Every factor below is a slider, because a market study whose assumptions can't be challenged is a press release. Move anything: the funnel, both SOM builds, the binding constraint, the convergence check, and the sensitivity ranking recompute live — and the reading rewrites itself.
| Project ceiling: FTE × hrs × rate | |
| Retainer base: clients × fee × 12 | |
| Capacity SOM |
| New projects: convos × 4 × close | |
| Bookings: projects × ACV | |
| + retainer base (45% conversion capped) | |
| Demand SOM |
Each driver shocked ±20%, holding the rest constant. Ranked by swing in SOM — spend diligence hours top-down.
A client version replaces every slider default with a figure traceable to one of these.
| Factor | Source | What it gives you |
|---|---|---|
| TAM | Fortune Business Insights · The Business Research Company · MarkWide · IDC Worldwide AI Spending Guide | Triangulate 3–4 estimates; use the definition closest to "implementation & deployment services," not total AI spend. |
| Geography split | Same reports, regional chapters; Eurostat digital-intensity stats | NA ≈ 38% of demand; EU adoption lag quantified rather than assumed. |
| Mid-market share | US Census SUSB + Eurostat SBS (firm counts by employee band) × gen-AI adoption rates (McKinsey State of AI, BCG AI Radar) | Build the segment factor from firm counts × adoption × services attach — a denominator you can defend. |
| Claude ecosystem share | Menlo Ventures State of Gen-AI in the Enterprise · Anthropic Economic Index · cloud-marketplace rankings | Enterprise LLM workload share by provider; re-base every two quarters — fastest-moving factor in the stack. |
| Blended rate | Upwork/Toptal published agency rates · SPI Professional Services Maturity Benchmark | Rate and utilization norms for boutique consultancies (SPI median utilization ~70%). |
| Close rate & ACV | Own CRM after 2 quarters; until then, SPI benchmark win rates for firms <10 FTE | The only honest source for these is your own funnel — flag them as lowest-confidence until measured. |
| Competitive density | Anthropic partner directory · Clutch/G2 agency listings | Count credible NA+EU Claude-focused boutiques; sanity-check whether your SOM implies outcompeting all of them. |
Method: SOM = min(capacity, demand), never a percentage of SAM. Convergence between the two builds is the credibility test; a gap > 25% means one leg is fantasy. All defaults illustrative. © Sergei Mochtchenkov, CFA — Mochtchenkov & Co.