01 · Problem statement
Northline Mechanical believes it has a lead-generation problem. Over twelve months, total enquiries rose while closed revenue fell. Three providers have proposed three different solutions: a website rebuild, an increase in paid media, and a CRM replacement.
The decision question: which single intervention should be funded first, and what evidence would show within 30 days whether it was the right one?
02 · Symptom versus constraint
- Reported symptom
- “We are not getting enough leads.”
- Measured reality
- Enquiry volume is up 12%. Enquiries meeting the company's own stated fit criteria are down 31%.
- Restated constraint
- The business cannot distinguish a qualified enquiry from an unqualified one at the point of intake, so it cannot tell which of its inputs is working.
03 · Hidden assumptions
- That “lead” means the same thing to marketing and to the estimating team. It does not.
- That the 2023 offer definition still matches what the market is buying in 2026.
- That response time is stable because nobody has reported it changing.
04 · Alternative explanations from the 216-state discovery
Six of the surviving cells, with the evidence that killed or supported each.
| Cell | Hypothesis | Status | Evidence |
|---|---|---|---|
| Why × Direct × All | Ad spend is too low across every channel. | Killed | Spend rose 40% year over year while qualified volume fell. More of the same input produced less output. |
| Why × Opposite × Many | Volume is fine; qualification is what changed. | Supported | Raw enquiries rose 12%. Enquiries meeting the stated fit criteria fell 31%. |
| Who × Other × One | A single unnamed actor sets the real conversion ceiling. | Plausible | One estimator handles every quote over $20k. Median quote turnaround is nine days. |
| What × Keep Same × All | The offer never changed while the market did. | Supported | Package definitions and price points are unchanged since 2023; two competitors repositioned in 2025. |
| When × Less Same × Partial | Follow-up decayed on part of the pipeline. | Plausible | First-response time on web enquiries moved from 4 hours to 31 hours after a staffing change. |
| How × Other × None | No mechanism exists to measure lead quality at all. | Supported | No field in the CRM records fit against the stated criteria. Quality is asserted, never recorded. |
05 · Intervention comparison
| Option | Cost | Reversibility | Time to evidence | Verdict |
|---|---|---|---|---|
| Rebuild the website | High | Low | 90+ days | Deferred — the site is not where qualification is failing. |
| Increase ad spend | Medium | High | 30 days | Rejected — amplifies an unqualified input. |
| Reposition the offer | Medium | Medium | 60–90 days | Second move — needs the measurement below to target it. |
| Instrument lead quality | Low | High | 14–21 days | Recommended first move. |
| Restructure the estimating queue | Low | High | 30 days | Parallel — cheap, independent, and testable. |
06 · Smallest sufficient intervention
Instrument lead quality at intake before funding anything that claims to improve it.
It costs the least, reverses freely, and is the only option that makes every other option measurable. Two of the three proposals on the table may still be correct — but neither can be evaluated until this exists, and funding them first means paying for an answer you cannot read.
07 · 30-day evidence plan
Add a fit-criteria field to every inbound record, scored at intake by the person who takes it.
Record first-response time automatically rather than by self-report.
Segment the last 90 days retrospectively using the same criteria to establish a baseline.
Review at day 21 with a single question: did qualified volume move, or only total volume?
08 · Stop conditions and review date
- If qualified volume is flat at day 21 but total volume rose, the constraint is upstream in positioning — proceed to the offer work, not the website.
- If qualified volume rose and close rate did not, the constraint has moved to sales response — stop marketing work entirely.
- If the estimating queue change alone recovers close rate, no repositioning is funded this quarter.
Review date: day 21. Decision to fund the second move is not taken before it.
09 · Routing recommendation and disclosure
The recommended first move is internal — no provider is required to instrument intake. If the day-21 evidence supports repositioning, the route would be a brand and offer specialist; if it supports sales response, a growth-execution provider.
Variety Portal evaluates multiple paths. When we recommend a company affiliated with Variety Portal, we disclose that relationship. When an outside specialist is a better fit, we can recommend or help source one. Referral or commercial relationships are disclosed before engagement.