Today’s issue is a slightly different format than normal. This is an in-depth tear down of the actual mechanics of a real Tiny Empire. If this format is helpful, let me know in the comments.
The Tiny Empire we’re looking at today is Exploding Topics.
Free event for Tiny Empires subscribers
A quick note before we dig in to this week’s issue.
If you’re interested in learning how to build agents, I’ve partnered with Octonous, a new agent-building tool from the Mozilla AI team. They’re hosting a free event on Monday where you’ll learn how to build complex agents using plain language.
There are limited seats so you can register here.
Now back to this week’s issue…
Brian Dean already had an audience when he co-founded Exploding Topics with Josh Sharp. He had built Backlinko into one of the most-read SEO blogs in the world, which gave him both distribution and a particular way of thinking, everything filtered through the question of what people are actually searching for and why.
The problem he kept running into was timing. By the time a topic appeared on Google Trends, it was usually too late to build around it. The content was already saturated, the product opportunity closed or closing.
Google Trends answers: what are people searching for right now? Exploding Topics was built to answer a different question: what will people be searching for in six months?
That reframe sounds small but the economics of it are significant. Reactive data is widely available and mostly free. Predictive data, or something that credibly approximates it, is scarce, and the people who want it most are exactly the people willing to pay for it: founders evaluating product ideas, investors looking for early signals in emerging categories, ecommerce operators deciding what to stock before demand peaks, marketers trying to write about something before everyone else does.
Every one of those users can turn a good trend insight directly into money. That’s the customer base Exploding Topics built for, and it’s not an accident. When your product has a clear financial return for the buyer, pricing becomes much easier and churn becomes much lower.
At a technical level, Exploding Topics ingested data from search, social platforms, ecommerce listings, blogs, and forums, then looked for topics showing early compounding growth. The distinction they drew was between spikes, short bursts of interest that don’t sustain, and genuine curves, where interest builds steadily over months and keeps accelerating.
Detecting that pattern in raw data isn’t especially hard. The hard part, and the part that made the product worth paying for, was the curation layer sitting on top of it. Humans reviewed what the algorithm flagged. They removed noise: viral moments, memes, things that spiked because of a news event and then collapsed. They added context, grouping individual signals into broader categories they called “meta trends,” the underlying shifts that a cluster of smaller topics all pointed toward.
That editorial judgment is what separated Exploding Topics from a data feed. Anyone can license raw trend data. Several companies sell it. What Exploding Topics sold was a considered opinion about which signals were real, delivered through an interface clean enough that a non-technical founder could act on it in an afternoon.
The product came in several forms: a searchable database of trends, curated reports, email alerts, and an API for enterprise customers who wanted to pipe the data into their own tools. The database and reports served the self-serve segment. The API served larger teams and commanded much higher prices, reportedly in the range of $1,000 to $4,000 per month, without adding significant cost to serve.
The three plans are Entrepreneur at $39/month, Investor at $99/month, and Business at $249/month. On the surface that looks like standard SaaS tiering, but the logic behind it runs deeper than it first appears.
Most SaaS products tier by features: pay more, unlock more. Exploding Topics tiered by user type. The plan names tell you exactly who each one is for, and each one is priced to match what that person can justify spending without thinking twice.
A solo founder at $39/month is cheap enough to expense or absorb personally. An investor at $99/month is well within what any fund analyst would put through on an expense report for a tool they use weekly. A business team at $249/month is paying for multiple seats and shared watchlists, which a VC firm with several analysts needs.
The question of why a VC wouldn’t just sign up for the $39 plan is a good one, and the answer is that some probably did, at least to evaluate the product. But the Entrepreneur tier limits how many trends you can track simultaneously and how far back the historical data goes. A solo founder checking in weekly can work within those constraints. An investor monitoring ten portfolio companies across three sectors hits the ceiling fast and upgrades, because the marginal cost of doing so is trivial relative to the decisions they’re making with the data.
That’s the conversion mechanic working as intended. Give people enough on the cheap plan to see the value, make the constraints feel real when their use case expands, and price the next tier low enough that upgrading is easier than tolerating the friction.
This structure meant the product could serve a wide range of willingness-to-pay without building dramatically different versions of itself. The underlying data was the same. The packaging and access level varied.
The economics of the model compound in a way that most service businesses don’t. Data, once collected and curated, has near-zero marginal cost to distribute. The hundredth subscriber costs almost nothing more than the first. When you combine that with customers who have a direct financial incentive to stay subscribed, you get a business with high margins, manageable churn, and revenue that grows without proportionally increasing the work required to sustain it.
Brian Dean has spoken publicly about bringing monthly churn from around 10% down to under 3% through a combination of newsletter engagement and onboarding improvements. That reduction is not a small thing. At 10% monthly churn, you’re replacing your entire customer base roughly every ten months. At 3%, you’re replacing it every three years. The same acquisition spend produces a fundamentally different business.
The growth engine was search, and it was more deliberate than it appears.
Exploding Topics published content about trending topics, emerging products, fastest-growing startups, up-and-coming consumer categories, and ranked well for all of it. Someone searching for “trending skincare ingredients” or “fastest growing SaaS companies” would find an Exploding Topics article, see that it was produced by a tool that surfaces exactly this kind of data, and convert into a trial or subscriber.
The content was produced using the same data asset the paid product was built on. They weren’t creating a separate content operation to feed a marketing funnel. They were publishing a subset of their product for free and letting it do the acquisition work. The blog and the product were the same thing at different price points.
The newsletter operated similarly. A free weekly email kept readers engaged with the product without requiring them to log in and use it. People who’d signed up for a trial and not converted stayed on the list. People who were on the list and curious about a trend would eventually find themselves back in the product. The newsletter was retention infrastructure as much as it was an acquisition channel.
The result was a growth model that compounded on two axes simultaneously. The SEO content accumulated domain authority over time, meaning older articles continued driving traffic years after publication. The trend database grew more valuable as it deepened, making the paid product harder to replicate. Both assets appreciated without proportionally increasing the cost to maintain them.
Semrush acquired Exploding Topics in August 2024 for an undisclosed amount. At approximately $1M ARR, and with private bootstrapped SaaS multiples sitting at a median of around 3x to 5x in that period, a realistic exit range is probably somewhere between $3M and $6M, with upside possible if growth rate and churn metrics were strong enough to command a premium.
Some coverage implies a $15M exit which would require a 15x multiple. That was plausible in 2021 when SaaS multiples peaked. By mid-2024 the market had compressed significantly, and bootstrapped acquisitions were trading at more grounded numbers. The undisclosed price means nobody outside the deal knows the real figure, but the conservative estimate is the more credible one.
For a two-person business built without external funding, even the lower end of that range is a meaningful outcome. Josh and Brian didn’t build a billion-dollar company. They built something much more achievable: a product with a clear value proposition, a customer base willing to pay for it, a growth engine that didn’t require a sales team, and margins high enough that the business was profitable long before the exit.
The decision to sell rather than raise capital and scale the team is consistent with how the business was run throughout. Every decision, staying at two people, keeping the product focused, maintaining high margins, pointed toward building something valuable and sellable rather than building something large.
That’s a legitimate strategy. It’s also one that most coverage of startups ignores entirely, because a $5M exit from a two-person company doesn’t make headlines the way a $100M Series B does. But for the people building it, the outcomes are probably not that different in practice.
Not all of this is replicable as described. The data pipeline took time and technical knowledge to build. The SEO moat took years to develop. Brian Dean’s existing audience gave the business a distribution head start that most founders won’t have.
But the underlying logic is cleaner than it looks. Find a category of information people currently access too late, too expensively, or with too much noise in the signal. Build a curation layer that removes that friction, the curation can be manual at first, before anything is automated. Package it as a subscription, with a free version that does acquisition and retention work simultaneously. Keep the team small enough that margins stay healthy.
That pattern, data gap, editorial layer, subscription, free content as funnel, isn’t specific to trend intelligence. It shows up in niche research newsletters, in industry data services, in curated job boards, in any business where someone with good judgment can take information that already exists and make it more useful to a specific group of people.
Exploding Topics ran that model about as efficiently as it can be run. The two-person constraint wasn’t a limitation they worked around. It was, to a significant degree, the point.
Most readers here aren’t building a $1M ARR SaaS. The same logic runs at a much smaller scale.
Say you work in procurement and you spend your days watching supplier markets in a specific manufacturing niche. You know which new suppliers are gaining traction before the rest of the market does, because you’re in the weeds of it every week. That’s a data gap. People in adjacent roles, other procurement managers, investors looking at the sector, founders deciding whether to enter the space, would pay for a curated monthly briefing that surfaces what you’re already seeing.
You write it up. You charge $99/month. Thirty subscribers gets you $3k/month at roughly 90% margin, because the cost is your time. The free version, a twice-monthly public post about something interesting in the sector, does the acquisition work. People read it, decide you know what you’re talking about, and convert.
That’s the same model. Data gap, editorial layer, subscription, free content as funnel. No pipeline, no team, no VC. The Exploding Topics version cost more to build and produced a much larger outcome, but the underlying logic is identical and the smaller version is achievable in months rather than years.
The more useful question is: what do you already see that other people in your field don’t, and would they pay for a monthly briefing that surfaces it?
Three things to take away
Pricing by customer type rather than feature set lets one product serve a wide range of willingness-to-pay without building four versions of itself.
Free content and a paid product can be the same asset at different price points. The blog wasn’t separate from the product, it was a public version of it.
Staying small was a strategy, not a constraint. High margins, compounding assets, and a clean exit are easier to achieve at two people than at twenty.
Is there a business you’d like to see pulled apart in a future issue? Leave a comment below.
If you enjoyed this newsletter, you may be interested in joining No Code Founders, our sister community of 38k+ founders and operators building businesses without code. It’s free → Learn more
SPONSOR THIS NEWSLETTER Learn more
Want to get in front of 38k+ other founders and makers building business faster using no-code and AI tools.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.