How do I get my first 100 users?
By Bonaventure Ogeto, Founder and CEO, Mctaba Labs
Published
Get from 10 to 100 users by finding one channel that repeats, not by doing more of everything. Test referrals, WhatsApp groups and Status, partners who already serve your customers, in-person selling, content and small paid adverts, two at a time for two weeks each. Keep the channel that brings paying, returning users at the lowest cost.
Ten customers is a list you can hold in your head. You know their names, you set most of them up yourself and half of them have your personal number. Then you try to repeat the trick and find that your phone book has run out.
That is the real change between 10 and 100. The first ten came from who you know. The next ninety have to come from something that keeps working when you are not personally calling a friend of a friend. This article covers six channels that suit a young product in Kenya, how to test each one cheaply, and how to tell which is working.
The running example is a simple app that helps duka owners in Eldoret record stock and customer credit, sold at KES 300 a month. The numbers are made up and round on purpose.
What changes between customer 10 and customer 100?
Three things change.
You need strangers. People who have never heard of you must find out, trust you enough to try, and pay. That calls for proof: names of current customers, a short voice note from one of them, a shop the new buyer can visit.
You need a repeatable action. "I called my uncle's friend" cannot be done forty times. "Every Tuesday I visit fifteen shops along one road" can.
You need to count. With ten customers you can feel what is working. With several channels running, your feelings will favour whichever was most fun. A sheet will not.
Do not skip ahead to this stage. If your first ten are not using the product and paying again, more users will only fill a leaking bucket faster.
Which channels are worth testing first?
Six channels are within reach of a founder with little money. You will not use all six. The work is to test them quickly and keep one or two.
| Channel | What it looks like here | A two-week test | What it costs you |
|---|---|---|---|
| Referrals | Current customers introduce other shop owners | Ask all ten customers for two names, with a reward for each one that pays | Time, plus the reward |
| WhatsApp groups and Status | Traders' groups, estate groups, your own and your customers' Status | Three useful posts a week and a clear way to reply | Time |
| Partners | A business that already serves your customers | One partner, one simple offer to its customers | Time, maybe a commission |
| In person | Walking a market, a road or a trade meeting | Two half days, fifteen shops each | Time and fare |
| Content | Short videos or posts that answer questions your customers ask | Six posts on one platform | Time |
| Small paid adverts | Boosted posts or search adverts that open a WhatsApp chat | A fixed small budget on one message | Cash |
Test two at a time. More than that and you will not do any of them properly. Give each test two weeks, a fixed number of hours and, where money is involved, a fixed number of shillings you can afford to lose.
How do I turn happy customers into a steady flow of referrals?
Make the referral easy, specific and worth the customer's trouble.
- Ask at the right moment. Just after the product has saved them money or trouble, for example on the day the app shows a duka owner exactly who owes her and how much.
- Give them the words. A two-line message they can forward beats "tell your friends".
- Reward the customer who refers. For example, a free month for that customer, given once the new shop has paid its first month. A reward in the product costs you less than cash, and tying it to a payment means you only reward introductions that turn into customers.
- Thank them in public, with permission. A Status post saying "Welcome to Mama Zawadi's shop in Langas, introduced by Mr Kiptoo" does more than any advert.
Record every referral against the person who made it. You will often find that two or three customers send most of the introductions. Look after them.
How do I use WhatsApp groups and Status without becoming a nuisance?
Give before you ask, and never post where you were not invited.
On your own Status, post proof instead of slogans: a screenshot of a credit list with names hidden, a short clip of a shopkeeper finding a product's stock level in seconds, a customer's voice note shared with permission. Ask your ten customers to share one of these on their own Status. Their contacts are other traders, suppliers and neighbours, which is exactly the audience you need.
In groups, follow the rules of the group. In a traders' group, answer questions about stock and credit when they come up, share a useful tip once a week, and mention your product only when it answers the question asked or when the admin has agreed. An admin who introduces you is worth more than fifty posts. An admin who removes you has closed that door for good.
Never copy numbers out of a group to message people privately. It is the fastest way to be reported and blocked, and it poisons your name in a small town.
Who already has my customers, and how do I approach them?
Somebody already visits, supplies or gathers the people you want. For duka owners, that could be a wholesaler, a distributor whose sales reps walk the same roads every week, or a traders' association. For other products, think of schools, saccos, clinics, landlords, churches, chamas and employers.
A partnership works when the partner gains something plain. A wholesaler gains shops that track stock and so order more regularly. A sacco gains a benefit to offer members. If you cannot say what the partner gains in one sentence, they will be polite and then do nothing.
Start with one partner and a small test. Here is a message to open the conversation:
Good morning Mr Barasa. I run a small app that helps dukas record stock and customer credit. Ten shops in Eldoret pay for it today, and three of them buy from you. Shops that use it know what is running low before it runs out. I would like to offer it to twenty of your customers for one month and share with you what we learn about their ordering. Could I come by for 15 minutes on Thursday?
Agree on the simplest possible first step: a mention by the sales reps, a poster at the counter, or five minutes at a members' meeting. Leave commissions and contracts for after the test has shown that the partner's customers do sign up.
When is selling in person still the best channel?
When your customers stand in one place, are busy during the day and do not trust what they cannot see. That describes shopkeepers, market traders, salon owners, clinic managers and many others.
In-person selling is slow per visit, but people say yes more easily to someone standing in the shop, so it can be fast in total. Pick one road or market, go at a quiet hour, and carry a phone with the product open on a real example. Mention a neighbour who already uses it. Offer to set it up on the spot and take the first payment on M-Pesa before you leave.
You can also make this repeatable. Once you know your pitch works, a part-time agent paid per active shop can walk the next road while you walk another.
Should I spend money on adverts yet?
Only in small tests, and only once the product keeps the users it gets.
A sensible first paid test is a single boosted post or advert aimed at one town, with one message and one action: open a WhatsApp chat with you. Set a total budget you would not miss, say KES 2,000, and run it for a week. Answer every chat yourself.
Content is the slower cousin of adverts. Short videos that answer a real question, such as how to stop losing money on credit sales, cost only time and keep working after you post them. Treat content as a three-month habit. A two-week test will not show its value.
How do I tell which channel is working?
Ask every new user one question on the day they sign up: "How did you hear about us?" Write the answer in a column. Then fill in a scorecard like this at the end of each test.
| Channel | Hours spent | Cash spent (KES) | Sign-ups | Paid first month | Still active after 30 days | Cash per paying user (KES) |
|---|---|---|---|---|---|---|
| Referrals | 6 | 2,100 in free months | 8 | 7 | 6 | 300 |
| Wholesaler partner | 10 | 0 | 20 | 12 | 10 | 0 |
| In person | 16 | 1,000 in fare | 14 | 9 | 8 | About 110 |
| Boosted post | 4 | 2,000 | 25 | 3 | 1 | About 670 |
Read it from the right. The referral cost is seven free months at KES 300, one for each new shop that paid. In this made-up example the boosted post brought the most sign-ups and looks like the winner on the surface. It is the worst channel on the sheet: few paid, one stayed, and each paying user cost the most. The wholesaler brought twelve paying shops for no cash and ten hours. That is where the next month of effort should go, with referrals running alongside.
Three rules for reading your own scorecard:
- Count paying, active users. Sign-ups flatter every channel.
- Count your hours. A channel that eats forty hours a week cannot grow past you unless someone else can do it.
- Double down, then test one new thing. Put most of your time into the best channel and keep a small slot open for one new test each month.
What does this mean for the product you build?
Build the pieces that make your best channel easier. If referrals lead, add a simple referral code so rewards apply by themselves. If a partner leads, the partner may need a list of which of its customers joined. If in-person selling leads, sign-up must take two minutes on a phone with poor network.
Add one field to your sign-up from the first day: where the user came from. It costs almost nothing to build, and without it every decision in this article becomes a guess.
Key takeaways
- The first ten came from your contacts. The next ninety need a channel that works without your phone book.
- Test two channels at a time for two weeks each, with a small fixed budget of hours and shillings.
- Ask every new user how they heard about you and record the answer the same day.
- Judge a channel by paying users who are still active a month later, not by likes, views or sign-ups.
- A partner who already serves your customers is often the fastest route in Kenya.
- Keep paid adverts small until you know your cost per paying user.
Frequently asked questions
- How long should it take to reach 100 users?
- There is no standard time, because it depends on your price, how often people need the product and how easy your customers are to reach. A more useful target is weekly progress. If the number of active paying users grows most weeks and you know which channel caused it, you are on track.
- Do I need an app in the app stores to get to 100 users?
- Usually not. A web app that opens from a WhatsApp link, or a service run through WhatsApp itself, removes the download step that many people abandon when data bundles or phone storage are tight. Build the store app when users ask for it.
- Should I give the product away to grow faster?
- Free users grow the count and hide the truth. A short free trial with a clear end date is fine, because it leads to a payment decision. An open-ended free product makes it impossible to tell which channel brings people who will ever pay.
- What is the difference between a user and a customer?
- A user is anyone who uses the product. A customer pays. In some products they are different people, for example parents pay and pupils use. Count both, and when you compare channels, compare them on whichever one keeps the business alive, which is nearly always the payer.
Sources
- Traction: How Any Startup Can Achieve Explosive Customer Growth, Gabriel Weinberg and Justin Mares (book, 2015). Checked on 19 November 2026.
- The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses, Eric Ries (book, 2011). Checked on 19 November 2026.
