I Dug Through 8,782 TrustMRR Listings. The Revenue Number Is Usually the Least Interesting Part.

TrustMRR is dangerously fun.

Open the leaderboard. Sort by revenue. Feel bad about your life for six minutes.

Then notice Gumroad near the top with $7.1 million in the last 30 days. Easytools with $2.7 million. A supplement platform doing nearly $1 million. A micro-influencer product at ~$900,000. An API product at ~$470,000.

The instinct is to compare them.

That instinct is wrong.

Gumroad's public TrustMRR page calls the number GMV. So does Easytools. That is money moving through the product, not necessarily money the company keeps. Elsewhere, the same column may represent subscription revenue, a one-time document sale, agency retainers, physical-goods fulfilment or campaign spend passing through a marketplace.

This matters because a leaderboard can make ten completely different businesses look like one sport.

I pulled the frozen TrustMRR dataset from July 30, 2026: 8,782 listings and 29 fields per listing. Then I stopped staring at the aggregate and opened the interesting products. Their pricing. Their websites. What customers are actually buying. How closely the subscription count maps to MRR. Where the 30-day revenue diverges from MRR. What gets easier as they grow, and what gets harder.

The aggregate still gives useful context. About half the listings reported zero 30-day revenue. The median among revenue-positive listings was only ~$185. The top line is very concentrated. But the closer I looked, the less I trusted a clean story built from the top line alone.

The individual businesses were much more useful.

Here are the ones I keep thinking about.

1Lookup: boring API plumbing can have beautiful economics

In the frozen dataset, 1Lookup showed:

• $391,321 in 30-day revenue
• $207,064 MRR
• 655 active subscriptions
• 4,273 X followers
• Domain Rating of 11

Its current public profile is even higher: roughly $470,000 in the last 30 days, $223,000 MRR and 664 active subscriptions.

This is the first listing I would show someone who thinks distribution means building an audience.

1Lookup has 40 data products behind one API: phone validation, email verification, IP intelligence, fraud checks, company enrichment, social data, SEO data and more. Customers buy one pool of credits. Every endpoint draws from that same balance. Failed lookups cost nothing. Responses are cached for seven days.

That sounds like a billing detail. It is actually the product strategy.

A developer can enter through one narrow job, say phone validation, without adopting a giant data platform. Once the API key is already inside production, the next data need is cheaper to satisfy with another 1Lookup endpoint than with a new vendor, contract and integration. The shared credit pool removes the tax of predicting usage across 40 separate products.

The low Domain Rating and modest founder following are not weaknesses in the normal sense. The product's distribution becomes invisible after integration. It lives inside API calls, not social posts.

The gap between MRR and 30-day revenue also makes sense for a usage product. There is a recurring base, then expansion as machines make more calls. This is a much healthier reason for revenue to exceed MRR than a one-off launch spike.

If I were diligencing it, I would care less about follower count and more about:

• revenue concentration across customers
• gross margin after upstream data costs
• which endpoint gets the first integration
• how many customers add a second and third endpoint
• how difficult it is to replace once embedded

My read: start with a sharp wedge, then expand wallet share inside the same primitive. One API key and one credit balance can be more of a moat than 40 disconnected features.

Postiz: open source is the acquisition layer, convenience is the business

Postiz is easier to read because the payment shape almost clicks into place.

Its public profile showed ~$186,000 MRR from 5,490 active subscriptions on August 9. That is about $34 per subscription. Its hosted plans begin at $29 and step through $39, $49 and $99.

You can see an actual self-serve SaaS business hiding in the arithmetic.

But pricing is not the interesting part. Postiz is open source and self-hostable. A social scheduling product sounds easy until you count the integrations: 30+ social networks, OAuth flows, expiring tokens, media quirks, platform policy changes, analytics, team permissions, rate limits and failed-post recovery.

The source code earns trust and developer discovery. The hosted version sells relief from operating the mess.

Then Postiz did something I think more SaaS founders should study. It did not stop at “we use AI to write posts.” That feature is already becoming a checkbox. Instead, it exposed the publishing rail through an API, SDK, webhooks, n8n, Make, Zapier, MCP and CLI integrations for tools including Claude, OpenClaw, Hermes and Codex.

The product is moving from social media manager to system of action for both humans and agents.

That creates two distribution loops:

1. Builders discover and trust it through open source.
2. Agents and automations create more publishing volume, which makes the hosted infrastructure more valuable.

The risk is real. Social platforms can break integrations or restrict access. Supporting this much surface area is expensive. But that pain is also why customers pay. The moat is not the post composer. It is the accumulated reliability work across a chaotic set of external APIs.

What I would steal from this: if the obvious AI feature is getting commoditised, move one layer down. Own the rail the AI needs to take action.

GojiberryAI: the useful AI SDR wedge is intent, not copy

The frozen GojiberryAI listing showed $385,633 in 30-day revenue, $494,668 MRR and 4,642 active subscriptions. Its Pro plan was $99 per month.

Divide MRR by subscriptions and you get ~$107.

That is almost suspiciously neat, in a good way. It looks like thousands of customers on a standard plan plus some add-ons, not four enterprise invoices dressed up as “SaaS.”

AI SDR is one of the noisiest categories right now. Most products lead with generated outreach, which is exactly the part getting cheaper and less differentiated.

Gojiberry's stronger wedge is before the message. Give it a website and the agent learns the business, watches intent and social signals, scores the ICP, enriches prospects, then runs multichannel outreach and learns from replies. Its current $99 plan includes up to 1,800 prospects, two agents and data from 15+ enrichment providers.

The promise is not “write an email.” It is “find people who have a reason to reply now.”

That moves the product closer to the booked meeting, where budget lives. It also compresses work that used to require a lead database, enrichment tools, signal products, sequencing software and human research.

There is plenty I would challenge before calling it durable. Lead-source access changes. Email deliverability can decay. Data costs rise with usage. A customer getting results may owe half the value to the underlying channel rather than the product.

But the product instinct is right: when a horizontal AI capability becomes abundant, package it around scarce timing and a measurable outcome.

The useful move is to stop selling the generated artefact. Sell the moment it is useful.

AEO Engine: AI may create better services before it creates pure software

AEO Engine had one of the most legible business models in the dataset.

On August 9, its profile showed about $77,000 MRR from 36 active subscriptions. That works out to ~$2,142 per customer, sitting neatly between its $1,597 and $2,997 plans.

Only, the company does not pretend this is a pure SaaS product. It calls the model “Service-as-a-Software”.

Human strategists and account managers stay in the loop. More than 40 agents handle execution. The offer promises 30% organic and AI traffic growth in 90 days or the team keeps working for free. New customers are limited to five per month.

This is a much more honest use of AI than a dashboard with a chat box bolted on.

The customer does not want an AEO tool. They want to show up when buyers ask ChatGPT, Perplexity or Google an important question. They want someone accountable when that does not happen. AI can lower the labour required to research prompts, analyse visibility, produce content and track citations. It does not eliminate judgement or trust.

The capacity limit reveals the service component. It can also protect quality. The guarantee reduces perceived risk. The high average revenue per account supports real human involvement.

If the business compounds, it will not be because “40 agents” is defensible. Anyone can say that. The useful assets will be its growing body of query data, the operating playbook that determines what agents do, case studies, and the judgement about which interventions produce visibility.

What I take from it: AI does not have to remove the service. Sometimes the better business is a service with software margins improving underneath it.

Stack Influence: the software is coordinating a network and a messy operation

Stack Influence showed roughly $902,000 in 30-day revenue, but only ~$25,000 MRR from 55 active subscriptions.

If you read that like a SaaS analyst, the numbers look broken.

If you open the product, they make more sense.

Stack Influence runs micro-influencer campaigns for consumer brands. It has a network of more than 400,000 creators. Brands send products. The company manages matching, shipping, campaign briefing, creator coordination, content rights, performance tracking and repeat ambassador relationships. Its pricing includes campaign and platform components, not only monthly subscriptions.

The recurring software is a small visible piece of a much larger flow of campaign money, products and operational work.

The actual product is not “AI influencer matching.” Matching will get cheaper. The product is an organised two-sided network plus logistics plus accumulated performance history. A brand is paying to avoid finding 500 creators, checking fit, collecting addresses, tracking shipments, chasing posts, securing usage rights and figuring out which people deserve the next campaign.

That is why founder audience size tells you almost nothing here. The distribution asset is the creator network and brand relationships.

It is also why I would not value the reported 30-day number until I knew what it includes. Is creator compensation passing through? Product or fulfilment cost? What is the take rate? What is gross profit per campaign? How much labour increases when campaign volume doubles?

The mistake is judging this through a subscription-only lens. Some excellent software businesses look mediocre there because the software coordinates an economically larger offline system.

Supliful: the subscription is an entry ticket to the supply chain

The same pattern appears in a more physical form at Supliful.

Its public profile showed ~$972,000 in 30-day revenue, ~$185,000 MRR and 3,609 active subscriptions. That is ~$51 per subscription, almost exactly its $49 monthly plan.

So the recurring layer is clean. The much larger 30-day number comes from what happens after someone subscribes.

Supliful lets creators launch supplements and packaged goods without holding inventory. It coordinates product sourcing, white-labelling, manufacturing partners and US fulfilment. The company says it has powered more than 1.6 million partner orders and $64 million in brand revenue.

The subscription is permission to use the operating system. The transaction and fulfilment economics expand when a customer succeeds.

That alignment is powerful. Supliful does not have to guess whether a dashboard seat is valuable. If a merchant sells more, more physical work flows through the platform.

It also makes the business harder than the MRR suggests. There are suppliers, fulfilment times, quality problems, refunds, labelling requirements and compliance risk. Software can make all of it easier to coordinate, but the moat is partly operational.

A normal SaaS buyer might want 90% gross margin. A builder should ask a different question: does controlling this operational layer create a stronger position and a larger wallet than a thin software tool ever could?

The opportunity is in reaching further into the real workflow and participating in customer success. But you have to measure the economics of the whole machine, not celebrate processed volume as revenue.

Rezi: a strong product can still have a natural retention ceiling

Rezi is a useful antidote to the idea that every flat month means a broken product.

Its current profile showed ~$243,000 in 30-day revenue, ~$260,000 MRR and 10,745 active subscriptions. The frozen July dataset was nearly flat on 30-day revenue and down ~5% on MRR.

Rezi has been building AI resume software since well before the current hype cycle. It now serves millions of job seekers with ATS-focused resumes, keyword targeting, job-search tools, interview support and human review. Pricing is Free, $29 per month or $149 lifetime, with a separate enterprise business for organisations and universities.

The MRR per active subscription is ~$24. That is consistent with a mix of discounted plans and lower-priced subscribers. Again, the arithmetic looks like a real product.

But resume software has a strange success condition: the user gets a job and leaves.

You can improve onboarding, output quality and brand trust. You cannot turn unemployment into a daily habit without becoming weird about it. A lifetime plan may improve conversion, but it also exchanges future recurring revenue for cash today.

Rezi's broader job-search and interview features extend the workflow. Its enterprise and university products shift the buyer from an episodic individual to institutions with an ongoing need. Those are sensible expansions because they respect the underlying job instead of forcing fake retention.

The retention lesson is uncomfortable: some products are supposed to be episodic. If customer success ends the subscription, expand into the surrounding workflow or an adjacent stakeholder. Do not manufacture engagement for its own sake.

NDISCompliant: urgency can beat sophistication

NDISCompliant looked tiny beside the top listings: about $14,000 in the frozen month's revenue, zero MRR and one $297 product.

It was also growing 209%.

The offer is a pack of 94 editable compliance documents for Australian NDIS providers. A regulatory deadline arrives on October 1, 2026. The site contrasts its $297 pack with consultants charging $3,000 to $8,000. Buyers get instant access, document previews, links to official regulatory material, lifetime updates and a 48-hour fix-or-refund promise.

At $297, the frozen month's revenue suggests roughly 48 purchases. The current public number implies closer to 58. No viral loop. No model benchmark. No giant audience.

Just a narrow buyer, an expensive problem, a clock, a concrete deliverable and strong risk reversal.

This is product work too.

The obvious risk is that deadline-driven demand falls off after October. Lifetime updates also create ongoing work without ongoing revenue. The durable version could become compliance monitoring, an annual audit workspace or a subscription for document changes. But that is the next product, not evidence that the current one is weak.

Urgency is a feature. A highly specific product attached to an external deadline can outperform a technically impressive product attached to “someday.”

Reddit Agency: high collections can still be a weak asset

Then there is Reddit Agency.

Its current profile showed nearly $100,000 in 30-day revenue, but only $3,250 MRR and one active subscription. The business was listed for sale at $200,000. The listing claimed about $60,000 in monthly revenue, ~$30,000 in costs and a 50% margin.

The low asking price is more informative than the revenue chart.

This is a done-for-you service with $3,000 and $5,000 packages. The site promises to get brands recommended by ChatGPT and says it can bury bad reviews and control how a brand looks on Reddit.

That creates several kinds of fragility at once:

• lumpy project revenue rather than a large recurring base
• labour and founder execution embedded in delivery
• dependence on one external platform
• reputational risk from tactics customers may not want attached to their brand
• unclear transferability when the operator changes

None of this means the cash is fake. It means the cash flow is not automatically an asset.

A buyer does not acquire last month's collections. They acquire the probability those collections continue without the seller. One active subscription and a platform-dependent service deserve a different multiple from hundreds of embedded API customers.

This is the asset lesson: revenue proves somebody paid. It does not prove why they paid, whether they will pay again or whether the business works without you.

What I now look for before I get impressed

After going through these listings, I use a different sequence.

First: what does the number represent?

Subscription revenue, GMV, campaign spend, physical-goods volume, annual prepayment and agency collections should not live in the same mental column.

Second: divide MRR by active subscriptions.

It is a crude check, but surprisingly revealing. Postiz lands near its lower plans. Gojiberry lands near $99. AEO Engine lands between its two core packages. Supliful lands on $49. The pricing page and payment data start telling the same story.

Third: explain the gap between 30-day revenue and MRR.

At 1Lookup, usage expansion is plausible. At Stack Influence and Supliful, transactions and operations explain it. At Reddit Agency, lumpy projects do. The same gap can mean a great expansion loop or weak recurrence.

Fourth: find the non-software work.

Data suppliers, creator coordination, fulfilment, compliance research, account management and deliverability do not disappear because the homepage says AI. Often that work is the moat. Sometimes it is the margin problem.

Fifth: ask what compounds.

An API embedded in production compounds through switching cost and usage. Open-source infrastructure compounds through developer adoption and integrations. A creator network compounds through relationships and performance history. A one-time compliance pack may not compound unless it becomes an ongoing system.

Sixth: ask whether the asset survives the founder and the platform.

Audience can help, but it is not the only distribution. Low-follower founders can build enormous embedded products. High revenue can still be untransferable if every sale depends on the founder, one channel or tactics a buyer will not touch.

The part the leaderboard cannot show

The most interesting TrustMRR businesses are not the ones with the biggest number.

They are the ones where the product explains the number.

1Lookup turns one API integration into multi-product usage. Postiz converts open-source trust into hosted convenience and is becoming a publishing rail for agents. Gojiberry packages abundant generation around scarce buying intent. AEO Engine uses agents to improve service economics while keeping humans accountable. Stack Influence and Supliful use software to coordinate valuable networks and operations. Rezi respects an episodic customer job. NDISCompliant packages a deadline better than most startups package AI. Reddit Agency shows why collections and enterprise value are different things.

That is the level I want when I study a business now.

Not “what category is hot?”

What precise mechanism keeps money showing up? What becomes easier after the 100th customer? What becomes harder? Which part is software, which part is operations, and which part is a temporary channel gift?

The revenue chart is the invitation.

The product is the answer.

Research note: I used the TrustMRR Full Dataset frozen on July 30, 2026, with 8,782 listings, then rechecked the cited public profiles on August 9–10. TrustMRR values are self-reported or connected payment data. The export's revenue fields mix revenue, GMV and other payment flows, so they are not perfectly comparable accounting revenue. Per-subscription figures are simple MRR / active-subscription calculations, not audited ARPU.