M&A across African tech nearly doubled in the first half of the year while equity dried up. The default way to grow flipped from raise to buy, and consolidation stopped being the risk and became the base case.
A weekly intelligence brief from Base X Studio. Pan-African tech, global brand strategy.
African tech raised $1.44bn in H1 2026, a hair above last year's $1.42bn, so the headline reads steady. It is not. The same money came through 146 disclosed deals against 252 a year earlier, which means capital is concentrating into fewer, larger, more heavily scrutinised bets while the number of companies that can raise at all is collapsing. Nigeria overtook Egypt as the most active market, and fintech and mobility took nearly half of every deal.
If you are planning a raise on the assumption that the 2021 market still exists, the deal count says otherwise. The money is there, but it is going to fewer, more legible companies, so your job this quarter is to be one of them. Tighten the story until an investor can repeat it back in one line, because the bar is now who gets read closely, not who shows up.
Across 40 deals. A stronger opening quarter that carried most of the H1 total and set the "flat versus last year" story.
Across 38 deals, down 40% on Q2 2025. Hold this pace and the full year lands near $1.28bn against 2025's $1.64bn.
Do not let a flat sector headline set your own runway assumptions. The quarter-by-quarter line is falling, so plan against Q2's pace, not H1's average. If your model needs the market to have recovered, it has not, and the safer read is that the next round is harder and further out than the annual number suggests.
The shared, neutral TowerCo model that let every operator rent the same masts on equal terms is reversing into captive ownership by the dominant carrier. Airtel now faces renting space from its biggest competitor, and Nigeria has opened a competition review. Across payments, licences, and now towers, partnership keeps being replaced by ownership.
If your cost base rests on neutral shared infrastructure, model the day a competitor owns it. The shift is from renting off a neutral party to renting off a rival, and that is a pricing and dependency exposure to surface now, not after the next tower deal closes. Ask what part of your rails you would want to own before someone else does.
The continent already carries 74% of global mobile-money volume, and the next phase is depth, credit, institutional design, cross-border integration, not just reach. A Rwanda-Kenya licence-passporting MOU now offers regulatory harmonisation as an alternative to buying your way across borders through M&A. Two routes to the same regional scale, with different economics and different consequences for your brand.
If regional expansion is on your roadmap, decide which playbook you are running before you spend on either. Passporting is cheaper and keeps your brand whole where the MOUs reach; acquisition is faster and buys licences but absorbs identities. Pick the route that matches what you are optimising for, speed or coherence, and stop treating "go regional" as a single decision.
With organic market entry costing $3-5m and equity dried up, acquisition has become the rational way to buy a licence, skip build time, or move into infrastructure. Every acquired brand gets folded, retired, or repositioned in weeks, by people who did not build it.
Jasmine Bina argues the market has left a twenty-year optimization plateau, where precision and specialist mastery were rewarded, and entered a roughly ten-year exploration phase driven by AI, where no best practices exist yet and old expertise can actively mislead. Her reframe is precision versus accuracy. A tight, confident, well-executed answer built on old data is precisely wrong once the target, the market, has moved. The response is to make experimentation a default and treat failure as data, not as a strategist's identity crisis.
If your strategy rests on being the most expert operator in a settled category, check whether the category is still settled. A precise plan built on last year's data can be confidently wrong, so build a cheap way to test your biggest assumption this quarter instead of polishing the plan. Treat a failed test as information, not as evidence you were not expert enough.
Six years after Obviously Awesome, Dunford is releasing an expanded edition on a harder claim. Positioning is not a one-time exercise. It has to be re-run every time the product, the competition, or the market shifts, and in 2026 all three are moving at once. She adds a sharp test for multi-segment companies: your differentiated value is usually identical across segments, and if it genuinely differs, that is a product-strategy problem wearing a positioning costume.
Re-run your positioning against your real, current competitors this quarter, and validate it in actual sales conversations rather than another website rewrite. If you run multiple segments on the same core value, stop writing three positions where one will do. Positioning that was right two years ago is quietly wrong now.
A differentiated value that a competitor could copy-paste onto their own site is not a position, it is a filler phrase. And "we use AI" has joined "we save time" in that bin, because everyone says it. The work is to name something concrete and specific enough that it could only be true of you, then hold it steady for at least six months instead of tweaking the copy every month.
Audit your homepage for placeholder value. If your differentiator would read as true for three of your competitors, it is not doing any work. Replace it with the one concrete thing a buyer could not lift onto a rival, and resist the urge to rewrite it again next quarter.
Naming the problem in new terms and owning the language an industry thinks in produces a more defensible advantage than positioning inside an existing category.
As AI increasingly filters buyer decisions, stable decision filters matter more. Brand systems that hold up under fast content cycles beat static guidelines.
Founders reach for category creation because it sounds visionary. Positioning is the unglamorous work that has to come first, and skipping it is avoidance.
If you cannot yet say who you beat and why in one sentence, you are not ready to name a new category. Clarity precedes scale, and the field keeps arriving at it.
Do not reach for category design before you can state your position clearly inside the category you are already in. Get specific first. If a sceptic on your team cannot repeat who you beat and why in one line, that is the work to finish before anything more ambitious.
Read the consolidation wave against the fewer-bigger funding pattern and the passporting-versus-acquisition fork, and it is the single most structural shift across both tracks. It is not a funding story, it is a change in how the market grows. With 63 M&A deals in the first half against 33 a year earlier, acquisitions have become the rational route to a licence, a market, or an infrastructure layer, precisely because organic entry costs $3-5m and equity has dried up. The brand consequence is immediate and mostly unmanaged: an acquired company's identity is decided in weeks, folded, retired, or repositioned, by people who did not build it. Consolidation is your base case now, so pick your side of it this quarter. Either become specific enough that folding you in would destroy the thing that made you worth buying, or become legible enough that an incumbent sees you as the capability they cannot build fast enough. The position that loses is the mid-scale generalist who is neither, because that is exactly who gets absorbed cheaply and erased.
Bina's precision-versus-accuracy reframe raises an honest question for any strategy discipline. Clarity is easy to demand in a settled category and much harder when the category itself is still forming. The move is not to abandon rigour, it is to trade a little precision for accuracy: hold the plan loosely enough to correct it, and test the one assumption everything else depends on before the market tests it for you.