Nine posts of architecture, infrastructure, failures and design. This is the tenth, and it is the one the series has been building toward: how a product like this actually finds the people it is for.
I want to be clear about what this post is not. It is not a revenue report, because there is no revenue. It is not a growth case study, because there is no growth to study yet. Kidslen is in early access. What follows is the commercial thinking behind it — the reasoning I would want to read from someone in my position, written before the outcome is known rather than reverse-engineered after it.
Who actually pays, and what they are buying
The buyer is a parent of a child roughly between nine and fifteen. Younger than that and the answer is usually a device-level restriction and a shorter leash. Older and the parent has mostly lost the ability to install anything on the phone without a conversation they will have to win on the merits.
But “a parent” is not a useful segment, because parents arrive in this category from very different places, and they want different things:
| Who | What triggered the search | What they actually want |
|---|---|---|
| The worried parent | A news story, a school email, a change in their child’s behaviour | To stop feeling out of control |
| The negotiating parent | A fight about screen time that keeps recurring | A shared set of facts so the fight is about rules, not about what happened |
| The technical parent | Curiosity; they have already read a privacy policy this week | A tool that does not embarrass them, with data practices they can inspect |
| The co-parenting household | Two homes, inconsistent rules | One agreed picture both adults can see |
What almost none of them are buying, when you listen carefully, is surveillance. They say things like “I just want to know” and “I don’t want to be the bad guy” and “I want to be able to talk to him about it without it turning into a fight”.
So the honest framing of the product’s value is: peace of mind, and a conversation. Not evidence. Not catching anyone. A parent who buys monitoring software to catch their child has a problem the software will make worse, and — commercially — is also the customer most likely to churn, because the day they find something is the day the product’s job ends badly.
The parent who buys a shared, visible record of how time gets spent has a tool they will still be using next year, and a child who is not planning how to defeat it. Those are the families the product is designed for and, not coincidentally, the ones with the longer lifetime.
This is why the messaging is what it is. “Monitoring your child sees too” is not a softened version of the pitch. It is the actual pitch, aimed at the actual buyer.
Positioning when you cannot outspend anyone
The incumbents in parental monitoring are large, well funded, and have been buying search ads in this category for years. A solo builder does not win an auction against them. Anything that requires paid acquisition at scale is closed.
What is open is being a different thing, clearly, so that the comparison is not on their terms.
The incumbents’ implicit promise is coverage: everything your child does, on every app, wherever they are. More data, more features, more control. It is a compelling promise and it has a structural weakness — it produces a product the child will resent and try to remove, a data set the vendor should be terrified of holding, and an app-store relationship that is permanently one policy update from trouble.
Kidslen’s promise is the inverse: less data, visibly collected, with the child’s knowledge. Titles, platforms, durations. Six platforms, done properly. A permanent banner. Deletion that works.
Stated as a feature list, that sounds like a weaker product, and against the coverage promise it genuinely is. Stated as a position, it is a different product for a buyer the incumbents systematically under-serve: the parent who wants some visibility and is uncomfortable with everything they have been offered so far. That parent currently resolves their discomfort by buying nothing. They are not a slice of the incumbents’ market; they are the people who bounced off it.
The three things I can say that the incumbents structurally cannot:
- Your child can see everything you see. They cannot say this, because for many of their customers the hidden mode is the reason they bought.
- Here is the complete list of what we collect, and it is short. They cannot say this, because their list is long by design.
- Here is how it is built. They cannot say this, because nobody publishes an architecture write-up including their own bugs. This entire series is a positioning asset — it is not a coincidence that part seven is about the week everything broke.
Being small is the enabling condition for all three, which is the only sense in which being small is an advantage.
Pricing a family product
Kidslen is free during early access. That is not a pricing strategy, it is an admission: I need families using it, reporting what is confusing, and telling me what breaks, and charging for that is charging people to do QA.
The pricing thinking for when that ends:
A family plan, not per-child pricing. Per-child pricing is the category default and I think it is a mistake in this product. It charges a family with three children three times as much for a service whose marginal cost to me is close to zero, and it makes the household with more children — often the household with less money and more need — pay the most. It also creates a genuinely perverse moment at the point of sale, where a parent decides which of their children to monitor based on price. I do not want to build the screen where that choice happens.
So: one price per family, a reasonable number of children and devices included, and if someone has six children they are an interesting family, not a revenue opportunity.
Monthly, with a meaningful annual discount. Monthly because trust in this category is earned in the first weeks and a parent should be able to leave cheaply if it is not for them. Annual because the families who stay are the whole business and the discount should be real, not cosmetic.
What a paid plan would have to be worth. The test I want to pass before charging is simple: a parent, three months in, should be able to name something concrete that the product changed. A rule they now both agree on. An argument that stopped happening. A week where they saw something early enough to ask about it gently. If a parent cannot say that, the subscription is a guilt tax and it will churn — and it should churn.
What I will not do. No free tier crippled to be useless — the free version is the whole thing during early access. No annual lock-in with a hostile cancellation path. No price increase applied silently to existing families. No selling, sharing or “anonymised aggregate insights” from children’s data, at any price, ever. That last one is worth stating in public precisely because it removes a future revenue line, which is what makes it credible.
Distribution that fits one person
Paid acquisition is closed. Here is what is actually available, roughly in the order I trust it:
Content and search. This series is the first instrument. Not keyword-stuffed listicles — engineering write-ups and honest documentation. It works on a long timescale and it compounds, and it reaches two audiences at once: engineers who respect the build, and parents who search for exactly the question a post answers. A parent searching “does parental monitoring software show my child what it collects” is closer to buying than anyone who clicks an ad, because they have already decided what kind of product they want.
Parent communities. School groups, local parenting forums, the group chats where one parent asks what everyone else uses. This is where recommendations in this category actually travel. It is also a place where marketing is instantly rejected, so the only viable approach is to be present as a person who built something and will answer questions, including the sceptical ones. Slow, unscalable, and the highest-quality channel there is.
Schools. Interesting and difficult. Schools are a natural point of trust and they talk to every parent at once. They are also risk-averse, have long procurement cycles, and are right to be cautious about endorsing anything that monitors students. The realistic version is not selling to schools; it is being the product a school’s digital-citizenship material can mention without embarrassment, because its data practices survive being read aloud at a parents’ evening.
App-store presence. Both stores are search engines that people already trust. Listing quality — screenshots that show the child’s view, a description that states plainly what is and is not collected, honest privacy labels — is distribution work disguised as compliance work.
Word of mouth between parents. The end state, and the only channel that scales without money. A parent recommends this kind of product only when they are not embarrassed to admit they use it. That is the single sharpest argument for the transparency design: a covert product is unrecommendable in public, because recommending it means admitting you spy on your child. A visible one can be mentioned at a dinner table.
Trust is the entire funnel
In most software categories, trust is a conversion factor — it makes a good product convert somewhat better. In this one, it is the funnel itself.
A parent evaluating monitoring software is running exactly one question: what does this company do with what it learns about my child? They will not ask it out loud. They will resolve it from signals — how the site looks, what the privacy policy says, whether the pricing is honest, whether the app-store reviews mention creepiness, whether the founder is a real identifiable person.
Which means the marketing asset is not a campaign. It is verifiable transparency, and the way to get it is to publish things that would be uncomfortable to publish if they were not true:
- The complete list of collected fields, and the longer list of what is not collected
- What the child sees, shown as the actual screen and not a mock-up
- The retention window, and what deletion actually does
- How the security works — rotating tokens, deny-by-default, audit logs
- The bugs. Part seven of this series describes a concurrency bug that lost events and a date-format mismatch that broke a screen, in my own product, in public.
That last one is the one people find strange, and it is the most effective. Anyone can claim to be careful. Publishing a specific failure and its fix demonstrates a way of working, and it is very hard to fake, because a competitor cannot copy it without also having to be honest about their own week.
The other half is the person. A product that handles children’s data from an anonymous brand is a harder sell than one from a named engineer with a public trail of how it was built. Being small and identifiable is, here, a feature.
What I would measure first
Vanity metrics are especially useless in this category, because signups mean nothing if families quietly stop using the thing. Here is what I would actually watch, roughly in priority order:
- Enrollment completion. Of parents who create an account, how many get a child’s device enrolled and reporting? This is the real funnel. Everything before it is intent; everything after it depends on it.
- Week-four retention of active devices. Not “account still exists”. Is a device still sending heartbeats a month later? A child who removed the app is the clearest signal a family exists that I failed.
- Whether parents open the dashboard after week one. Monitoring products have a known decay curve: install, check obsessively for a week, never open it again. If that happens, the product is a reassurance purchase, not a tool, and it will churn at renewal.
- Policy usage. Are families setting limits, bedtimes, alerts — the features that require a conversation with the child — or only looking at the feed? Policy usage is my proxy for “this became a shared arrangement” rather than “a parent is watching”.
- Deletion and export requests. A low number is not a good sign; it may just mean people cannot find the buttons. I want these to work and to be used, and I would rather see a family export and leave than quietly drift.
- Support questions, read as a list. Which ones are confusion, which are missing features, and which are someone asking for a hidden mode — that last count tells me how well my own positioning is filtering the market before the sale.
What I would not measure first: revenue, signups, or anything with a growth curve attached. Those are downstream of whether the product works and whether families keep it, and optimising them early in a category built on trust is how you end up with a business that has to be marketed with fear.
That is the series. Ten posts: the architecture, the capture mechanism, the pipeline, the apps, the infrastructure, the failures, the privacy model, the brand, and now the commercial thinking. Written before it worked, rather than after, which means some of it will turn out to be wrong — and when it does, that will be a post too.
If you are a parent who has read this far, that is the pitch. There is no fear paragraph and no invented statistic. The product is at kidslen.app, and everything it does to your family’s data is written down before you sign up.