I spent the last month working on a research paper for my MBA that tried to answer one specific question: what does the future of e-commerce look like in Kenya, given everything Industry 4.0 has promised us? I went into it expecting to find the familiar story — physical retail on the way out, e-commerce on the way up, everyone online by 2030. The data told a different story. A much more interesting one. And the more I looked at what is actually happening in Kenyan retail in 2026 — not what is being predicted, but what is being lived — the more I became convinced that the next decade here will belong to a specific kind of business. Not the pure online player. Not the stubbornly offline shop. The blended operator.
I want to share what I found, and more importantly what I think it means for anyone running a retail or service business in Kenya today. There is no magic formula. There is a set of patterns that keep showing up in the businesses that are winning, and a set of mistakes that keep showing up in the businesses that are struggling. The sooner Kenyan operators recognise both, the better.
What the data actually shows
Between 2019 and 2024, the companies we think of as the future of commerce did not behave like the future. Jumia's 2024 revenue was below its 2019 peak. Amazon's retail business grew at 7.4 per cent annually — essentially the same rate as Walmart, which is a physical retailer. eBay's revenue in 2024 was lower than it had been five years earlier. Alibaba's GMV went sideways. Meanwhile Walmart, Costco, Carrefour, Naivas, Quickmart and the rest of the physical retail world kept growing profitably. In Kenya specifically, Carrefour doubled from 14 outlets to 28. Naivas went from 64 outlets to over 100. The supposedly dying model kept expanding. The supposedly ascendant model kept stalling.
The businesses that actually grew fastest were neither pure online nor pure physical. Shopify, which powers millions of small brand-owned stores, grew at 24 per cent compound annually. TikTok Shop went from zero to 33 billion dollars in global sales. MercadoLibre in Latin America grew at over 40 per cent. What these winners have in common is that they are not selling commerce as a destination — they are selling commerce as a tool. You open your own shop. You meet customers where they already are. You integrate payment, delivery and communication into something that feels personal rather than transactional.
And then there is the Kenyan story underneath all of this. Masoko failed. Copia collapsed. Kune Foods closed. Twiga is in what its former CEO described diplomatically as a transition, and what whistleblowers described as a soft liquidation after burning more than 160 million dollars of investor capital. Every one of these was a well-funded, well-publicised, venture-backed attempt to build a Kenyan online retail giant. Every single one failed at the same thing: the economics of acquiring and retaining customers online, at scale, without a physical anchor, simply do not work in this market. The money ran out before the model proved itself. It will keep running out for anyone who tries the same thing the same way.
What does work, quietly and without fanfare, is the boutique in Karen with an Instagram account and a WhatsApp catalogue. The hardware dealer in Industrial Area whose regulars call before they walk over. The salon with a booking app and a loyal clientele. The small food brand that started on TikTok and now sells out of three supermarkets. None of these are trying to replace physical commerce. They are layering digital tools on top of relationships they have built face to face. This is the pattern. This is what the paper ultimately argued. The centralised marketplace as a dominant model is fading. What is rising is something quieter and more distributed — a mesh of small, physically-grounded, digitally-competent businesses.
Before the tactics, have a strategy
Before any of what follows is useful, one thing needs saying. You cannot fly blind. Too many Kenyan businesses are run on instinct and urgency — react to what is in front of you, chase what seems to be working, drop what feels slow. This is not strategy. It is drift. A business without a strategy is a business that cannot tell the difference between a good decision and a lucky one, and cannot learn from either. When the lucky streak ends, there is nothing to fall back on.
A strategy does not need to be a fifty-page document. It needs to answer a handful of uncomfortable questions honestly. Who is your customer specifically, and who is not? What are you the best choice for, and what are you not? Where will growth come from over the next three years, and where will it not? What will you stop doing to make space for what matters? These questions sound obvious, but most Kenyan businesses cannot answer them cleanly. They know what they sell. They do not know who they are for.
I will go deep on strategy in a future post, because it deserves more space than this piece can give it. For now, the single point to carry forward is this: none of the tactics below will save a business that has no strategic frame. The tactics are tools. The strategy is the reason you pick up one tool and not another. Without it, you will spend money, time, and attention on things that feel productive but do not add up.
Pick the right channel for your actual business
The first mistake I see Kenyan businesses making in 2026 is channel drift. The owner reads an article about TikTok Shop, signs up for Jumia Seller Center, opens a Facebook page, posts occasionally on Instagram, puts their business on Jiji, and ends up doing six things badly instead of one thing well. Every channel has its own rhythm, its own audience, and its own content demands. You cannot run six of them in parallel as a side effort.
The question is not "which digital channel should I be on?" It is "where are my actual customers already spending their attention, and how do I show up there properly?" A hardware store's customers are searching on Google. A fashion boutique's customers are scrolling Instagram and TikTok. An electronics dealer's customers are asking questions in WhatsApp groups. A beauty brand's customers are watching creator reviews. A furniture maker's customers are looking at Pinterest and Google Maps. These are different places, and each one requires a different presence.
Pick one channel. Go deep on it for twelve months. Understand its algorithm, its content style, its measurement tools. Only once you are running that channel competently do you add a second. Businesses that try to be everywhere at once are usually present nowhere, because none of the channels get the investment needed to actually work.
Your customers are your brand ambassadors — and most Kenyan businesses do not notice
Within every Kenyan customer base there are a handful of people who already do the work of marketing for you without being paid. They use your products. They believe in what you do. They talk about you naturally in conversations with friends. They post you on their stories without being asked. They have the platforms and the credibility that paid influencers struggle to manufacture. Most Kenyan businesses have no idea who these people are.
The research on this is clear: 56 per cent of Gen Z and Millennial consumers globally buy based on a content creator's recommendation, but the distinction that matters in Kenya is between paid influencers and natural ambassadors. A paid influencer is a cost centre with questionable conversion. A natural ambassador is a customer who already talks about you, and whose friends already believe them. The first is expensive and often ineffective. The second is nearly free and genuinely effective — but only if you identify them and treat them well.
The practical move is this: observe your customer base carefully. Who posts your products without being asked? Who tags you? Who leaves detailed reviews? Who walks their friends into your store? These are your ambassadors. Give them early product trials for free, before anyone else gets to see a new launch. Invite them to closed events. Send them the occasional gift with a handwritten note. Book this spend as a marketing line item with expected return, not as charity. A brand ambassador programme of thirty carefully chosen real customers will outperform a budget of KSh 500,000 on paid influencers, and will do so quarter after quarter at a fraction of the cost.
Every shilling of marketing must have a question attached to it
The second mistake I see is marketing by whim. An owner spends KSh 50,000 on boosted Facebook posts one month, KSh 80,000 on a TikTok campaign the next, KSh 100,000 on new display equipment the month after, and a discount promotion somewhere in between. Nobody asks what each of those was supposed to achieve. Nobody measures whether it did. The money is gone and the lesson is not learnt.
If you are serious about marketing in 2026, set a budget and set expectations against every line in it. When you spend money on TikTok content, what specific outcome should you see in six weeks — follower growth, store visits, website traffic, direct sales? Write the number down before you spend. Then check it. When you run a discount promotion, what margin are you sacrificing and what volume should it produce to justify the sacrifice? Calculate this in advance. When you buy new equipment, what customer experience or throughput improvement does it enable, and how will you know it worked?
The businesses that get this right are not the ones with the biggest marketing budgets. They are the ones with the most disciplined measurement. A KSh 50,000 monthly budget, tracked carefully and adjusted monthly based on what actually produced results, will outperform a KSh 500,000 monthly budget spent on whatever feels urgent that week. Hope is not a strategy. Review is.
Stay in the game long enough to win it
Closely related is the mistake of treating digital advertising like a billboard. A Kenyan business owner puts a Facebook ad live, runs it for two weeks, sees no sales, takes it down, and concludes that Facebook does not work for their category. What actually happened is that they quit before the system could deliver value. Digital ad platforms are not billboards — they are learning systems. In the first two weeks of a Facebook or TikTok campaign, the algorithm is still figuring out who your customers are. It is showing your ad to a broad audience, watching who clicks, watching who converts, and gradually narrowing toward the people most likely to buy. Month one is tuition. Month three is when the system starts paying you back. Month six is when your cost per acquisition becomes genuinely efficient.
Quitting at week two is like paying a university first-term fee, walking out after the first two lectures, and then complaining that education is a scam. The money went in. The value had not yet come out. It never would, because you left before the learning could compound. There is a similar problem with brand recall. Kenyan consumers typically need to see a brand seven to twelve times before they take action. A two-week ad flight might deliver three impressions per person. You stopped exactly when the customer was starting to notice you. The customer who would have bought in month four never got to month four because you cancelled in month one. And the algorithms themselves penalise inconsistency — when you pause, platforms quietly deprioritise your content; when you restart, you are starting from zero again, not resuming where you left off. The intermittent marketer pays the learning cost every time.
The implication is simple but demanding. If you commit to a channel, commit to it for at least six to twelve months at a consistent spend, not three bursts of two months. Money spent at KSh 20,000 per month over twelve months will almost always outperform the same total budget spent in two-month bursts. Consistency builds a data pool, a retargeting audience, a content library, a follower base, and an algorithmic preference. Each month makes the next month cheaper. Inconsistency rebuilds everything from scratch every time. If you cannot afford to commit to a channel for a year, do not start on that channel yet — start on a cheaper one and save up.
Use discounts sparingly — they teach customers you are cheap
The reflex to discount is strong, especially when sales slow. Resist it unless you are in fast-moving consumer goods, where price is genuinely the product. For most other categories — fashion, electronics, furniture, services, specialty food — discounts quietly erode the brand. A customer who bought your product at KSh 3,000 three months ago and sees it at KSh 2,400 today learns two things: that the real price was always 2,400, and that if they wait, they can always get a discount. You have just trained them to never pay full price.
Instead of discounts, think about value additions. A free accessory with a purchase. Complimentary wrapping. Faster delivery for repeat customers. An extra year of warranty. Something that costs you little but feels meaningful to the customer. Value additions preserve price integrity while delivering the same emotional pull. Customers remember being given something; they also remember being given a discount, but the memory of a discount makes your original price feel fake.
The quickest win is customer service etiquette
This is the single cheapest, most neglected competitive advantage in Kenyan retail today. In a world of chaos, of apps crashing, of unanswered calls, of rude sales people, of confused service counters — a polite, honest, attentive human voice is a minor miracle. The businesses that get this consistently right have customers who forgive almost any other flaw.
After a long day, on the way home, when I need to grab something I need a human touch. Whether on a phone call or in a shop. That is what I am paying for as much as the product itself. And yet Kenyan retail in 2026 is full of staff who answer the phone with an exhausted "eeeh", who make customers feel like interruptions, who give wrong information confidently rather than correct information slowly. Every interaction like this is a customer considering a different shop next time.
Training your customer-facing staff in the basics of etiquette — how to greet, how to listen, how to apologise, how to escalate, how to close a conversation warmly — is probably the highest-ROI investment any Kenyan retailer can make this quarter. It costs little and it compounds forever. Customers tell their friends about rude service more often than they tell them about good service, which means the downside of getting it wrong is much larger than the upside of getting it right. Start there before you start anywhere else.
AI is here — but do not rush it into the places where humans matter
I have used AI extensively while writing the paper, and I am not going to pretend it is not a remarkable tool. But one of the sharpest lessons from working with it is knowing where it belongs and where it does not. AI belongs in the back office. It is excellent for drafting content ideas, managing inventory, analysing customer data, enhancing product photography, answering first-pass queries in written channels, and generating the variations of text that marketing teams spend hours on. Use it there. Use it heavily.
AI does not belong in the places where the human being is the product. Do not generate AI faces for your team photos. Kenyan customers know your staff, and they will notice if the person in the "meet the team" photo is someone the algorithm invented. Do not fill your TikTok feed with AI-generated humans pretending to be real people. Short videos of your actual team members, however imperfect the lighting, will always outperform polished AI content in this market — because the audience is buying the humanity as much as the product. Do not replace your customer service calls with a bot unless the bot is clearly labelled as such and the option to reach a human is one tap away. The moment a customer realises they are being fooled, the trust is gone.
A useful test: use AI where it removes friction from your work, not where it removes humans from your customer's experience. Use it to make your team more productive. Do not use it to pretend you have a team. The businesses that get this balance right will outrun both the ones who refuse to adopt AI and the ones who shove it into every customer interaction. We are years away from a world where AI can replace the human touch in retail. We are here, now, in a world where AI can amplify human work dramatically. Those are very different things.
The single action to take this week
If you read this and do only one thing, let it be this. Go back to your customer records — whatever form they take, whether a proper database or a WhatsApp contact list or a book of phone numbers. Identify the ten customers who you know would talk about your business without being paid. The ones who already do, quietly. Then call them. Not a text. Not a broadcast. A phone call. Ask how they are. Ask what they think of your last interaction with them. Tell them you are grateful and that you would love to get their input on something you are working on. Do not pitch, do not sell, do not ask for anything.
That call is your retail strategy for 2026, in its simplest form. It is the blend of physical and digital the paper argues for. It is the recognition that customers are the ambassadors. It is the marketing budget spent with expected return. It is the customer service etiquette made visible. It is the deliberate refusal to let AI do the thing only a human being should do. Ten calls, one week. If you find that meaningful, the rest of this piece will find its way into your business naturally.
The future of Kenyan retail is not being decided in Silicon Valley boardrooms or venture capital pitch decks. It is being decided in thousands of small daily choices by thousands of business owners who have not yet realised how much power they have. The blended decade belongs to the ones who figure it out first.
Comments (0)