The State of African Business 2026: One Million Companies, Analysed

Africa's formal economy is usually described with estimates and projections. This report uses records. Fylings maintains the largest open collection of African company data we know of, 1,006,104 company records across 20 national registries, and this is the first time we've stepped back to read the whole dataset at once. The picture that emerges is of a continent registering businesses faster than ever, formalising at scale, and, for anyone who has to trust these companies, carrying real risk that only shows up when you check the record. Here is what a million African company records actually say.

1. Company formation has grown roughly eightfold since 2000 in the registries we track

The clearest signal in the data is momentum. Among the 832,150 companies whose registration date is published, annual new registrations have climbed from around 9,300 in 2000 to 76,812 in 2025, the highest year on record, and roughly eight times the level at the turn of the century. The curve is not smooth growth; it bends sharply upward after 2016, and again after the pandemic:

• 2000: ~9,300 new registrations
• 2010: ~16,900
• 2016: ~28,800, the inflection point
• 2020: ~41,600
• 2023: ~55,800
• 2024: ~67,700
• 2025: ~76,800, a record

Registrations more than tripled in the decade to 2025 (up from ~21,000 in 2015). The post-2020 surge is striking: rather than the pandemic suppressing company formation, the years since have produced the four largest registration cohorts in the dataset, consistent with a wave of digitisation, online registration systems coming online, and a young population starting businesses.

An obvious objection deserves a direct answer: could this curve simply reflect registries moving online, with older records never digitised? Partly, yes, and we say so. The aggregate 8× conflates real formation growth with expanding registry coverage. But the trend survives when we hold coverage constant. Within Mauritius's CBRD alone, a registry with consistent published records back to 1990, annual registrations grew 2.8× over the same period, from ~4,600 in 2000 to ~13,100 in 2025, with the same post-2020 acceleration. Zambia's PACRA shows the same shape at 4.9× (~1,500 → ~7,400). Digitisation amplified the curve; it did not create it. The underlying signal, Africans registering formal businesses at a historically unprecedented rate, holds inside individual registries with consistent coverage, at three to five times the level of 2000.

2. Governments have awarded half a million public contracts

Beyond the registry itself, we match companies to public procurement records. Across the continent, the dataset now holds 525,149 government contract awards, linked to 39,591 distinct companies, a rare, structured view of where public money actually goes. By volume, Tanzania leads with more than 308,000 awards, followed by Kenya (~78,000), Uganda (~66,000) and Nigeria (~31,000). In Tanzania alone, awards carrying a published value total over 22 trillion shillings (~US$8.7 billion).

What governments buy is just as telling as how much. The most active public buyers in the data cluster around a handful of essential functions:

• Health, hospitals and national drug authorities are among the single largest buyers by contract count
• Roads and infrastructure, national roads authorities award thousands of contracts each
• Agriculture, agricultural research and development bodies are consistently high-volume
• Elections and revenue, electoral commissions and revenue authorities feature prominently
• Water and utilities, national water and sewerage corporations round out the top buyers

3. Only about half of registered companies are active

The most important finding for anyone relying on these companies is also the least visible: a registry entry is not a guarantee that a business is trading. Across the full dataset, status breaks down as follows:

• Active: 49% (492,003 companies)
• Status unknown or not published: 25% (248,435)
• Dormant: 13% (135,130)
• Dissolved / struck off: 13% (130,536)

Put plainly: more than one in four registered companies in the data is dormant or has been struck off the register. A name on a certificate, or even a real registration number, tells you the company existed, not that it exists today. This is the entire case for checking a company's live status before paying it, lending to it, or onboarding it, and it is why every record on Fylings carries the status and the date it was last verified.

4. Country snapshots

The pan-African picture is made of very different national ones. Five markets, five stories, each figure below comes from that country's official registry and open-contracting data, with the caveats stated where the source is incomplete. US-dollar equivalents use July 2026 rates.

🇳🇬 Nigeria, the verification-risk story

207,675 companies on record · CAC. More than half, 52%, are flagged inactive by CAC itself, against 45% confirmed active. (Method note: this is CAC's own status flag, not our inference; CAC typically marks a company inactive for unfiled annual returns, so it signals a lapsed company, not necessarily a formally struck-off one, either way, a company you should check before trusting.) Nigeria's open-contracting portals add 30,760 contract awards matched to 12,697 companies, worth ₦6.30 trillion (~US$4.6 billion) where values are published. Our Nigeria set is a search-based sample of CAC, not a full sweep, so we quote shares, not national totals.

🇹🇿 Tanzania, where the public money shows

156,122 companies on record · BRELA. Tanzania publishes the deepest open-contracting data on the continent: 308,769 contract awards, more than the rest of our dataset combined, matched to 10,265 registered companies, with disclosed values totalling TSh 22.83 trillion (~US$8.7 billion) plus US$258 million in dollar-denominated awards. BRELA reports 97% of companies as active, though it rarely records dissolutions, so read that as an upper bound.

🇬🇭 Ghana, a business-name economy

107,224 entities on record · ORC. Ghana's register is dominated by sole-trader business names, 86,109 of them against 21,008 limited companies, roughly four to one, the highest ratio among our large registries and a snapshot of an economy still early in formalising. Ghana's procurement portal adds 9,651 awards worth GH₵2.25 billion (~US$198 million) in disclosed value. ORC publishes neither status flags nor incorporation dates in bulk, so Ghana sits outside our formation and status analyses.

🇲🇺 Mauritius, the deepest register, and half of it dissolved

227,308 companies on record · CBRD. Mauritius runs the most complete open registry in our dataset, with consistent records back to 1990, which is why it anchors the robustness check in Section 1 (formation up 2.8× since 2000). Its most striking number cuts the other way: 52% of every company ever registered, 118,608, is dissolved, the highest proportion anywhere in our data. A register this honest about corporate death is rare; it is also a reminder that most companies, everywhere, do not last.

🇿🇲 Zambia, formalisation in motion

111,807 entities on record · PACRA. Zambia splits into 65,483 sole-trader business names and 46,309 limited companies, and its full-history register shows formation up 4.9× since 2000 (~1,500 → ~7,400 a year), the second leg of our robustness check. PACRA also publishes real status data: 75% active, 24% dormant. Zambia's procurement records add 23,588 awards to 5,453 companies, worth K80.7 billion (~US$4.4 billion) in disclosed value.

5. The trader economy: what African businesses actually do

The dataset also settles a simple question, what is the typical African business? By legal form, formal companies (494,162) outnumber sole-trader business names (377,608) by roughly four to three, a snapshot of an economy still formalising. By activity, the backbone is trade and services rather than industry. The most common declared activities are:

• Non-specialised wholesale and general trade, the single largest category
• Business and professional services
• General contracting and merchandise
• Retail, stores, stalls and non-specialised outlets
• Mixed farming and agriculture
• Construction of buildings and road freight transport

The long tail is human and specific, hairdressing and beauty, tailoring and fashion, small-scale mining support, microfinance. It is the texture of a continent of traders, service providers and builders, not the resource-and-conglomerate caricature.

Where the data comes from

The one million records span 20 national registries, 18 fully tracked, plus South Africa and Kenya as smaller paid-source samples (which is why other pages on Fylings say "18 registries": that count is the fully-tracked ones). The dataset is weighted toward the largest open registers: Mauritius (227k), Nigeria (208k), Tanzania (156k), Zambia (112k) and Ghana (107k), with Madagascar, Uganda, Seychelles, Namibia, Malawi, Senegal, Angola, Burkina Faso and the OHADA francophone states adding depth. Every record is sourced directly from an official national registry, Nigeria's CAC, Tanzania's BRELA, Mauritius's CBRD, Senegal's RCCM and others, and is free to search by name or registration number at fylings.com.

Methodology and limits

We believe in showing our working. This is the largest open dataset of African companies we know of, but it is not an official census, and it should not be read as one:

• Coverage varies by country. The dataset reflects what each national registry publishes openly; it does not yet include full South Africa, Kenya, Egypt or the DRC, and no registry is 100% complete.
• Dates exist on ~83% of records (832,150 of 1,006,104), skewed toward registries that publish incorporation dates, so the formation trend is directional, strongest for those registries, not a continent-wide count. A handful of erroneous future/ancient dates were excluded.
• Sector labels are inconsistent free text across registries and are grouped only approximately.
• Procurement values span currencies and are not summed across them; per-country figures are quoted in the local currency of the award.
• Figures are current as of publication and grow as we add registries and records.