A Number Worth Examining Carefully
Sparkle has 1,900 plus signups. Around 1,000 of those users are active. We spent nothing on paid acquisition to reach either figure.
That is a number worth taking seriously, and also a number worth examining carefully rather than simply celebrating.
What it tells you is that the product works well enough for people to find it, try it, and return to it without being incentivised to do any of those things. That is genuine validation. It is also a limited kind of validation. It tells you the product has value. It does not tell you whether the product is reaching the people who will eventually pay for it, or whether the way we have positioned it maps accurately to the problem it is actually best at solving.
Both of those questions are still open. This piece is an honest account of where Sparkle sits, what the organic growth revealed, and what the next phase of the product looks like.
What Sparkle Is and the Problem Behind It
Sparkle is an AI-powered design tool. The problem it addresses is one that exists at considerable scale across India and most emerging markets: a large population of small businesses, independent professionals, and individual creators who need to produce high-quality visual content regularly, without the resources to hire a designer or the time to learn professional software.
A real estate agent listing five properties a week needs visual output for each one. A fashion seller running their own catalogue needs product images at volume. A small business owner managing their own marketing needs social graphics, promotional banners, and event materials on a recurring basis. None of these people have a design team. Most of them cannot justify one.
The standard answer to this problem has been tools like Canva: simplified interfaces that lower the barrier to entry. That answer is partially right. The limitation is that simplified interfaces still require creative judgment. They require you to know what looks good, what format works for which channel, what hierarchy of information makes sense for a particular kind of content.
Generative AI removes that requirement. You should not need design intuition to produce design that looks like someone with design intuition made it.
That is the thesis behind Sparkle: AI that makes the creative judgment on your behalf, so the person producing the content can focus entirely on the information they want to communicate.
What the Organic Traction Actually Revealed
When you grow without paid acquisition, the users who find you are the users who genuinely needed you. Nobody arrived at Sparkle because an algorithm served them a targeted ad. They arrived because they had a problem, went looking for a solution, and found something that addressed it.
That makes the usage data unusually honest. The people using Sparkle are using it because it works for something specific in their lives, not because a marketing cycle persuaded them to try it.
What the usage patterns showed us, fairly clearly over time, is that the users getting the most value from Sparkle are not generalists. They are people in specific industries, producing specific types of content, at high frequency. Real estate, fashion and apparel, and small business marketing recur consistently as the contexts where Sparkle is doing its most useful work.
That pattern is a product insight of real consequence. It means Sparkle is not best understood as a general-purpose design tool. It is best understood as a high-frequency visual production tool for people in industries where visual output matters but design expertise is unavailable.
The product did not change when we understood that. The way we need to talk about it did.
The Positioning Problem
Positioning a research-built AI product as a generic design tool means competing on a shelf with tools that have been around longer, have larger teams, and have spent considerably more on brand recognition.
That is a losing position, not because Sparkle is weaker than those tools, but because the comparison is wrong. Sparkle is not trying to be a general-purpose tool. It is trying to be the right tool for a specific kind of user with a specific kind of need. Positioned correctly, those users would recognise immediately that Sparkle was built for them.
Positioned generically, they have no particular reason to prefer it over whatever they already know.
The next phase of Sparkle involves narrowing the message. Speaking directly to real estate professionals, fashion sellers, and small business operators in the language of their industry rather than the language of design software. Making the value specific and immediate rather than broad and abstract.
This is not a product problem. The product works. It is a communication problem, and communication problems are considerably more solvable than product problems.
The Market the Product Sits Inside
India has a content creation growth rate that is among the highest of any country globally, at 16.2% projected concentration in the content creation market through 2036. The shift from ad-based revenue to direct commerce through content is pushing individual creators and small businesses toward owning more of their visual production. When your content is also your storefront, the quality of every image carries direct commercial weight.
Individual creators represent over 58% of the content creation market in 2026. These are solopreneurs and small operators. They are precisely the people Sparkle was built for, and they are the fastest-growing segment in the market.
Generative AI tools are removing technical skill barriers, allowing solopreneurs to produce agency-quality output without overhead. That sentence describes exactly what Sparkle does. The task now is making sure the people who need it most know that it exists and understand that it is specifically for them.
The Commercial Gap and How We Think About It
No one has paid for Sparkle yet. That is the honest version of where the product sits commercially.
The organic traction proves the product works. The absence of paid conversion tells us the value proposition, as it currently reaches people, is not clear enough to convert. Those two things together point toward positioning as the primary problem to solve, not product development.
There is a version of this story where the response to no paid conversions is to add features, reduce friction in the signup flow, or run a promotional campaign. We do not believe any of those are the right response to the specific problem Sparkle has. Adding features to a product that people are already using but not paying for does not address why they are not paying. It just adds complexity.
The work is on making the product legible to the right audience. Making it clear, at the moment of first encounter, that this is precisely the tool they have been looking for and here is what it will do for them in the first ten minutes.
That work is in progress.
What Builds Underneath the Product
Sevorse is a research company before it is a product company. That shapes how Sparkle was built in ways that are not immediately visible but matter considerably.
The generative AI underlying Sparkle's design output was not built by connecting to an external API and adjusting the prompt until the results looked acceptable. It was built from the model layer up, with the same discipline and rigour that we bring to published research. The behaviour of the model, the quality of the output, the handling of edge cases: all of it is understood and controlled at a depth that lets us improve it with precision rather than hoping an upstream model update goes in a useful direction.
Startups with genuine research foundations are more likely to build products that perform as described. That is the honest reason research matters commercially, not as a credibility signal but as a quality guarantee.
Where Sparkle Goes from Here
The immediate priorities are two.
The first is re-engagement with the existing user base. We are in direct conversation with Sparkle users about what problems they are solving, where the product is working, and where it falls short. That conversation is the most valuable product research available to us right now, and we are treating it as such.
The second is targeted outreach to the three industries where the usage data shows Sparkle creating the most value: real estate, fashion and apparel, and small business marketing. Not broad outreach. Industry-specific communication, in the language of each industry, that makes the value of Sparkle specific and immediate rather than general and theoretical.
The freemium model continues for now. The data suggests it is an effective discovery mechanism. The focus is shifting toward what happens after discovery: how someone who finds Sparkle useful becomes someone who finds it indispensable.
A Note on Honesty in Product Development
The startup convention is to talk about traction in purely positive terms. Signups go up. Users are engaged. The product is resonating. The round is coming.
That convention serves nobody particularly well. It makes founders reluctant to discuss the actual problems they are working on, which makes those problems harder to solve, and it gives the ecosystem a distorted picture of what building a product actually looks like.
1,900 organic signups with no marketing spend is meaningful. No paid conversions is a real problem that requires real work. Both of those things are true simultaneously, and pretending only one of them exists does not help us build the right next thing.
We would rather be honest about where we are than appear further along than we are. That instinct comes from research culture, where intellectual honesty is not optional. We carry it into the product work for the same reason: because you cannot fix problems you are not willing to name clearly.