When your data quietly becomes an asset.

The most valuable asset in a mature SaaS company is usually the one nobody’s turned into a product.

It’s not the feature set. It’s not the brand. It’s the data the company has been quietly piling up for years just by being the system of record its customers rely on. Every transaction, every workflow, every edge case its users run through the product leaves a trace. Individually those traces are noise. In aggregate, they’re something almost no competitor can replicate.

And most of the time, the company sitting on top of it can’t see it.

This isn’t a failure of intelligence. It’s a phase, and it’s worth understanding why it happens, because the window where the data becomes valuable is the same window where you’re least equipped to notice.

The data is worthless right up until it isn’t.

Every data business goes through the same arc, and the first stage fools people into ignoring it forever.

Early on, you don’t have enough users to have anything worth looking at. Ten customers doing ten things a day is not a dataset, it’s an anecdote. Founders who obsess over “our data moat” at this stage are usually kidding themselves. There’s nothing there yet, and building for a moat you don’t have is a good way to neglect the product that would actually get you users.

So you do the right thing. You put your head down and build for the audience in front of you. You solve their problem better than anyone else. You earn retention. You become the tool they open first thing in the morning.

Then one day, without any single moment marking it, the data crosses a line. You have enough users, doing enough things, for long enough, that the aggregate starts to describe the whole industry. What was exhaust becomes inventory.

Stage	What the data is	Why you can’t see it
Too early	Thin, anecdotal	There’s genuinely nothing there yet
Maturity	Rich, representative	You’ve trained yourself to serve the existing roadmap
The unlock	A product in its own right	The moment it flipped, nobody was watching for it

The cruel part is the timing. The data becomes valuable at exactly the point where you’re least likely to notice, because by then you’ve spent years getting very good at one thing: serving your existing audience.

The focus that got you here is the reason you’re blind.

I want to be fair to the teams that miss this, because I’ve watched good ones do it.

The reason a mature product company overlooks its own data isn’t laziness. It’s the flip side of the thing that made them successful. To win a market, you have to develop an almost obsessive focus on your core customer and their problem. You build the muscle of asking “what does this user need next” and you fire it a thousand times until it’s reflex.

That muscle is what beat the competition. It’s also what keeps you staring at the roadmap when the more valuable opportunity is sitting in your database, describing an entirely different business.

Nobody on the team is assigned to notice the moment the data flipped from noise to asset. The product people are shipping. The engineers are scaling. Leadership is protecting the core. The dataset just quietly compounds in the corner while everyone does the job that made the company great.

Reporting is not a product. Prediction is.

Here’s how I’d tell the difference between data that’s just a dashboard and data that’s a business.

Most companies already surface their data back to customers as reporting. Charts, exports, “here’s your activity this month.” That’s table stakes, and it’s not what I’m talking about. Reporting tells a customer what they already did. A data product tells them something they couldn’t have known on their own, usually by comparing them to everyone else, or by predicting what happens next.

Stripe is the cleanest example. Every business processing payments through Stripe generates fraud signals. On its own, one merchant’s fraud data is thin. Across the whole network, it’s a model no single company could ever build alone. Stripe turned that aggregate into Radar and sells fraud prevention as a product. The data was a byproduct of the core business. The product was hiding inside it.

Zillow did something similar with the Zestimate. The value of a home is exactly the kind of thing that’s hard to know and valuable to estimate, and Zillow had the listings, the histories, and the transactions to model it. Bloomberg built an empire on the idea that financial data, aggregated and made instantly accessible, is worth more than any individual piece of it. Plenty of developer and marketing tools now publish industry benchmarks drawn from their own users, and those benchmarks quietly become one of the stickiest reasons to stay.

The pattern is the same every time. The company was collecting the data anyway. Someone recognized it had crossed from exhaust into inventory, and treated it like the asset it had become.

AI closed the insight gap, not the monetization gap.

This is where I think the industry has genuinely gotten better, and where it’s still leaving most of the value on the table.

AI has opened a lot of doors for internal insight. Companies are much better than they were a few years ago at leveraging their own data to run smarter: catching churn earlier, spotting patterns in support tickets, surfacing what’s working. The tooling has made internal intelligence almost a default expectation now.

But being good at using your data internally is not the same as being good at selling it. Those are two different muscles, and the industry has built one and barely touched the other. Most companies have quietly become sharp at internal insight while never once asking whether the thing they’re mining for their own decisions could be a product someone would pay for.

That gap is the opportunity. The internal-insight problem is largely solved. The monetization problem is wide open, and it’s where the actual new revenue lives.

Try your existing audience first.

Say you do notice. You recognize there’s a product hiding in the data. The first question is who you sell it to, and the instinct usually points the wrong way.

The instinct is to sell the new data product to someone new. A different buyer. A new market that would find your aggregated insights valuable. It feels ambitious, and it’s the harder road. A new buyer means a new funnel, a new sales motion, new positioning, new everything. You’d be starting a second company from scratch and stapling it to your first.

So start closer to home. See if you can monetize the data with the audience you already have. Same customers, same funnel, same trust you’ve spent years earning. You already have distribution to the exact people most likely to find the insight valuable, because it’s built from their own world. Turning your existing audience into the buyer for the new product is a dramatically shorter path than going and finding strangers.

Sometimes that isn’t an option. The insight is genuinely more valuable to someone outside your customer base, and you don’t get to choose. That’s fine, it just means you’ve signed up for the harder version: a new market, a new motion, and a longer road to revenue. Worth doing when the prize is big enough, but go in knowing what it costs. When you can sell it through the channel you already own, that’s usually the version that works with the least friction. Not a new business bolted onto the old one, but a new layer of value sold to people who already trust you.

Don’t torch the core on your way to the new thing.

When you can sell to your existing audience, it’s the shorter path, and it’s also where you can do real damage if you’re careless. There are a few ways to get it wrong, and I’ve seen each of them break trust that took years to build.

The first is making it feel like price-gouging. If customers who already pay you suddenly feel like they’re being charged a premium for something they assumed was included, you haven’t launched a product, you’ve launched a resentment.

The second is subtler and worse: making it feel like you’re selling their own data back to them. When you’re the system of record, your customers gave you that data to run their business, not to have it repackaged and sold back at a markup. Cross that line and it reads as a betrayal, even when the product is genuinely useful. How you frame it, and how you handle the aggregation and privacy, matters as much as the product itself.

The third is the quiet killer: complexity. People pay for predictable pricing. They’ll often choose the simpler plan over the cheaper one just to know what the bill will be. A clever new data tier that fragments your pricing into a maze can cost you more in confusion and churn than it ever earns in new revenue. Simplicity is a feature, and every new SKU spends some of it.

The best data products I’ve seen respect all three. They feel like a natural extension of the relationship, not an extraction from it.

Not every dataset is a business.

I’d be doing the same thing I’m criticizing if I told you every company is sitting on a goldmine. Plenty aren’t.

Some datasets are rich internally but useless to anyone else. Some are too small, too narrow, or too tied to one customer’s context to generalize. Some would be genuinely valuable but can’t be sold without stepping on privacy commitments you should never break. Recognizing which datasets are actually a business, and having the discipline to say the rest aren’t, is half the job.

Which is why this deserves the same rigor as any new product. Talk to customers before you build. Find out whether they’d actually pay, and how much, and whether the insight changes a decision they care about. Size it honestly. The fact that the data is proprietary makes it a candidate, not a guarantee. This connects to a broader point I’ve written about, that proprietary data is one of the few real moats left in a world where anyone can build software in a weekend, and to the reason founders struggle to step back from the core in the first place.

The business case is already written. Someone has to read it.

The strange thing about all of this is that the hard part is already done. You’ve spent years building the product, earning the users, and accumulating the data. The asset exists. The distribution exists. The trust exists.

What’s missing is the person willing to step back from the roadmap long enough to notice that the byproduct became the prize. It takes the same high standards you’d hold on the core product and points them at a question the team has been too busy to ask.

Your richest dataset is a business case waiting to be written. The company that reads it first wins a market its competitors don’t even know is there.

If you’re staring at a mature product and wondering where the next line of growth comes from, that’s the kind of problem I like to sit down and work through. It’s usually closer than it looks.