Finding your first users: A seed-stage guide for consumer tech founders
For consumer tech brands, go-to-market used to be the thing you figured out after the product worked. It has now become the product.
.png?w=320&h=320&auto=format)
Over the past year, the product half of a consumer tech pitch has become the least interesting part of the meeting for me. Most decks now show clean onboarding, a chat interface and a flow that two people built in a few weeks. The products work. But they also resemble each other, and I now assume any feature I see in a demo will show up in a rival app within a month.
Increasingly at seed in consumer tech, a well-built product has stopped being what separates one company from another, because software that AI tools can produce in weeks can be reproduced by a competitor just as quickly. The separation now comes from how a team reaches its first few hundred/thousand right users, how fast those users teach it something, and how quickly it turns that lesson into the next version. Go-to-market used to be the thing you figured out after the product worked. It has now become the product.
Why product stopped being the moat at seed
Consumer tech has always been easy to copy. A format or feature that one app invents has a habit of turning up in a dozen others soon after. What has changed by 2026 is how long that process takes. A feature that once took a quarter to replicate can now be rebuilt over a weekend. The clone also arrives running on the same model as the original, since nearly every team rents its intelligence from the same handful of labs.
Bryan Kim at a16z has argued that in consumer AI the advantage now goes to whoever builds first, iterates fastest and distributes best, and of those, distribution is the one a competitor cannot rebuild as quickly as the product itself.
Andrew Chen made the more structured version of this argument when he wrote that a startup needs two theories, one about its customers and one about how it will reach them, and that the best distribution is baked into the product idea rather than bolted on at the end. Dropbox is a great example. Sharing folders was part of the product from the start, so every folder a user shared put Dropbox in front of someone who was not yet using it, and in later years that feature drove a significant share of the company’s growth.
The bottleneck has moved to the audience
Most seed companies in consumer tech run the same cycle. They put the product in front of people who have the problem, watch how those people use it, change the product in response, and then go and find the next group. AI has made iterating the product close to free, which leaves finding the right people as the slow and expensive step. A team that can show each new product version to a small group of the right users every week will learn faster than a team with a stronger product shown to the wrong users, because everything the latter team hears comes from people the product was never going to serve.
Rahul Vohra’s account of how Superhuman found product-market fit is the clearest example I know of what choosing those users looks like in practice. Rahul asked users how they would feel if they could no longer use Superhuman and tracked the share who answered “very disappointed,” which in 2017 stood at 22%. Instead of trying to win over everyone who had answered, he looked at who the very disappointed users were, narrowed his focus to founders, managers, and executives, and built for them alone. Counting only that group lifted the score to 33% with minimal product work, and three quarters of building around what those users wanted took it to 58%.
The question a seed founder in this space has to answer has shifted from how good the product is to how quickly they can find the people for whom it is already good enough, and how much those people can teach them about what to build next. When building takes a matter of days, the team that is best at finding its right users is the one that improves fastest.
Five go-to-market playbooks that still set companies apart
Each of the playbooks below is a way of finding the first users a product is truly built for, and learning from them before a competitor does. These are known playbooks with the difference that they now matter right at the start.
- Know what kind of problem you are solving before you choose a channel. A user might already be searching for a fix, might have accepted the pain as permanent, or might not see the problem at all, and each calls for a different approach. Many consumer AI products in India face the second case, where the user has ajugaadthat works well enough and no awareness that anything better exists. Reaching that user takes patient explanation and a narrow wedge that proves the point.
- Let the product do the distributing. When a Meesho reseller forwarded a product catalogue to a WhatsApp group, using the product and marketing it were the same act, and every forward put Meesho’s catalogue in front of new buyers. A useful test is to follow one completed task and count how many people saw the product at work who are not themselves users; if the answer is none, every new user will have to be paid for.
- Do things that do not scale, for one specific user. That might mean spending a week in a Tier 2 kirana store, handling the first deliveries yourself or answering the early support calls in person. Since building the product is becoming the equalising step, a founder’s time is better spent being present when the first users show what they need, and the lesson from Superhuman is to decide exactly who that user is right at the start.
- Go where your users already are. Asking people to download a new app or visit a new site is expensive, while reaching them on a platform they already open every day costs far less. PhysicsWallah spent years teaching on YouTube, where students were already looking for help, before it asked them to pay for anything. Choices like these matter, because a narrow community made up of exactly the users you want will teach you more at seed than a much larger one where most people were never going to stay with you anyway.
- Make access scarce at the start. CRED initially opened only to people with a credit score above 750, which brought in exactly the users the product was built for and made an invitation something people sought out. For a company at seed, the value of scarcity is control over who gets in, so that the feedback in the early months comes from the people whose behaviour matters. Don't do this to be cool, but to ensure you don't spend valuable time and effort speaking to people who were never meant to be your consumers.
Distribution in India is harder, which makes solving it worth more
Many of India’s most successful consumer tech companies grew by inventing their own way of reaching users, because the obvious channels were either closed to them or too expensive. Those channels have narrowed further since. Apple and Google control the app stores, WhatsApp no longer allows general-purpose AI assistants on its business platform, and paid ads cost every competitor the same, so they don’t help iterate a product more meaningfully to stand apart at the start. Furthermore, asking a new user to download an app adds one more step at which they can drop off before they have seen what the product does.
But this difficulty is also an opportunity. A channel all competitors can use doesn’t separate anyone, whereas a company that finds a way of reaching households nobody else has tried, at a fraction of the cost, has built something a clone cannot simply buy. In India that kind of advantage used to matter mainly once a company was at scale, but AI has now made it critical earlier.
There are three places I would look. The first is the supply side, where a tutor, technician or seller using the product brings in the next customer because doing so earns them money. The second is the household, where a single task involves several family members, so each of them sees the product work. The third is language. A product that reaches a regional-language speaker in a format they already use, whether a voice note, a video or a forwarded message, finds an audience that many English-first products have yet to serve well.
What are the key questions to focus on then?
At seed, four things deserve more of your time than another round of product polish.
- Name your first hundred users. Describe them as people you could go and meet: where they live, what they use today in place of your product, and what would make them switch. If you can describe them only by age bracket and city tier, you haven’t found them yet.
- Test one channel with real numbers. Find where those users spend their attention today and put the product in front of them there. One channel you’ve tried, with a cost per user and a retention figure, will teach you more than five channels in a plan. It also gives investors something they can check.
- Ship to real users every week. Count how many versions reached users in the last month, and write down what you changed because of what they did. AI has made building cheap. The teams that pull ahead turn each release into a lesson about their users.
- Answer the clone question for yourself. If a competitor copied your product on Monday, what would you have left? A feature is something they can take. The users you’ve reached and what they’ve taught you take far longer to copy.
If you’ve found a way to reach your users that nobody else is using, I’d like to hear how it works, even if the product is unfinished.
Keep reading
All perspectives →
The New Physics of Influence: How consumer brands are built in 2026
Influence used to be something a brand could buy outright. Today, the physics of influence has moved.
04.11.2026 · 12 min read

Insights · Consumer Tech
The Deliberate Aspirant: How Young India Shops for the Life It Wants
22.10.2026 · 11 min

Insights · Consumer Tech
AI in Recruitment: A Market Looking For A New Winner Since 2003
05.10.2026 · 4 min

Insights · Consumer Tech
Agents in India Will Be Judged on Atoms
10.09.2026 · 8 min
.png?rect=0,182,1644,1043&w=320&h=203&fit=min&auto=format)
Insights · Healthcare
What It Actually Takes to Deploy AI in Indian Hospitals: A Conversation with MOC Cancer Care
03.07.2026 · 8 min
.png?w=128&h=128&auto=format)
.png?w=256&h=256&auto=format)

.png?w=128&h=128&auto=format)
.png?rect=0,135,1080,810&w=320&h=240&fit=min&auto=format)

