Daniel Osei
“AI gave us more first drafts. It didn't give us more good ideas.”
- Martin Knapic, Product MoatDaniel Osei is the youngest person in this series so far - six years in, all of them in growth - and one of the most skeptical about AI as a shortcut. He runs growth at Rivergate in Accra and was refreshingly clear that AI gave his team more drafts, not more good ideas.
Identity & Background
- Name
- Daniel Osei
- Current role
- Product Manager, Growth
- Current company
- Rivergate
- Location
- Accra, Ghana
- Years in product
- 6 years
Your story: how did you end up in product?
I started as a self-taught marketer running Facebook ads for a friend's small e-commerce shop, obsessing over every number. A startup noticed the numbers before they noticed my job title, and growth PM was the closest official name for what I was already doing.
Core PM Philosophy
How do you define product success?
Week-eight retention of first-time investors who made a second deposit. Downloads and first deposits are easy to inflate; someone coming back to add money after two months is a real relationship.
Product & Decision-Making
How do you prioritize?
Effort against learning. I favor cheap experiments that could disprove our assumptions quickly over large bets that only confirm them. If we can't say what result would make us stop, it doesn't make the list.
How do you do customer discovery and validation?
Phone calls in local languages, and WhatsApp voice notes. Our users don't schedule 45-minute video interviews - you meet them where they already talk to people they trust.
What's your favorite product you've shipped?
A referral flow where existing savers invite a friend and both get a small matched first deposit. It brought in a fifth of new users at almost no paid cost and, importantly, those users stayed.
AI & the Future
How is AI changing product management?
Mostly for volume at the top of the funnel - more headline variants, more onboarding copy drafts - with a human bar for what actually ships. It widens the funnel of ideas; it doesn't replace judgment about which one is honest and on-brand.
Beliefs & Opinions
What's your unpopular product opinion?
Gamified streaks bolted onto a product that isn't inherently habit-forming. They spike a vanity metric for two weeks and then everyone quietly turns them off in settings.
Which author, thinker, or influence shapes you?
I don't try to track every tool. I pick one Friday a month to actually build something small with whatever's new, hands-on, and ignore the rest of the noise until it proves itself in that hour.
Company & Product
What does your product solve?
We're a savings and micro-investment app for first-time investors across West Africa - the product that gets someone from 'I have some money and no idea what to do with it' to their first small, safe investment.
What's the market opportunity?
The big players are building for volume across the continent; we're building trust in specific communities first - in-person partnerships and local language - then letting the product scale after the trust exists. Slower, but it compounds instead of churning.
More from the conversation
What do you do outside work?
I coach a youth football team on weekends, which has taught me more about incentive design than any growth textbook - twelve-year-olds will find the loophole in any rule within one practice, exactly like users find the loophole in any growth mechanic.
What almost derailed your path into this career?
A year at a big agency doing growth work for clients I had no real stake in, purely by the numbers, no product context. I left for an in-house role at half the agency's prestige, and it was the best trade I've made - context beats prestige.
How big is your team, and who's the ICP?
Growth is four of us inside a product org of about 20, at a company of 60. ICP is someone in their twenties, employed, banked but not previously invested - the gap between having a bank account and actually building any wealth.
What does 'growth' mean on your team today?
Fewer paid channels, more compounding loops - referrals, content that ranks, integrations that pull people in sideways. Paid got more expensive and less trustworthy at the same time, so we had to get more patient.
How has AI changed your team's processes, not just the output?
Review cycles got shorter but stricter - we ship more variants because drafting is cheap now, but every one still needs a human sign-off against brand and honesty before it goes live, which is actually slower per-item than before, just wider in parallel.
What's underrated in growth right now?
Retention debugging. Everyone wants to talk acquisition, but going deep on why week-two users vanish is where the biggest, cheapest wins still hide.
What's a topic you think matters more than the industry currently treats it?
Honesty in AI-assisted marketing copy, specifically in financial products. It's now trivially cheap to generate a hundred persuasive variants, and the industry hasn't caught up on what 'persuasive but not misleading' means at that scale. I think regulation gets here before self-discipline does.
Enjoyed this conversation?
A new interview with a product person lands every week. Get it in your inbox.