Analyze Lost Deal Reasons for Patterns

You are an analytics-minded sales leader for a digital marketing agency. You cluster messy reason text into themes without overreaching beyond what's written.
Agency: [AGENCY_NAME] — [SERVICES] | Period: [PERIOD] | Data:
[LOST_DEAL_REASONS]
(Each row: deal name, ACV, service requested, raw loss reason text or picklist value.)
Cluster the loss reasons into a clean theme list. For each theme, give: theme name, plain-English definition, count of deals, total ACV at risk, and 2 representative quoted reasons from the input. Rank themes by ACV at risk, not just count.

- Themes must be grounded in the actual reason text — quote at least 2 examples per theme verbatim.
- Don't merge themes that look similar but mean different things (e.g., "too expensive" vs "no budget this quarter").
- If more than 25% of deals have blank or single-word reasons, flag data quality at the top.
- No rep-bashing — frame as process/qualification themes.

Output a ranked table: Theme | Definition | # Deals | ACV at risk | 2 example quotes. Top of output: 1-line data-quality flag.

• [AGENCY_NAME] — Your agency name
• [SERVICES] — Service lines covered
• [PERIOD] — Date range of losses
• [LOST_DEAL_REASONS] — List of lost deals with raw loss reason text from your CRM

Agency: Northstar Digital — SEO, Paid Search | Period: Q1 2026
Losses:
BrewHaus | $72k ARR | PS | "chose AdVenture, said our reporting felt thinner"
MetroMattress | $108k ARR | PS | "went with AdVenture again — existing relationship"
Lumen Lighting | $32k proj | Web | "pushed launch to Q3, no budget now"
Cedar & Sage | $48k ARR | SEO | "need someone with more CPG experience"
Northwind | $60k ARR | SEO | "scope didn't match — they wanted content production too"
Kite | $54k ARR | PS | "price — 20% over their internal cap"

Data quality: clean — all reasons have free-text context.

Themes ranked by ACV at risk:
1. Competitive loss to AdVenture (PS) — $180k — "chose AdVenture, said our reporting felt thinner" / "went with AdVenture again — existing relationship."
2. Scope mismatch — $108k — "need someone with more CPG experience" / "scope didn't match — they wanted content production too."
3. Price ceiling exceeded — $54k — "price — 20% over their internal cap."
4. Timing/budget shift — $32k — "pushed launch to Q3, no budget now."

Top action implied: the AdVenture pattern is the biggest single ACV bucket — worth a dedicated competitive teardown and a reporting-experience upgrade before reframing pricing or scope.

• Feed in the raw notes field, not just the picklist — picklists hide the real story.
• Sort by ACV at risk, not deal count, so big-ticket themes don't get buried under small-deal noise.
• If "price" dominates, drill in — it's almost always a value-articulation problem, not a real price problem.