Zimbabwe Market Context (2026)¶
Why this exists: Every blueprint in this repo makes assumptions about connectivity, payments, device quality, and regulation. Those assumptions are stated here once, with sources, rather than repeated (and drifting) across five documents.
Rule: If a number here isn't sourced, it's marked [ESTIMATE] with its basis. Do not upgrade an estimate to a fact without a source.
1. Connectivity and devices¶
| Metric | Value | Source |
|---|---|---|
| Population | ~16.8M | POTRAZ population base used for penetration ratios |
| Active mobile subscriptions | ~16.8M lines, ~107% penetration | POTRAZ Q4 2025 |
| Active internet/data subscriptions | 12.99M (Q3 2025) → ~84.55% penetration (Q4 2025) | POTRAZ Abridged Sector Performance |
| Household internet access | 75.5% (2025), up from 50.1% in 2020 | ZIMSTAT household survey |
| Smartphone share of connections | [ESTIMATE] 55–65% |
Derived from POTRAZ broadband vs total subscriptions; feature-phone base remains material |
| WhatsApp share of mobile data traffic | 20.69% of all mobile internet traffic (Q4 2025) | POTRAZ |
| Mobile data cost | High relative to income; ~US$1–2 per GB depending on bundle | [ESTIMATE] operator bundles |
Engineering consequences:
- Android-first, low-end. Assume a US$60–120 Android device, 2–3GB RAM, limited storage. Budget: initial JS < 200KB gzipped, session data < 500KB.
- WhatsApp is the operating system. 20.69% of all mobile data traffic is WhatsApp. Any notification strategy that doesn't route through WhatsApp is fighting the grain. Push notifications are unreliable on cheap Androids with aggressive battery managers.
- Offline tolerance is mandatory. Data is metered and coverage is patchy. Core write actions (log a fill-up, log a service) must work offline and sync later.
- USSD still matters at the edges. For garage-side or non-smartphone users, USSD/SMS remains the only universal channel. Consider it for v2, not v1.
2. Payments¶
| Rail | Position | Integration path |
|---|---|---|
| EcoCash | ~86% of mobile money market; >70% of all national payment transactions | EcoCash Open API v2.0 (sandbox + certification) or Paynow |
| OneMoney (NetOne) | ~11% of mobile money | Paynow aggregation |
| InnBucks, O'mari, others | ~3% combined | Paynow where supported |
| Visa/Mastercard | Low penetration among consumers; common for business | Paynow / ZimSwitch |
| Paynow | Leading aggregator; 1.5% + US$0.50 per transaction | Node SDK available |
| Cash USD | Still dominant for informal commerce | Manual reconciliation required |
Engineering consequences:
- Never build card-first. A checkout that requires an international Visa will lose most of the market.
- Paynow first, direct EcoCash later. Paynow gets you EcoCash + OneMoney + cards in one integration and removes the certification burden. Revisit direct EcoCash Open API only when volume makes the ~1.5% worth optimising.
- Price in USD, display ZiG. USD is the mental unit for most pricing decisions. ZiG display should be a live-rate conversion, never stored as the source of truth.
- Annual, not monthly. Recurring small debits are disliked and technically fragile on mobile money. Sell US$8/year, not US$1/month.
- Cash collection is a real workflow. Garage partners may pay in cash. The system must support "marked paid, cash received by rep" as a first-class state, not a hack.
3. Energy (this changes the ranking)¶
| Fact | Value | Source |
|---|---|---|
| Load-shedding status | 304 consecutive days without scheduled load-shedding as of 28 Oct 2026 | ZESA via Pindula / The Herald |
| ZESA target | End load-shedding by Dec 2026; end power imports by 2027; full electrification by 2030 | ZESA Group CEO, May 2026 |
| Supporting facts | 138 days without load-shedding by May 2026; US$210M Afreximbank facility; active SAP trading | Energy News Africa |
| Risk factor | Lake Kariba levels receding seasonally (Oct 2026) — drought can reverse the position | Zambezi River Authority via Maravi Post |
| Fuel | Petrol ~US$2.23/L, diesel ~US$2.11/L (Apr 2026); all diesel taxes removed | ZERA |
Consequence: The original "ZESA + water + daily life" idea was built on a 2023–2024 reality that no longer holds. Load-shedding is not currently a daily pain. An app whose core promise is "power going off in 30 minutes" is solving a problem that has, for now, stopped occurring — and ZETDC ships its own official app. See 03-parked/zesa-water-daily-life.md for the full verdict and the revival trigger.
The water half is still live. Harare's water supply and rationing remain a genuine daily problem, but it is a different product with a different data source (city council schedules, borehole crowdsourcing) and a much weaker monetisation path. Do not conflate the two.
4. Vehicles (the strongest market in the idea set)¶
| Metric | Value | Source |
|---|---|---|
| Registered vehicle fleet | 1.52 million (2025), +44.7% since 2020 (+469,000 vehicles) | ZINARA / Equity Axis |
| New monthly registrations | 10,000–11,000 | ZINARA |
| Licensed vehicles | ~827,000 (2024), target 1,000,000 (2025) | ZINARA |
| New light vehicles registered (Q4 2025) | 19,715, +18.4% QoQ | ZIMSTAT Transport Statistics Q4 2025 |
| Harare car repair businesses | ~155 (146 single-owner = 94.19%) | Rentech Digital business census |
| Dominant vehicle profile | Used Japanese imports; high CVT/automatic share; aging fleet | [ESTIMATE] from import patterns |
Interpretation:
- The fleet is growing at roughly 9% per year and adds 120,000–132,000 vehicles annually. Every one of those needs servicing, licensing, insurance, and parts.
- The repair sector is almost entirely micro-business: 94% single-owner. This means (a) no incumbent software, (b) high fragmentation, © low ability to pay individually, (d) enormous value in aggregation.
- ZINARA licensing and insurance expiry are recurring, date-driven, government-mandated obligations. These are perfect app hooks because the deadline is externally enforced — the user doesn't have to care, the law does.
- Used Japanese imports with CVT gearboxes are a specific maintenance trap: the fluid is expensive, widely misunderstood, and skipped by owners until the gearbox fails. This is a sharp, credible wedge.
5. Competition snapshot¶
Revised 2026-11-10 after discovering ZiMINDEX, ZWPRICE, DotCompare, and ZimSeek. The price-intelligence space is far more contested than the first pass found.
| Space | Existing players | Implication |
|---|---|---|
| Price index / cost of living | ZiMINDEX, DotCompare, zimpricecheck.com, shopzim.co.zw, ZimSeek, zimqoute.co.zw | Six active players. The category is occupied, not empty. See §5.1. |
| Quote / RFQ marketplaces | ZWPRICE (building, fuel, wholesale, farming) | The RFQ mechanic is proven in-market — see §5.2 |
| Load-shedding | ZETDC official app, poweroutage.live, ZESA self-service portal | Crowded by the incumbent with authoritative data |
| Transport | Hwindi (taxi/food/delivery super-app), Whindy, ZimTap | Well-funded-ish, already moving |
| Car care | No Zimbabwe-specific consumer product found; only global generic trackers (MyAutoLog, Vehicle Maintenance Tracker, WhenService) | Genuine white space — and the only major category on this list that is still empty |
| Household management | Global free apps (AnyList, OurGroceries, Spondle, NestSync) + WhatsApp itself | Saturated, free, and good enough |
5.1 ZiMINDEX — the one that changes the analysis¶
zimindex.co.zw describes itself as "the intelligence layer for everyday life in Zimbabwe." That is an accurate description of what it actually does. It is not a price comparison site; it is a cost-of-living data platform.
What it ships today:
| Surface | Detail |
|---|---|
| Macro dashboard | ZiG/USD (26.52), ZAR/USD, diesel and petrol per litre, USD and ZiG inflation, provincial inflation breakdown |
| Poverty lines | Per person, food-only and food-plus-everything, translated to household size |
| Product price database | Per-product US$ and per-unit pricing, multi-shop ranges (e.g. US$0.60–0.65, 2 shops), with staleness shown per item (4 days ago) |
| Baskets | "Medium family basket" — 25 items, totalling US$155.23, in USD and ZiG |
| Cost calculators | Electricity by appliance (stove, fridge, kettle, lights, TV, iron → kWh and US$), rent by property type |
| Household planner | /plan — build a monthly picture by household composition (adults, children) |
| Statutory fee index | Indexes traffic fines as priced products (e.g. US$30 for a beacon-light offence under S.I. 129/15) |
| Crowdsourcing | "Know a different price?" → /contribute |
| Data provenance | ZIMSTAT cited as the statistical source |
Why this matters more than the others: four of the five attributes I designed into the Mota price index as differentiators are already live here, executed well:
| Feature I specified | ZiMINDEX status |
|---|---|
| Every price shows its date | Shipped — per-item staleness (4 days ago) |
| Price ranges from multiple suppliers | Shipped — US$0.60 – US$0.65 · 2 shops |
| Crowdsourced price contribution | Shipped — /contribute |
| Category-level normalisation (per kg, per litre) | Shipped — US$13.20/kg |
| Sponsored placement labelling | Not observed |
The Cloudflare failure is a demand signal, not a weakness. The site is returning Worker exceeded resource limits on several routes (/baskets/*, /calculators/*, /plan partially). That is a free-tier worker limit being exceeded — which means real traffic. A dead side project does not exhaust Cloudflare workers. An incumbent failing under load is a scaling problem, not an opening. Building a business whose only advantage is "the incumbent is down right now" is not a strategy.
The specific threat to the vehicle idea: ZiMINDEX already indexes statutory fees as priced line items. That is the same data model required for ZINARA licensing, insurance premiums, and roadworthy fees. The gap between "we index traffic fines" and "we index ZINARA renewal costs" is small, and the domain knowledge is already in-house. This is the most credible adjacent-entrant path into Car Care's territory found so far — not because they would build a maintenance tracker, but because they could credibly own the deadline and cost half of it.
What ZiMINDEX does not have: vehicles, service history, maintenance intervals, garages, parts pricing, or any mechanic-to-owner relationship. The per-vehicle record — the actual thing Mota exists to hold — is absent, and it is not a natural extension of a basket-costing engine.
5.2 ZWPRICE — independent validation of the lead mechanic¶
zwprice.com is an RFQ (request-for-quotation) marketplace, not a price index. The buyer describes what they need in natural language, ZWPRICE routes the request to relevant suppliers, suppliers submit quotes, and the buyer compares and orders. Free for buyers; suppliers pay for access.
Coverage: building and hardware (PPC Cement, Union Hardware, Beta Bricks, Nash Paints), fuel (TotalEnergies, Puma, Engen, Zuva), wholesale groceries (N. Richards, National Foods, Gain Cash & Carry), farming (Seed Co, ZFC, Windmill). No automotive service or repair category.
Why this is good news for Mota: ZWPRICE independently arrived at the same commercial mechanic the Mota blueprint specifies. Their published trust model is a near-exact match for the design in 07-interface-design.md §8:
| ZWPRICE behaviour | Mota equivalent | Read-across |
|---|---|---|
| OTP to phone/email before a request is sent | FR-VEH-01 (phone OTP) | Validated |
| "Suppliers see what you need and where, never your contact details" | FR-GAR-07 / PII boundary | Independently validated — this is the correct design |
| Contact released only to the supplier you choose | contact_released_at gate |
Independently validated |
Supplier labels: Listed / Claimed / Reviewed by ZWPRICE |
verification_status (binary) |
Better than mine — three tiers is right. Adopt. |
| "Delivery not yet quoted" is shown honestly and never counted as free; an option with a missing price is never ranked above a fully priced one | Not specified | Adopt. Directly applicable to parts-vs-labour pricing |
| "We are not the seller… no escrow or payment protection" | Not specified | Adopt as the disclaimer template |
| Free for buyers, suppliers pay | Consumer-free, garage-paid | Validated |
What this changes: assumption A2 ("garages will pay US$2–5 per attributed booking") is no longer purely speculative. A comparable Zimbabwean marketplace operates on supplier-paid, buyer-free economics across 26 supplier brands. It is evidence, not proof — different category, different buyer — but it moves the assumption from [ESTIMATE] to [SUPPORTED BY ANALOGY].
What it does not change: ZWPRICE proves that suppliers will join. It does not prove they will pay, and it does not prove the model works in a category where the buyer is an individual consumer rather than a contractor. Both remain Gate 0 / Gate 3 questions.
5.3 What the expanded competitor set actually tells us¶
Three conclusions, each of which shapes the ranking:
- Price intelligence in Zimbabwe is a crowded, well-executed category. Six players, one of them operating at a level of polish and data breadth that would take months to match. Idea #4's competition score must fall accordingly.
- The RFQ/lead mechanic is commercially proven in this market. This is upside for Car Care — it de-risks the business model, not the competitive position.
- Vehicle maintenance remains the only major everyday category with no local software. Everywhere else in this document, someone is already there. In car care, no one is. That asymmetry is now much more visible than it was in the first pass — and it is the strongest single argument for building Mota.
6. Regulation¶
Cyber and Data Protection Act [Chapter 12:07] (2021)
- POTRAZ is the designated Data Protection Authority.
- Data controllers must be licensed/registered — see POTRAZ Guidelines on Licensing of Data Controllers (2025) and SI 155/2024.
- Lawful processing requires a stated basis, purpose limitation, and data-subject rights (access, correction, deletion).
- Transfers of personal data outside Zimbabwe are restricted.
Engineering consequences:
- Hosting choice is a legal decision, not just a technical one. Supabase/Vercel default to US/EU regions. Whether that's permissible under the CDPA cross-border provisions needs a lawyer, not a guess. Flagged as an open question in every blueprint; interim mitigation is to minimise PII and prefer an African region where available.
- Minimise collected PII. Phone number + first name + vehicle is enough to run the product. Do not collect ID numbers, addresses, or date of birth unless a feature genuinely requires it.
- Ship a privacy policy and a deletion path before launch, not after. Deletion must actually delete.
- Payment data stays with the processor. Never store card or wallet credentials.
7. What this means for the idea set¶
Three conclusions fall out of the data before any scoring:
- The load-shedding thesis is dead for now. The pain it monetised has been absent for ten months, and the incumbent utility ships an official app.
- Vehicles are the strongest structural market. 1.52M and growing 9% a year, 94% of the service sector unsoftwareed, government-mandated recurring deadlines, and no local competitor.
- Price comparison is a crowded category with a brutal data-freshness cost. Six active players, one of which (ZiMINDEX) already ships per-unit normalisation, multi-shop ranges, per-item staleness, crowdsourced contribution, and ZIMSTAT-sourced macro data. The mechanic is viable as a feature attached to a paying audience, not as a standalone startup.
Conclusion 3 is the seed of the portfolio strategy: car care is the beachhead, and parts pricing is the moat. That is the thesis the ranking formalises.