🆕 The Promote Playbook

Hey there 👋,

In 2017 I had built FunnelEngine.com to $5,000 a month promoting ClickFunnels and SamCart. That year I finished in SamCart's top ten affiliates. Since then I have been a top affiliate for Odys and Content at Scale for a period each, and I am currently Work Hero's number one.

The software review site sold for six figures, and almost the entire valuation was the two affiliate accounts.

Not the rankings. Not the content. The recurring commissions attached to customers I had already introduced.

I spent three years thinking I was building a website. What I had actually built was two revenue streams and a 15,000 person email list that fed them.

Back then I built a mini-course called Promoting Software where I showed exactly how I did this, including video walkthroughs inside my affiliate accounts.

But every one of those customers still had to do the work themselves, exactly as I had. That is the part that changed. A lot of it can now be handed to an AI worker that does not get bored on step four.

I’ve just built out the Promote Playbook, plus a free Affiliate Grader tool which did a test run on Kinsta’s program, details in the BUILD section.

The trackers through mid-August: AI is explicitly cited in around 55% of 2026’s cuts so far, some 160k+ roles, with tech well past 125k for the year and May the peak month for AI-reasoned cuts. Product, engineering and content roles are taking the worst of it at the scale players. Companies aren’t trimming. They’re reallocating around agentic systems. Same pressure is already hitting freelancers harder — Upwork GSV and active clients down as AI automates exposed skills.

🛒 Etsy: 220 (12%), mostly product & engineering, restructuring into flatter teams for speed (AI noted as changing work, not the driver). Link

📱 TikTok: 250, full Nashville office closure (content moderation), streamline operations. Link

📊 VideoAmp: 50-60 (20%, including CTO), explicit pivot to agentic software and AI-powered media platform. Link

This week my workers read every live Flippa listing between $20,000 and $200,000 across ecommerce, SaaS, content, newsletters and YouTube. A hundred-odd of them. Ten made the board.

These are first glances, not teardowns. Most of what I threw out failed the same way it always does: sub-1x dropship stores where the traffic is rented and the seller knows the lease is up. I also dropped four that I showed you last time and are still sitting there unsold at the same price, and one where the seller's headline claimed 658 subscribers while Flippa's own verified panel said 141. I am not passing you a number I cannot stand behind.

Every price below comes off the listing page itself, not the search card, because on one of these the two disagreed by $37,000. [Each of these links are affiliate, the next section gives you a playbook on this.]

.

Two facts that sound like they cannot both be true.

US businesses will spend $13.81 billion on affiliate marketing this year, up 11.3% on 2025. Around 84% of ecommerce brands run a program. B2B participation grew roughly 17% last year, driven by SaaS.

And the way people used to get a share of it is collapsing. Google AI Overviews now appear in 25.8% of US searches, 82% of B2B technology searches. When one shows up, organic click-through falls 61%. Wirecutter, the best resourced review operation on the internet, lost over 60% of its search visibility in four months.

Then the merchants started quietly paying less for what was left. Amazon Associates cut commission rates by up to 50% through early 2026. Categories that paid up to 10% dropped to 4 and 5%. Milestone bonuses removed. Reporting degraded, so affiliates can no longer see which products drive sales. The April operating agreement ended halo sale commissions, so you now earn on the item you promoted and nothing else in the basket.

None of it was publicly announced. Creators reported income drops of around 25%.

Every one of those changes was permitted by terms those affiliates had already accepted.

So the structure to hold in your head: The click got cheap and the trust got expensive.

The money did not leave. What broke was the mechanism: rank a page, collect the click, pass it on, take a percentage. That was arbitrage. You were buying attention cheaply from Google and selling it at a markup, and the supply of cheap attention has been withdrawn.

What is scarce now is a person whose opinion somebody actually takes. An AI Overview can summarise a review. It cannot be someone who used the software for two years and will tell you which part is annoying.

Almost nobody reads affiliate programme terms. The terms are where the money quietly goes missing.

Paste in a programme and this grades it on the eight things that decide whether you actually get paid. Each one PASS, WARNING or FAIL, with the clause quoted back to you.

Attribution, and whether a browser coupon extension can take credit for a customer you found and warmed up. Cookie window, matched against how long your audience actually takes to decide. Whether renewals pay, for how long, and whether that survives an upgrade. How long they can claw a commission back. Payout threshold, schedule, and whether they can pay into your country at all.

Then a verdict, and six to ten questions to put to the programme manager in writing. A manager who will not answer them has answered them.

I built it because of the paragraph above. Every affiliate Amazon cut this year had already agreed to terms permitting all of it. Not one of them was ambushed in any legal sense. They just had not read the thing, because nobody reads the thing, and the terms only ever matter on the day the company changes its mind.

Run it before you point an audience at somebody's checkout, not after.

It closes on the bigger question: is what your audience buys earnable on at all? A 25% recurring commission on a subscription somebody keeps two years is worth roughly a hundred times a 4% one-off on a physical product. Same work. If your audience's spending is mostly one-off, low value or impulsive, it says so.

Ten workers, the playbook, a two-sheet toolkit. $299 with a year of Pro, same shelf and same price as the other three.

The Agency and Productize playbooks sell your own labour. Promote sells your judgment about other people's products. You never build it, host it, support it, or take the refund call.

The downside is you control nothing. The merchant can cut your rate on customers you already referred, change the product until it no longer suits the people you recommended it to, or close your account and keep the balance. Amazon demonstrated the whole sequence in public this year.

I ran all ten workers end to end, against one made-up freelancer with no list and no ad budget, to see what they actually do rather than what I intended them to do. Three results I did not expect:

The paid traffic worker did the sums and told her not to spend anything. She could afford £0.46 per subscriber. A subscriber would cost her £3.33. Short by 7.2 times, and it did not clear on a thirty-month payback window either. I wrote that gate because I did not have one. I built my income on SEO and email. Paid worked on the SamCart landing pages and most other times it lost me money. The diagnosis, nine years late, is that I was buying traffic before I knew what a subscriber was worth. SEO forgave that because the traffic was free. Paid turned the same ignorance into an invoice.

The negotiator told her she was asking the wrong programme. She wanted a better affiliate rate on four conversions, which is too few and reads as not understanding the business. What it found instead was that Kinsta runs a separate agency partner programme that qualifies on hosting under management rather than referrals, and her existing client accounts already cleared the middle tier. She had been eligible for three years and never looked.

The traction worker refused to find a problem. Four months in, no money to speak of, and the honest read was slow rather than wrong. What it named instead was that she had shipped nine articles and kept deferring the one worth the most, three times in a row, which is a scheduling problem wearing a strategy problem's coat.

And the part nobody tells you, which is why the document does not end at "keep promoting": recurring affiliate accounts are an asset, not income. They transfer in a sale. I have now sold those accounts twice, the software review site and the beehiiv account as part of the LetterOperators sale.

If you run WordPress sites, your own or your clients', and you are the person who ends up fixing them at eleven at night, Work Hero is what I would point you at. Flat monthly fee, you send them the job, it comes back done. They support over 1,200 sites for more than 60 agencies.

That is an affiliate link. I am their number one affiliate, confirmed by the founder Kevin, a long-time friend.

Work Hero is also the cleanest example of the offer type the playbook argues for. Recurring revenue, a professional audience , a researched purchase, and customers who stay for years. It out-earns things with far more brand behind them, because the shape is right.

Every single sale in the top ten of the latest domain chart was six figures, and .ai and .com took all ten spots between them.

This week alone: Perform.ai at $200,000, Kingdom.ai at $130,000, and on the 11th the day's top sale was a .ai at $136,500. On the 6th, Namecheap founder Richard Kirkendall published thirty .ai sales done through Spaceship Sellerhub, the disclosed prices running from $40,000 to $200,000.

Sitting above all of it, the number everyone quotes: AI.com sold for $70 million, the largest domain transaction ever recorded, more than double Voice.com's $30 million in 2019. Buyer was Kris Marszalek, the Crypto.com CEO, paid in crypto, brokered by Larry Fischer. The deal actually closed in April 2025 and stayed quiet until he launched the thing during the Super Bowl in February.

So the domain market looks like the best trade on the board. Here is why I am not taking it.

You are looking at the sales. Nobody publishes the portfolio.

Thirty .ai sales got posted. The number of .ai names being renewed that month by the same people did not. And .ai is not a cheap thing to hold: roughly $70 to $110 a year depending on registrar, against $10 to $22 for a .com. The wholesale price went up another $10 a year in March, a 14% rise, on names people had already bought.

Do the arithmetic on a real portfolio rather than a headline.

Hold 200 .ai names at $85 a year and your renewal bill is $17,000 before you sell anything. Domain sell-through runs at roughly 1 to 2% of a portfolio a year, so 200 names produces two to four sales. Meaning every sale has to average $4,250 to $8,500 just to cover the renewals. Not to profit. To break even. Before what you paid to acquire them.

You’re likely going to need to run outreach to sell enough of these names. Fortunately I have a great Domain Outbound Worker in the Domain Operator Pack who can do this for you. It’s been iterated on and battle-tested by a top domainer who only sells 6 and 7 figure domains.

Employed, freelancing, buying, or automating yourself out of what you already own. The community has a channel for each, and they're the same four sections you just read.

Cheers!

Richard

The obvious bit: nothing in this email is financial, investment, or legal advice. The deal reads are produced by AI workers from public listing data, unverified, and I hold no position in anything listed. Valuations are opinions, multiples move, and no verdict here replaces your own due diligence. If you're buying anything, verify the numbers yourself before money moves.