Choose an expensive problem

A problem becomes commercially interesting when a costly event meets a reachable buyer, a reason to act, and a solution you can test safely.

ReconImagine an operations manager says, “Month-end is painful.” That sentence is a lead, not evidence. struct the last month-end and the shape may become clear: twelve hours of reconciliation, two people pulled from other work, an invoice delayed, and a finance lead who owns the result. Now there is an event, a cost, a buyer, and a path to verification.

An expensive problem need not involve the largest rupee amount. It is expensive when the customer feels enough consequence—money, time, delay, error, risk, or lost opportunity—to change behaviour and allocate resources.

Reconstruct the event before naming the market

Begin with the last real occurrence. Ask the customer to walk through the trigger, steps, people, systems, hand-offs, workarounds, failure, and aftermath. Ask what they tried, what it cost, and why the problem remains.

Past behaviour is stronger than an imagined future. “Would you use this?” invites politeness. “What happened the last three times?” invites evidence. Where permission allows, inspect sanitized artifacts: a blank form, an anonymized report, a workflow map, an error log, or the calendar where the recurring fire drill appears. Build the cost of inaction from separate, supportable parts:

Cost of inaction = labour and rework + delay + failure + risk exposure + forgone contribution − overlaps

Do not turn every inconvenience into a dramatic financial claim. Employee hours count only at an agreed loaded cost and only if those hours have an alternative use. Delayed revenue differs from lost revenue. Risk exposure is not an expected loss until probability and impact are responsibly estimated. Keep facts, customer estimates, and your assumptions in separate columns.

WORKED ARITHMETIC · ASSUMPTIONS SHOWN Worked example · the cost floor, not a sales promise

Assume a verified process consumes 12 hours each month at an agreed loaded cost of ₹1,200 per hour. The visible labour cost is ₹14,400 per month, or ₹1,72,800 for twelve occurrences. If the customer also claims a delay cost, verify that it is separate before adding it. This arithmetic establishes one part of the current burden. It does not prove your solution will remove all of it.

Pass the five commercial gates

Use scores to compare candidates, but use gates to prevent an attractive total from hiding a fatal weakness.

After the gates, rank survivors on frequency, severity, ability to pay, founder fit, proof speed, retention potential, and adjacent expansion. Keep the components visible. Multiplying arbitrary 1–5 scores can create impressive precision from opinions.

Find the entire buying system

The user and buyer may be different people. For a business purchase, map at least:

• • the user who experiences the workflow;
• • the process owner accountable for the result;
• • the economic buyer who can release money;
• • technical, security, legal, or procurement gatekeepers;
• • the person whose workload, authority, or status may be threatened;
• • the alternatives: current process, another vendor, internal build, delay, and doing nothing.

Budget is sometimes pre-allocated. Sometimes a strong business case creates it. What matters is whether the consequence is owned, the organization can pay, and the decision path is real. A friendly champion without access, authority, or a personal reason to act is useful context—not yet a deal.

Prefer a beachhead you can learn faster than others

“Small businesses” is too broad to teach you anything repeatable. A useful beachhead shares a workflow, buyer, vocabulary, trigger, and delivery environment. Narrowing increases the chance that one conversation improves the next, one result becomes relevant proof, and one introduction reaches a similar buyer.

The beachhead must still be large and reachable enough to support the intended business. Count plausible accounts from an explicit list or defensible source. Then test access. Thirty names in a spreadsheet are more useful than a giant market estimate if you can contact the thirty this month.

Look for operational signals:

• • repeated manual work or unofficial tools;
• • recurring rework, refunds, penalties, escalations, or missed deadlines;
• • budget already spent on staff, vendors, insurance, or workarounds;
• • a dated compliance, contract, capacity, or customer trigger;
• • a buyer who will share evidence, define success, or make an introduction;
• • a willingness to sign a bounded paid proof.

Compliments, survey clicks, and feature requests are weaker. A paid pilot proves one purchase under one set of conditions. Renewal begins to prove repeated value. Profitable repeatability in a defined segment is stronger still.

Reject seductive problem choices

Several traps recur:

Novelty without consequence. Interesting technology is looking for a reason to exist.

Large market without access. The theoretical spend is enormous; you cannot reach the buyer.

User love without payer value. Adoption may grow while the economic buyer sees no return.

Pain without priority. The cost is real, but three other problems outrank it.

Urgency created by your pitch. Pressure produces a signature and then regret, delay, or churn. A proof that carries unacceptable risk. Sensitive data, automated high-stakes decisions, regulated activity, or safety-critical delivery may require controls and professionals before any experiment.

· EF pp. 13–18, 30–33 · SALES pp. 12–13, 20–21 · TABLE pp. 11–20 · GE chs. 8–10. Mechanisms paraphrased; the arithmetic is an illustrative cost floor, not a value claim.