Kombi / Transport — PARKED¶
Original idea: Real-time kombi route availability, today's fare, queue location. Plus a daily takings tracker for drivers.
Rank: 4 of 5. Score: 5.40.
Verdict: Best market on the list, worst fit for a 1–2 person team with no capital.
Reopen if: a kombi association partnership is in hand, or a funded team is assembled.
1. The market is genuinely the biggest¶
This is not a marginal idea. It scores 9 on pain and 10 on market size — the highest of any idea evaluated.
| Fact | Implication |
|---|---|
| Kombis are the backbone of urban transport in Harare | Near-universal daily use among the working population |
| No metro, no reliable schedule, no official route map | The information vacuum is total |
| Fares change with fuel prices (petrol ~US$2.23/L) | Riders genuinely don't know today's fare |
| Queues are long and unpredictable | Time cost is real and daily |
| Drivers/conductors track takings manually | A cash business with no visibility |
The pain is daily, universal, and unsolved. If the only criterion were "size of problem," this wins.
2. Why it still loses¶
The score collapses on buildability (2) and monetisation (3). Those aren't soft factors — they're the difference between a plan and a fantasy.
2.1 The data problem is a hardware problem¶
Real-time kombi tracking requires knowing where each vehicle is. That means one of:
| Approach | Reality |
|---|---|
| Install GPS devices in every kombi | Capital cost per vehicle, installation labour, theft, maintenance. For thousands of vehicles this is a hardware company, not an app company. |
| Drivers run a smartphone app | Drivers are cash businesses with thin margins. They have no incentive to share location, and every incentive to hide their route from competitors and from enforcement. |
| Crowdsource from passengers | Requires dense simultaneous usage to infer a route's state. Cold start is severe, and data quality is poor (a passenger's GPS says where they are, not where the kombi will be in 20 minutes). |
| Conductor reports at each stop | Depends on a paid operator per route — a field workforce. |
| Partner with the association | The only viable path. Also the hardest to obtain. |
There is no cheap way to get this data. That single fact determines the outcome.
2.2 The incentive problem is worse than the data problem¶
Kombi operators are not neutral about information.
- Route information is competitive advantage. Telling everyone where your kombi is means competitors intercept your passengers.
- Fare transparency is against the operator's interest. Published fares reduce pricing power during peak times and disruption.
- Queue information invites enforcement. Police and council enforcement use visibility of where kombis are loading.
- Associations are political. Route allocation is contested and regulated. A neutral technology platform is rarely neutral in that fight.
An app that makes the market more transparent is an app that takes value away from the people who control supply. They will not cooperate, and without their cooperation there is no data.
2.3 The competition is already moving¶
| Player | Position |
|---|---|
| Hwindi | Super-app: taxi booking, food, groceries, delivery. Already operating, already has drivers and payments. |
| Whindy | Mobile-first bookings, route optimisation, pricing, tracking, payments. |
| ZimTap | Explicitly an ecosystem play — cards, mobile money, kombi associations, city authorities. |
| Informal WhatsApp groups | Route-specific groups already share queue and fare information for free. |
The last one is the most underrated. Riders already solve this in WhatsApp groups, imperfectly but at zero cost. Any product must beat that, not just beat nothing.
2.4 Monetisation is thin¶
| Stream | Assessment |
|---|---|
| Consumer subscription | No. Riders will not pay for fare information. |
| Ads | Low value, high annoyance, and riders are in a hurry. |
| Driver SaaS (takings tracker) | Plausible, but drivers are cash businesses with low willingness to pay, and a simple notebook works. |
| Commission on bookings | Doesn't fit the kombi model — you don't book a seat in advance; you queue and board. |
| Association/authority contracts | The only real revenue. Requires the partnership we don't have. |
3. The honest comparison¶
| Dimension | Car Care | Kombi |
|---|---|---|
| Data acquisition | User enters their own data, free, willingly | Requires hardware, field staff, or operator cooperation |
| Data freshness needed | Monthly (service intervals) | Seconds (real-time position) |
| Incentive alignment | Garage gains customers; owner gains protection | Operators lose pricing power and privacy |
| Cold start | Useful at 1 user (it's personal) | Useless below dense network coverage |
| Capital required | ~US$100 | Thousands in hardware and field ops minimum |
| Solo buildable | Yes | No |
| Competitors | None local | Three funded-ish players |
| Realistic 12-month outcome | US$600 MRR | Zero revenue and a stalled pilot |
The comparison isn't close, and it isn't close because of the market — it's close because of the shape of the market. Car Care's data is volunteered. Kombi's data must be captured, and the people who control it are structurally opposed.
4. What would change the verdict¶
| Trigger | Why it changes things |
|---|---|
| A kombi association offers a data partnership | Solves the incentive problem and provides the data source. This is the only realistic path. |
| A funded team with hardware experience | Makes the capital and field-ops problem tractable |
| City of Harare / ZUPCO digitisation mandate | Creates a compliance-driven data source, similar to how ZINARA mandates licensing |
| ZimTap or Hwindi opens an API with route data | Removes the data problem entirely; a thin layer on top becomes possible |
Note: all four are external events. None is something we can create by building. A business that requires someone else's decision to become possible is not ready to start.
5. The reducible slice — if we ever revisit it¶
Not the full vision. The smallest thing that could work:
A fare index, not a tracking app.
| Aspect | Detail |
|---|---|
| What | Crowdsourced current fare by route, updated daily by riders. "CBD → Chitungwiza: US$2.00 as of 07:40 today." |
| Why it works | Only needs one report per route per day, not real-time positions. Massively lower data density requirement. |
| Where the data comes from | Riders, in a 15-second submit flow, rewarded with contribution credit |
| Monetisation | None at first. Later, sponsored placement or a partnership with a delivery operator. |
| Why it's still hard | Cold start per route; accuracy disputes; no revenue |
| Why it might be worth trying | It's a weekend project, not a company. It could validate whether riders will contribute data at all. |
This is recorded as a possible cheap experiment, not a plan. If Car Care reaches a stable state and there's appetite to test a second idea, a fare index is a far better first move than a tracking app — it tests the core assumption (will riders contribute?) for near-zero cost.
6. The transferable lesson¶
A big problem is not the same as a good market.
This idea has the largest problem in the set and scores fourth. The reason is that the constraints — data acquisition, operator incentives, capital — are all outside the product. No amount of good design, clever engineering, or founder hustle fixes an incentive misalignment with the people who control the supply.
The diagnostic that would have caught this early:
Who has to cooperate for this to work, and do they want it to?
- Car Care: garages cooperate because they gain customers. Aligned.
- Kombi: operators must cooperate but lose pricing power and privacy. Misaligned.
Ask that question before falling in love with the size of a problem.