I got my first 100 paying customers for Relatable because of WhatsApp. Not because I ran some clever WhatsApp marketing campaign. Because I made it stupidly easy for people to talk to me, and then I actually talked to them.
That sounds simple. It is simple. But the way most startups use WhatsApp in Africa is either too aggressive (spamming broadcast lists) or too passive (a number on the contact page that nobody checks). The version that works is more deliberate than both. It changes at every stage of the business, and it serves a different purpose each time.
Here's how I've used WhatsApp from day one through to now, and what I've learned about where it works and where it doesn't.
Product validation: WhatsApp as your first sales channel
Before I had a polished product, before the onboarding flow worked properly, before pricing was finalized, I put a WhatsApp CTA on the landing page. Not buried in the footer. Front and center. "Have questions? Chat with us on WhatsApp."
The goal wasn't customer support. It was sales and learning. When someone clicked that button and sent a message, I got something no analytics dashboard can give you: a real conversation with a real potential customer. I could hear how they described their problem in their own words. I could ask what they'd tried before. I could understand what they were actually looking for, which was often different from what I assumed.
And then I could sell to them manually. Walk them through the product, answer their objections in real time, help them through the payment flow if they got stuck. Those first sales were not scalable. They were invaluable. Every conversation taught me something about the product, the positioning, or the audience that I would have taken weeks to learn from analytics alone.
If you're pre-product-market fit and you don't have a WhatsApp link on your landing page, you're leaving your best learning channel on the table.
Early traction: WhatsApp as a safety net
Once the product started working and users could sign up and pay without my help, WhatsApp shifted roles. I kept a support link on every page of the app and the website, but the purpose changed. It became the safety net for when things didn't work.
This matters more in Africa than in most markets. When a user in Nairobi hits a confusing screen, encounters a bug, or doesn't understand what they're paying for, they're less likely to submit a support ticket or search a help center. They're much more likely to either message you on WhatsApp or leave. By making WhatsApp available everywhere, I caught the users who would have otherwise just disappeared.
The conversations at this stage were different from the validation phase. Users would send screenshots of error messages. Voice notes explaining what went wrong. Photos of what they expected versus what they got. This raw, unfiltered feedback was more useful than any structured feedback survey I could have designed. It surfaced bugs faster, revealed UX problems I hadn't noticed, and often saved customers who were about to churn.
Post product-market fit: WhatsApp moves down the funnel
This is the transition most startups don't think about. When your product is working and your funnel converts reliably, WhatsApp shouldn't be the first thing users interact with anymore. You want them to sign up for the app first. You want the product to answer their questions, not you.
I moved WhatsApp further down the funnel. It's still there, still available, but I want users to experience the product first. If the app answers their questions and delivers value, they never need to message me. If it doesn't, WhatsApp is there as the fallback. But it's no longer the front door.
This shift is important because WhatsApp conversations don't scale linearly. At 50 customers, you can reply to everyone personally. At 500, you can't. The product needs to do the work that WhatsApp was doing in the early days. WhatsApp becomes the exception handler, not the default experience.
Churn prevention: WhatsApp as a relationship layer
This is where WhatsApp is most underrated. When a customer hits a problem and is about to give up on your product, the difference between them churning and them staying is often just the friction of complaining. If complaining is hard (fill out a form, write an email, wait 48 hours for a response), they'll just delete the app. If complaining is easy (send a WhatsApp message, get a reply in minutes), you have a chance to save them.
I've kept customers who were frustrated, confused, or ready to cancel simply because they could send me a voice note about their problem and I responded quickly with a fix. Those customers don't just stay. They become loyal. They've seen that there's a real person behind the product who cares enough to solve their problem in real time.
For a startup in Africa, this kind of relationship is disproportionately valuable. Word of mouth is still the strongest distribution channel on the continent. A customer who had a problem and got it solved through WhatsApp will tell their friends. A customer who had a problem and got a "we'll get back to you" email will tell their friends something very different.
Where I'm taking it next
For Relatable, I want the line between the app and WhatsApp to blur. When a customer's AI-generated images are ready, I push them directly to their WhatsApp. The notification, the delivery, and the conversation all happen in the same place the customer already spends their time. Retention, support, and relationship-building collapse into a single channel.
This is what lifecycle management looks like when you build it for how African users actually communicate. Not email drip campaigns that go unread. Not push notifications that get ignored. A WhatsApp message that feels like it came from someone they know, delivering something they're waiting for.
The one thing I won't do
I won't force users to generate images through a WhatsApp bot. I've seen this pattern in several African AI products: the entire product experience lives inside a WhatsApp conversation. You send a message, the bot responds, you send another message, the bot processes it.
It's annoying. The UX constraints of a chat interface are severe. You can't browse, you can't compare, you can't go back easily, you can't see your history at a glance. For anything more complex than a single question and answer, a proper app or web interface will always be better.
WhatsApp is extraordinary as a communication layer, a delivery mechanism, and a support channel. It's terrible as a product interface. The companies that understand this distinction will use WhatsApp to make their product feel more personal and more accessible. The ones that try to replace their product with WhatsApp will end up with something that's neither a good app nor a good conversation.
Use WhatsApp to talk to your customers. Don't use it to replace your product.