Pain Discovery

#concept #sales

Summary

The offer is found by interrogating the market, not by introspection. Sources across all three traditions describe the same loop — find the bottleneck, name it back as the offer — while the distillation adds a qualification filter that decides whether a pain is worth selling to at all, and oskar-hartmann adds the acute-pain test (revealed preference at its purest).

Current Understanding

The loop: talk to prospects → find the bottleneck → name it back to them as the offer (2026-07-17-design-the-perfect-offer). Worked example: "business owners aren't using AI → they'd want workflow automation → the offer is a calendar audit that identifies 3 workflows and buys back 10 hrs/week." The prescribed question is "what's your biggest bottleneck around [your domain]?", asked of 5–10 prospects in the niche.

Watch spend, not talk — origin: 2026-06-15-rodenko-selling-development-expensively (the Zendesk motion). What people say they want and what they already pay for are different data. Find an operational cost line, then:

"How much do you spend on X?" → "And if you spent 10× less — would we talk?"

The insurance built into the 10× ask: even if you're off by half, you still delivered 5×. This is a strictly better opener than the bottleneck question, because a budget line is evidence and a stated bottleneck is an opinion. Rodenko maps the biggest cost line by company type: service firms → payroll; manufacturing → materials; narrow industries → sector-specific SaaS.

"Pain with money" — the four-part filter. A pain qualifies only if all four hold:

It recurs often.
It costs the business money.
It's already being solved with duct-tape workarounds.
The founder understands its value.

Missing any → not a client. Origin: the B2B pipeline in 2026-06-15-making-money-with-ai-2026 (the distillation inherited it). Criterion 3 is the subtle one: an existing workaround is proof of both budget and felt pain, whereas an unsolved problem may simply not matter enough. Criterion 4 is why niche-selection says sell to leadership — the same pain fails this filter when pitched at the wrong seat. That same source adds the discovery motion around it: find many companies → DM 3–4 questions about manual tasks (~10% reply) → validate the pain has money before building.

The acute-pain test — revealed preference at its purest (2026-07-26-main-principle-of-successful-business, oskar-hartmann, added 2026-07-26): "If the business only works when everything is perfect — it's a bad business. A good business is when everything is bad and people still come and pay." His exemplar: hospitals — 1.5-hour waits, zero good reviews, you pay anyway. When pain is that acute, people pay for the concept, before polish exists — which is what makes sell-before-build's pre-payment tests possible at all. This extends the "watch spend, not talk" line to its endpoint: the strongest qualification isn't what they spend on, it's what they keep paying for despite bad service. His targeting question sharpens the filter further: who — by name — is the smallest group with the most acute pain? (A SOM statement; see tam-sam-som.)

Your old clients are the warmest pain data. Call your 3–5 best and ask "what result were you paying me to get?" — pointedly not "why did you choose us?", which only returns "good team" (2026-06-15-rodenko-selling-development-expensively; echoed in 2026-06-15-konspekt-aphorisms: "the money is with your old clients"). The first question recovers the outcome you were actually bought for, which is the raw material for outcome-based-selling. The vault's most immediately actionable item.

AI as a free diagnostic — the 30-day giveaway sorts prospects by their reaction; see ai-market-shift.

Naming the pain in the buyer's own words (content-side). A distinct application of the same skill: 2026-07-18-information-is-free-implementation-is-paid mass-produces nuanced, observable pains with an AI prompt — "Make a list of 10 nuanced but observable problems business owners between $500K and $2M in revenue have around [hot button]." The modifiers do the work: "nuanced but observable" forces specifics over clichés, and the revenue band anchors to the ICP, so the output "describes the viewer's world better than they can describe it themselves." This is pain articulation for marketing (open a video on the pain, then teach the fix), not live qualification — but it's the same core move as the bottleneck question and the old-client question: surface the pain, name it back. See information-vs-implementation.

Evidence

"Talk to prospects, find the bottleneck, name it back to them as the offer" + the calendar-audit worked example — 2026-07-17-design-the-perfect-offer
"Ask 5–10 prospects: what's your biggest bottleneck around [your domain]?" — checklist item 3, 2026-07-17-design-the-perfect-offer
Primary (spend/10× opener, old-client script, cost-line-by-company-type): 2026-06-15-rodenko-selling-development-expensively
Primary (four-part "pain with money" filter + the DM-many-companies motion): 2026-06-15-making-money-with-ai-2026
"The money is with your old clients" — 2026-06-15-konspekt-aphorisms
The nuanced-observable-pain AI prompt (content-side pain articulation) — 2026-07-18-information-is-free-implementation-is-paid
The acute-pain test (hospitals; "pays even when everything is bad"); smallest-group-by-name targeting — 2026-07-26-main-principle-of-successful-business (oskar-hartmann)
Condensed restatement of all of the above — 2026-06-15-selling-development-services-in-the-ai-era

niche-selection — choose the audience before interrogating it
outcome-based-selling — the discovered pain, inverted, becomes the outcome
productized-service — repeated identical pains are the productization trigger
ai-market-shift — the AI giveaway as a qualification instrument
information-vs-implementation — the same pain-naming skill, pointed at content instead of a sales call
sell-before-build — downstream: the paid experiment that confirms what discovery suggested
overview

Contradictions / Uncertainty

Stated vs. revealed preference. The video's method (ask about bottlenecks) is exactly the kind of self-report the distillation warns against ("watch what they spend on, not what they say"). Not a flat contradiction — the video's example does end at a budget-relevant outcome — but the distillation's method is the more rigorous of the two. Status: tentative on the bottleneck question as a standalone tool.
The 10× framing assumes the cost line is compressible by an order of magnitude. Neither source discusses what to do when it isn't.

Next Questions

Which operational cost lines are actually 10× compressible with AI, and which just look it?
How do you run the "what result were you paying me for?" call without seeding the answer you want?
What replaces old-client mining for someone with no past clients?

7.1 KiB Raw Permalink Blame History Unescape Escape