wins.

Learn the language of the person in front of you

CFOs want revenue numbers.

Developers want clear, concise instructions.

If you tell a developer “make the site fast,” they’ll ship you a JavaScript-heavy site that’s technically fast but Googlebot can’t crawl. Be specific.

Find the gap

I waited one and a half years for a simple WordPress checkbox fix. I caught a developer on a Friday when he had some downtime. He fixed it in 4 hours.

Rapport, patience, and building relationships with the right people wins battles.

Author’s note: Sara made a great point here that I want to reinforce. When a stakeholder blocks a small thing today, your next suggestion gets harder. Technical SEO debt compounds. So does political debt. Get the small wins early. Earn the capital. Then spend it on the bigger fights.

Why Most Founders Shouldn’t Start with SEO

There are three kinds of founders.

The technical one, who only thinks about product and numbers.

The seller, who goes out and closes deals.

And the promoter, who understands marketing.

The third one is rare in tech. When you find one, they usually grow their companies fast.

Technical founders love numbers. But SEO gives them uncertainty in the short term, and they don’t like uncertainty.

So they default to product.

Features over pipeline.

Every time.

If I’m walking into an early-stage B2B SaaS cold, I usually tell the founder don’t start with SEO.

• Start with direct sales.

• Hit doors.

• Go to events.

• Send LinkedIn messages.

• Build an email list.

• Run a newsletter.

Get ten customers who can tell you what they actually want. Then think about scaling a channel.

Nik Vujik said his agency has built attribution models around SEO that tie back to pipeline, and for B2C and small B2B that absolutely works. But for enterprise B2B with 50 to 150 touches per deal, trying to attribute a sale to one SEO page is a fool’s errand.

Which leads to the next point.

The Death of Attribution

Here’s my controversial take.

Attribution is dying, and I’m fine with it.

Between GDPR, cookie blocking, the messy middle of the buyer journey, and enterprise sales cycles that hit 150 touches, pinning a closed-won deal to one channel is a fiction you sell to a CFO so you don’t get your budget cut.

The reality:

• Self-attributed data is honest. When we surveyed Cyberbit customers, most said “I found you on the internet” or “I googled you.” Not helpful for a dashboard. Very helpful for understanding reality.

• Multi-touch attribution models are a defense mechanism. If 95% of your energy goes to defending every dollar you spend, 5% goes to actual marketing. That’s not growth. That’s survival.

• Fundamentals work. Write good content. Send good emails. Show up at the right events. Talk to customers. The basics outperform most attribution dashboards.

Do the work. Trust the fundamentals. Stop trying to prove every click.

Executing fast will get you the wins needed to grow.

Author’s note: There’s nuance here. If you’re in B2C or small B2B with short cycles and few touchpoints, attribution is doable and useful. Nick’s Zapier-based touchpoint tracking for his ten-to-twenty-thousand-visits-per-month clients sounds genuinely useful. I’m not anti-measurement. I’m anti-wasting-your-life measuring the unmeasurable.

The Side Project Rule

Every in-house marketer needs a side project with no limitations.

In corporate, product says no. The CEO says no.

You can’t get budget for link building.

You can’t test that tool.

You can’t publish that piece.

You stop learning because someone upstream decided you shouldn’t.

A side project fixes this.

Newsletter — Help grow this one ;) it’s my side-project.

Personal website.

Chrome extension.

Whatever. It forces you to stay sharp and keeps you learning at a pace the corporate job can’t.

My side projects are the newsletter you’re reading right now, my LinkedIn presence, marketingexpertshub.com, and LibrarIn (a Chrome extension for saving and categorizing LinkedIn posts, currently in review for the next version).

Sara has her courses and her own newsletter.

Nick has his agency and is pushing into building his own product.

The pattern is the same. In-house gives you the real-world problem set. The side project gives you the sandbox to figure out how to solve it.

An AI Take from a Skeptic

I’m an AI skeptic. I’m also a Claude user.

These two things aren’t contradictory.

I don’t think AI is replacing marketers.

I think it’s Zapier on steroids.

A productivity tool that saves me a few hours a week.

I use it to write regex for Search Console, fix code snippets, draft first versions of emails, and build small tools with Lovable.

For months I stayed on Gemini because I was already paying for Google. Then I went to Greece, talked to a bunch of people who wouldn’t shut up about Claude, and finally tested it.

Four hours in, Claude fixed a Chrome extension I’d been fighting with Lovable on for two weeks.

Author’s note: Sara made the right cautionary point here. AI is great until you stop checking its output. I’ve seen people ship work with obvious hallucinations because they trusted the first draft. AI is a productivity tool. It is not a replacement for thinking. Treat it like a very fast, very confident junior.

The Takeaway

Three things stuck with me from this conversation.

In-house politics

If you work in-house, you’re going to lose fights. Pick the ones worth winning. Drop the rest. Build enough rapport and wins that when the important one comes, you have the political capital to take it.

No attribution is perfect

Stop waiting for perfect attribution. Do the work. Trust the fundamentals. Measure what you can, ignore what you can’t, and spend your energy on marketing instead of defending marketing.

Side-projects are the best learning experiences

Have a side project. One with no limitations. It’s the only way to keep learning when the corporate world says no.

What’s one fight you’ve decided to stop picking with your product team to save your sanity?

Let me know in the comments.

Generative AI raised the floor on content. It also raised the bar.

Most content fails now. Everywhere.

Google, AI search, LinkedIn — the same posts, the same takes, the same recycled frameworks rewritten by the same tools.

And the people still winning all have one thing in common: they’re doing the work nobody else wants to do.

Read more

...and almost no one actually fixes.

For this episode, I sat down with Antonis Askianakis, Head of Sales at LearnWorlds, and a former colleague of seven years. We worked together at LearnWorlds, had real disagreements, and came out of it as friends. That’s the only kind of relationship where this honest conversation is even possible.

It’s what the conversation actually sounds like between two revenue people who trust each other enough to be honest.

In this episode, you will hear us discuss:

• What sales actually wants from marketing (and why “more leads” is never the full answer)

• Why misalignment is almost always a leadership failure before it’s a team failure

• The MQL/SQL debate — and why shared definitions matter more than the labels

• Why marketing’s responsibility doesn’t end when the lead is handed off

• The dual-pipeline target fix that forces both teams into accountability

• When and why reducing marketing touches mid-deal is the right call

The Real Ask: Alignment, Not Just Leads

We started where most of these conversations should start — what does sales actually need from marketing?

Both teams share the same goal, even if they sit at different parts of the funnel. What sales is asking by “more leads”, “better quality” might be better alignment.

The funnel feels linear, but it isn’t. And when marketing’s North Star metric diverges from sales’, you’ve already lost.

The fix sounds simple:

• agree on definitions,

• set common goals,

• write it down.

Alignment is a Leadership Problem

We spent a lot of time here, and it’s worth spending time on.

When two teams are fighting over lead quality, pipeline ownership, or CRM hygiene, the instinct is to fix the process. The problem is around accountability and leadership.

When leadership doesn’t take a clear position on who owns what it creates a blaming circle.

Marketing blames sales for not following up.

Sales blames marketing for bad leads.

Nobody fixes anything because nobody has to.

The only sustainable fix is a shared revenue goal with clear ownership of each stage. Whether that comes from a CRO, a combined commercial team, or two directors who decide to align themselves, somebody has to draw the line.

Author’s note: At Cyberbit, the first thing I did was sit down with sales and agree on definitions. Not to be bureaucratic about it, just to have something written that both sides owned. It didn’t solve everything. But it removed the most common arguments.

The ICP and Lead Quality Problem

Experienced AEs have a warm network, a preferred industry vertical, and leads they can close with high certainty. When marketing brings in leads from a different profile, the easiest thing to do is ignore them.

That’s not a sales problem — it’s a trust and data problem. If an AE consistently ignores a lead type, there’s a reason. Either the leads genuinely aren’t converting, or the AE hasn’t been given a reason to trust the process.

The right response isn’t to complain about it in a marketing meeting. It’s to ask:

“Can you walk me through why these three leads didn’t go anywhere?”

That conversation usually reveals more about your ICP than six months of analytics.

The Pipeline Target Fix

The most practical takeaway from the episode.

Marketing tracks pipeline. Sales tracks pipeline and revenue. When the only metric marketing owns is top-of-funnel, the incentive to stay involved after handoff disappears.

Antonis’ suggestion: give marketing a pipeline target that extends deeper into the funnel. When marketing has a stake in whether those leads convert, not just whether they arrived. Then, the dynamic changes.

Marketing starts asking why BDRs aren’t converting the leads.

Sales starts explaining what’s actually happening in the conversations.

Both teams start working on the same problem instead of arguing about whose fault it is.

When Marketing Should Pull Back

Not everything calls for more marketing involvement.

In complex enterprise deals — where buying committees are large, relationships are fragile, and timing is everything — an unsolicited marketing email to the wrong stakeholder can derail a deal in progress.

Antonis’ approach: reduce direct touches at the deal level, but don’t disappear.

• Brand awareness,

• retargeting, and

• content still run in the background.

The distinction matters.

Marketing that stays visible at the brand level creates the familiarity that makes future deals easier to close.

Marketing that steps on an active deal creates friction.

Wrapping Up

The closing line from Adonis landed the clearest:

That’s it. That’s the whole job.

Not leads for the sake of leads. Not MQLs to hit a target. Bring marketing and sales close enough to their total addressable market that the right conversations happen with the right people at the right time.

Everything else — the definitions, the targets, the processes — is just scaffolding around that one idea.

The internet, helped measure everything:

• clicks,

• impressions,

• view time,

• scrolls, etc…

…

Read more

A user at their started plan will pay $10.99 per user, per month.

That would mean…

• A small team of 5 would pay: $54.95 / month

• A medium company of 20 would pay: $219.80 / month

• A big company of 1000 would pay: $10,990 / month

Costs rises with the company, becoming a non-insignific…

Read more

Going into 2026, most teams have tested AI to write content and writers have either added it to their workflows or position themselves as 100% human writers.

For this episode, I sat down with Pawel Tatarek, an experienced B2B Content Writer & Editor to discuss how content marketing has changed.

In this conversation, you will listen to

• Attribution for content (and how to sell it internally),

• Human vs AI,

• Key trends for content in 2026,

• SEO vs multi-channel content, and

• Data-driven, thought leadership content.

AI search is changing the game.

Minuttia helps B2B companies dominate search—not just traditional SEO, but AI-driven discovery too.

• Develop AI + Google search strategies

• Create high-quality human and AI content

• Amplify visibility through Digital PR

• Continuously improve with agent analytics & reporting93% of our clients stick around.

If you want to win on AI search before it’s table stakes, talk to Minuttia.

Get in Touch

The Attribution Nightmare & Value of Content

Content has always had SEO and Social Media traffic to help assign attribution to content efforts. Today, proving the value of content is becoming much more difficult.

CMOs, CFOs, CEOs, and the Board are asking the ROI of content.

How do you prove a content writer’s worth with performance metrics, if visits to the website from organic channels declines?

Author’s Note: In my opinion, we are going deep down in a low attribution world, where every channel struggles to attribute results.

Here are a few ways to prove attribution:

• Self-attribution - adding a step for users to self-attribute where they found you.

• Assign value to other metrics - for example impressions, shares, mentions, etc.

• Leads - lead magnets can still provide value through the numbers of leads attracted, or leads that came from other content pieces.

Author’s Note 2: The best CMOs are not the ones who measure performance at every turn, they are the ones who follow their guts and take risks to non-attributable channels they know will work, but able to defend their work to CFOs with overall business performance.

Human vs. AI Writer

Through the conversation, we both agreed AI has a place in the content writing process.

• Supporting on research

• Drafting

• Spinning, or rewriting

Even some top-of-funnel content can be mostly written with AI, but would need a human editor to ensure quality.

But, the lower you go into the funnel, the more SMEs are important.

Writing convincing copy with industry expertise is something AI cannot be trusted to do. Being 90% accurate does not cut it, especially in specialized businesses where the client expects industry knowledge.

But… there are two “armies” the pro-AI and the anti-AI. Pro AI business owners will always prefer to not pay for human writers, no matter the cost or arguments.

If you are a B2B Writer, go towards the route of least resistance. Work with people who believe, trust, and support your craft!

2026 Content Trends

We also discussed a few trends we see emerging around content:

• Shift to Data-Driven & Thought Leadership: The demand is moving away from generic SEO posts toward “Data-Driven Content” (original research studies) and deep thought leadership. This is the only way to stand out in a sea of commoditized AI content.

• Repurposing is the New Standard: You can no longer write just for Google. Content must be repurposed for newsletters, social media, Reddit, and YouTube to be effective.

• Personal Brands: The founder’s brand, employee advocacy, and thought-leadership from the company on social media is the way to go. Especially today, where user generated content (UGC) is more important to get beyond the AI hype.

• Visuals & video: With AI overviews (like Google SGE) citing images and videos more frequently, visual assets are becoming critical for SEO. Content marketing has to evolved and adapt fresh new formats that artificial intelligence has more trouble creating.

As we look toward 2026, the hype cycle is breaking.

Trust becomes the expensive currency for marketers, brand is how you grow that currency.

As every channel is saturated and AI content has commoditized even professionally created content, we (marketers) need to find better ways to distribute our work.

High competition and AI-generated noise ca…

Read more

Usually, because of fear. As the company grows and ownership, regulations, and the email list all grows together.

But, that’s essential to keep being top-of-mind for your audience.

Why Tech and B2B Fail in Email

…

Read more

An honest, bare truth moment.

These are the good, the bad, and the tactics you can use to do it too.

The secret?

• Consistently giving value,

• engaging, and

• connecting with people … emotionally,…

Read more

No longer.

Vibe coding tools are here to help you make:

• simple calculators,

• data visualizers, or

• niche name generators…

…that would be an absolute perfect lead magnet for your ideal customer.

We sat down with Zacharias Xiroudakis and George Fakorellis to guide you through the process of building your first mini-app.

Join Growth Vibers

Before you go further, be sure to check our community of Vibe Code marketers!

We will be sharing the resources we talked about on the webinar inside the slack community:

• Best tools we tried

• Resources to take you further

• Our own experiences

• How to improve your skills in vibe coding

P.S.: I am not an AI-first marketer, I do believe in strong fundamentals and good knowledge first, then AI to help you build up and deliver on those. I am an AI-skeptic using the tools to test their potential and deliver truthful, fact-based, real uses.

You Don’t Need to Be a Developer (But You Need to Know the Lingo)

You don’t need to be fully technical or know coding.

But, knowing some basic coding and thinking more technically will go a long way to debug and correctly build what you need.

Zacharias takes us through the basics for building an app:

• Front-End: This is simply what your user sees and clicks on. The buttons, the text, the colors.

• Back-End: This is the server, the “brain” that does the processing. It’s where the logic happens.

• Database: This is just the “filing cabinet” where data is stored (like user info or content).

• Libraries: This is pre-written code that does a specific thing (like create a map or a chart). You don’t have to reinvent the wheel.

You don’t need to know how to write this code. You just need to know what these things are, so you can tell the AI what you need.

Stop Thinking “SaaS,” Start Thinking “Mini-Tool”

Complex, multi-feature SaaS products usually fail badly with vibe-coding. But, making a simple mini-app, chrome plugin, web-app, it’s achievable!

Mini tools can be:

• productivity boosters

• lead magnets

• SEO traffic gold-mines

These mini-tools are powerful because they:

• Capture High-Intent Demand: People are searching for “keyword generator” or “backlink checker.” Ahrefs captures that search, solves their problem, and introduces them to the main product.

• Showcase Your Core Value: The mini-tool gives users a small, free taste of the power behind your full, paid platform.

• Act as World-Class Lead Magnets: A tool that solves a real, recurring problem is something your audience will actually value.

Instead of trying to build a massive, all-in-one solution, focus on solving one tiny, specific problem for your ICP.

Define the Problem - Get Better Results

This was the most practical lesson of the day. You cannot just type “build me a cool map for my blog post” and expect a good result. The AI will make assumptions (or “go rogue,” as they said), and they will almost always be the wrong ones.

You have to be the project manager, not just the requester.

In the webinar, I showcase the example of building my mini-map for Country Tier Lists.

Knowing the terminology of “choropleth maps” and specific libraries like leaflet.js was the point where I truly unlocked the capabilities of vibe coding for the specific problem.

How it impacted it:

• “I need a Chloropleth map.” (The specific type of map)

• “Use the Leaflet.js library.” (The specific technology to use)

• “Use OpenStreetMap as the baseline data source.” (The specific resource)

The AI is a powerful intern. It can code incredibly fast, but it has no business context or strategic vision. You have to provide that.

The more specific your brief, the better your outcome.

Use the AI as a Guide

This might have been my favorite takeaway. When the AI builds something for you, it’s tempting to just copy the code and move on.

Try to understand what it’s building... Ask it.

Why did you choose to do it this way?

What is this back-end structure you’re using?

This transforms the AI from a simple factory worker into a 24/7 personal tutor. In my experience, this is how you build real, durable skills.

By asking “why,” you go from being a button-pusher to being a true builder.

Where to Find Problems Worth Solving

“But what do I actually build?”

It’s the most common question. Here are the answers:

• Solve Your Own Pain: This is the easiest one. Look at your own team’s tedious, manual workflows.

• Productize a Service: Go to Upwork or Fiverr. Find simple, high-demand gigs (like “I will build a basic mortgage calculator for your site”). If people are paying for it as a service, it’s a great candidate for a mini-tool.

• Search Reddit: Use search operators like site:reddit.com “is there a tool for [your niche]” to find what people are desperately asking for.

• Check G2 & Capterra: Look for software categories with high demand but low user ratings. Read the 1- and 2-star reviews. You’ll find a list of common feature gaps and pain points that you can build a mini-tool to solve.

Wrapping Up: Just Start Building

The main lesson here is that the barrier to entry for building useful tools has never been lower.

It’s not “easy,” and it’s not magic. It still requires you to think, to research, and to be a good project manager. But for the first time, you don’t have to wait for a developer to be available.

My advice? Pick one, tiny, annoying problem that you or your customers face.

Then, go try to build the solution yourself.

And… join our Growth Vibers slack community to upskill in Vibe Coding for Marketers: