Thirty pages of AI research and still no decision: researching and reading numbers the smart way

Ask a language model to evaluate an investment opportunity in Zambia, and it hands you thirty pages of beautifully structured text. It cites international development reports, discusses sub-Saharan macroeconomic trends, and presents attractive financial projections. You reach the final page with an exhausted brain and zero clarity on what commercial decision to take on Monday morning.

Generic AI research generates analytical fog. It blends outdated overseas commentary with hypothetical figures, hiding the practical operational variables that determine whether a Zambian business generates cash or goes broke. Zambian commercial operations turn on specific primary sources: Bank of Zambia policy announcements, Energy Regulation Board monthly fuel pricing reviews, road toll schedules, and Gazetted Statutory Instruments governing cross-border transit and tax obligations.

Commercial management requires clearing away decorative prose. You must force language models to extract verified regulatory data, stress-test underlying financial assumptions, and condense voluminous research into an actionable one-page decision memo.

The Kasumbalesa Expansion at Nkandu Freight

Nkandu Freight operates clearing and forwarding offices in Lusaka, Nakonde, and Chirundu. In early 2025, Managing Director Ba Nkandu considered opening a dedicated transit clearing yard at the Kasumbalesa border between Zambia and the Democratic Republic of Congo to capture growing copper concentrate and acid tanker traffic.

Nkandu asked an external business consultant to prepare an operational feasibility study. The consultant used a public language model to generate a 35-page market assessment. The document praised the commercial prospects of the regional trade corridor. It projected a 28% return on investment, recommending an immediate capital commitment of K2.8 million for land leasing, security fencing, heavy gravel paving, and office container deployment.

Nkandu examined the financial models behind the recommendations. The AI report based its projections on three fatal assumptions:

• Diesel fuel priced at K22.50 per litre, whereas the Energy Regulation Board had adjusted retail automotive diesel beyond K31.00 following international crude shifts and Kwacha movements.
• A truck border turnaround time of eighteen hours, ignoring the four-day vehicle queues common during Congolese customs system outages and clearing agent strikes.
• A flat Kwacha exchange rate against the US Dollar across a three-year operating period, disregarding seasonal currency volatility and Bank of Zambia reserve pressures.

Committing capital based on that report would have drained Nkandu Freight's working capital within five months. Nkandu scrapped the document and forced the consultant to rebuild the analysis using verified Zambian primary sources.

The Three Primary Sources of Zambian Commercial Reality

When using language models for market research or strategic planning, ground your queries in three statutory authorities.

1. The Bank of Zambia (BoZ). Monetary Policy Committee statements determine commercial borrowing costs and liquidity conditions. When evaluating capital investments requiring bank financing, pull the latest BoZ Monetary Policy announcement. If the policy rate rises or statutory reserve ratios tighten, local commercial bank lending rates increase. Any feasibility study assuming fixed 18% commercial borrowing rates during monetary tightening is commercial fantasy. Feed the actual policy announcement into the model to extract the prevailing commercial lending base rate.

1. The Energy Regulation Board (ERB). Fuel prices drive distribution and operating expenses across every Zambian province. The ERB adjusts retail pump prices monthly based on international oil prices and exchange rate movements. When modeling transport routes, delivery fleets, or generator running costs during load-shedding, input current ERB pricing decisions. Require your analysis to include a sensitivity model testing a K3.50 per litre increase in retail diesel.

1. Zambian Statutory Instruments and ZRA Customs Directives. Government regulations regarding transit permits, road tolls, carbon tax, and customs processing change through official Gazetted Statutory Instruments. An AI model trained on general internet sources will cite outdated regulations or regional templates from Kenya or Ghana. Download the specific Statutory Instrument from the National Assembly portal or Ministry website, upload the PDF directly into your workspace, and instruct the model to verify compliance requirements against that exact text.

The Decision-Grade Summary Protocol

Transform multi-page research documents into a concise executive decision framework using three operational rules.

Rule 1: Enforce comparative tabular structures. Ban narrative essays. Require the model to present investment options in side-by-side matrices comparing capital requirements, regulatory hurdles, and payback horizons.

Rule 2: Expose and isolate operational assumptions. Every commercial projection rests on assumptions regarding pricing, volume, turnaround times, and regulatory clearances. Require the model to state every assumption in an itemized table, citing the primary source for each figure.

Rule 3: Execute downside stress testing. Test every commercial proposal against severe local shocks: a 15% Kwacha depreciation against the US Dollar, a 10% ERB fuel price increase, and a forty-eight-hour customs border delay. Viable projects withstand these localized shocks; fragile projects collapse immediately.

Copy-Paste Prompt: Zambian Commercial Feasibility Sparring Partner

Act as a skeptical commercial director evaluating an investment proposal for a Zambian business.
Review the attached project proposal: [Insert Project Description or Text].

Execute a rigorous commercial evaluation under the following constraints:

1. Assumption Extraction:
List every operational and financial assumption in a table:
- Assumption Parameter (e.g., Fuel Price, Exchange Rate, Turnaround Days).
- Value Used in Proposal.
- Primary Zambian Source Benchmark (ERB, Bank of Zambia, ZRA, RTSA).
- Plausibility Check (Realistic / Understated / Highly Speculative).

2. Downside Sensitivity Stress-Test:
Recalculate project monthly net margins under three simultaneous stress events:
- Event A: 15% depreciation of the Zambian Kwacha against the US Dollar.
- Event B: An increase of K3.50 per litre in ERB automotive diesel pricing.
- Event C: A 72-hour operational delay at provincial borders or customs checkpoints.

3. Executive Decision Matrix:
Provide a 1-page summary table:
- Option Name
- Initial Capital Expenditure (ZMW)
- Monthly Fixed Overhead (ZMW)
- Breakeven Volume / Units
- Statutory Approvals Required (PACRA, ZEMA, Local Council, ZRA)
- Final Recommendation: PROCEED / REVISE ASSUMPTIONS / REJECT.

Format Rules: No conversational pleasantries. Output only structured tables and concise operational recommendations.

The Primary Source Verification Checklist