Part 2 of 15 · Before you start writing
Part 2: The Research That Makes a Plan Credible
Analysts can tell within two paragraphs whether you did primary research or paraphrased a Google result. This part shows you how to validate demand, size a market you can actually reach, analyse competitors properly, and cite South African data sources that carry weight in a credit committee.
The gap between an amateur plan and a professional one is almost entirely a research gap. The writing is the easy part. What follows is the fieldwork that produces sentences a stranger will believe.
Chapter 9Validating the idea before you fund it
Validation means obtaining evidence that people will pay, at your price, at a volume you can serve. Opinions are not evidence. In descending order of strength, here is what actually counts:
| Rank | Evidence type | Weight with funders |
|---|---|---|
| 1 | Historical revenue from paying customers, visible in bank statements | Decisive |
| 2 | Signed contracts, purchase orders or offtake agreements | Very strong — often unlocks contract finance on its own |
| 3 | Letters of intent or conditional supply commitments on client letterhead | Strong, but discounted — assume 30–50% conversion |
| 4 | Paid pilot or pre-orders, even small ones | Strong for start-ups; proves willingness to pay |
| 5 | Waiting lists, deposits, or a paid pre-launch cohort | Moderate |
| 6 | Structured customer interviews (30+ target buyers) | Moderate — must include price reaction, not just interest |
| 7 | Survey results | Weak unless the sample and method are disclosed |
| 8 | “Everyone I’ve spoken to loves the idea” | Zero — actively harmful |
Chapter 10Sizing a market you can actually reach
The most common market-analysis error in South African plans is quoting a national market figure and implying a share of it. Analysts have seen it thousands of times and discount it entirely. Use the TAM–SAM–SOM structure and build the bottom-up version, because that is the one that will be interrogated.
Market sizing in four steps
- 1TAM (Total Addressable Market). Everyone who could theoretically buy this category nationally. Cited from StatsSA, an industry body or a paid report. Used for context only.
- 2SAM (Serviceable Addressable Market). The portion reachable given your geography, licensing, language, price point and channel. A bakery in Mthatha does not serve the national bread market.
- 3SOM (Serviceable Obtainable Market). What you can realistically win in three years given capacity and marketing spend. This is the number that must reconcile to your revenue forecast.
- 4Bottom-up reconciliation. Rebuild the same revenue from units: customers × frequency × average basket. If the two methods differ by more than 25%, your forecast is wrong — find out which one.
Bottom-up sizing — worked example (fast-casual restaurant, Gauteng)
Capacity approach Seats 48 Covers per seat per day (turns) 2.4 Trading days per month 27 Average spend per cover R 148 ------------------------------------------ Theoretical monthly revenue R 460,339 Realistic utilisation (yr 1) 62% ==> Forecast monthly revenue R 285,410 Catchment approach (cross-check) Households within 4km 18,400 % in target LSM band 34% ==> Target households 6,256 % that visit at least monthly 21% Visits per month 1.6 Average spend per visit R 148 ==> Forecast monthly revenue R 311,088 Variance between methods: 8.9% -> ACCEPTABLE Adopt the lower figure and explain the cross-check.
Chapter 11Competitor analysis that is not a list
A competitor section that names three companies and says “we will offer better service at lower prices” is worse than no competitor section, because it proves you have not thought about it. Better service at lower prices is not a strategy; it is a description of a business that will run out of money.
Use a structured comparison, and be honest about where competitors beat you. Analysts trust plans that concede ground.
| Factor | You | National franchise | Local incumbent | Informal operators |
|---|---|---|---|---|
| Typical monthly price (500m²) | R14,500 | R19,800 | R15,200 | R8,000 |
| B-BBEE level | Level 1 | Level 4 | Level 2 | Unverified |
| Compliance (UIF, COIDA, PPE) | Full | Full | Partial | None |
| Response time on escalation | 4 hours | 48 hours | 24 hours | Ad hoc |
| Contract flexibility | Month-to-month | 24-month lock-in | 12-month | None |
| Their advantage over you | — | Balance sheet, national coverage, insurance cover | 15-year relationships, known to procurement | Price — 45% below you |
| Your counter | — | Level 1 procurement points + faster escalation | Target their weakest accounts on compliance risk | Compete on liability transfer, not price |
Chapter 12South African data sources that carry weight
Citing credible, checkable, local sources is one of the cheapest credibility gains available. Here is where the usable data actually lives.
| Source | What you get | Best used for |
|---|---|---|
| Statistics South Africa (statssa.gov.za) | Census and Community Survey data, QLFS employment, CPI/PPI, retail and manufacturing sales, building plans passed, GHS household data | Catchment demographics, income bands, sector growth rates, inflation assumptions |
| SARB (resbank.co.za) | Repo and prime rates, credit extension, exchange rates, Quarterly Bulletin, sector credit data | Interest rate assumptions, financing cost, macro outlook |
| National Treasury | Budget Review, MTBPS, infrastructure pipeline, municipal finance data (via the Treasury local government database) | Public spend pipeline for construction, logistics and supplier businesses |
| the dtic (thedtic.gov.za) | Sector master plans, incentive programme guidelines, industrial policy action plans | Incentive eligibility, sector policy alignment for grant applications |
| IDC (idc.co.za) | Sector research reports, economic trends, sectoral value-chain studies | Industry analysis with a citation a DFI already respects |
| CIPC (cipc.co.za) | Company registration, directorship, annual return status | Competitor verification, own compliance evidence, director history |
| SARS | Tax compliance status pin, VAT registration, trade statistics | Compliance pack; import/export volumes by tariff code |
| CSIR / HSRC / DSTI | Applied research, technology and innovation studies | Technical validation for manufacturing and innovation plans |
| Municipal IDPs and SDFs | Integrated Development Plans, spatial frameworks, capital budgets | Local demand, zoning, planned infrastructure — hugely under-used |
| Industry bodies | SAPOA (property), NAAMSA (auto), SAPIA (fuel), BUSA, Agbiz, SAAFF (freight), SEIFSA (metals) | Sector-specific volumes, pricing and outlook |
Chapter 13Using AI for research without wrecking your credibility
AI tools are genuinely transformative for the research phase — and genuinely dangerous if you let them generate facts. The rule is simple: use AI to structure, challenge and draft; use primary sources for every number that enters the plan.
| Use AI for | Never use AI for |
|---|---|
| Building interview guides and survey instruments | Generating market size figures or statistics |
| Structuring the competitor matrix and identifying factors you missed | Naming competitors and their pricing without verification |
| Stress-testing your assumptions (“attack this forecast as a credit analyst”) | Producing the financial model without you understanding every formula |
| Rewriting waffle into plain, specific language | Citing regulations, incentive limits or programme criteria |
| Generating the risk register long-list, which you then edit down | Writing the executive summary from nothing — it will be generic |
| Summarising long PDFs (IDPs, sector master plans, annual reports) | Anything you cannot personally defend in a meeting |
Chapter 14The research pack you should have before writing
Research completeness checklist
- Catchment or customer-base data from StatsSA, with the specific table referenced
- At least 15 structured buyer interviews, with price reaction recorded
- A competitor matrix with verified pricing (obtained as a mystery shopper, not estimated)
- Three supplier quotes for every major capital item, on letterhead, dated
- A written cost breakdown of one unit of your product or service, including owner time
- Municipal IDP extract or sector master plan reference relevant to your demand
- Your regulatory and licensing list, with current status and cost of each
- Twelve months of bank statements and management accounts, if trading
- Signed LOIs, contracts or pre-orders, scanned and dated
- The current prime rate and your assumed lending margin, sourced and dated
The plan is a report on the research. If the research is thin, no amount of writing will save it — and if the research is strong, the writing almost does itself.
The operating principle of this guide
