Methodology

A formal description of the valuation, buyer-archetype, confidence, and platform-fit models that produce Will I Sell IT? outputs. Every weight, threshold, and benchmark is documented and tunable. Calibrated against 2026 indie-business sale comparables across multiple data sources.

version: v0.6  ·  revised: 2026-05  ·  license: CC BY 4.0 for the methodology document; algorithm available under MIT for non-commercial integration (see API)

1. Pipeline overview

Inputs are a structured answer set a from a 12–17 question form (revenue path) or 11 question form (pre-revenue path). The pipeline produces a single immutable result object containing the verdict, valuation range, archetype, confidence, fix-list, platform recommendations, and a sectioned narrative.

Resolution order matters: archetype is determined first (from answers and range), then platforms are biased by archetype, then firstMove may reference the top platform. All downstream sections read from the unified synthesis object.

2. Valuation model — revenue path

The revenue-path valuation is a multiplicative composition of an asset-type baseline and 14 factor weights, applied to monthly revenue (annualized for non-recurring asset-revenue mismatches).

Annual-basis divisor

When the user selects a recurring-style asset (SaaS, AI, API) but reports one-time, ad-based, or mixed revenue, the asset multiplier (calibrated for monthly recurring) overshoots when applied directly to monthly-equivalent of annualized non-recurring revenue. The divisor of 2.2 brings the math back into the correct annualized range.

Score clamp

Compounded multiplication of 14 factors can produce scores in the range 0.05–8.0 for extreme answer combinations. We clamp to [0.30, 3.00] to keep deal estimates within plausible market bounds. The unclamped rawScore is preserved in the output for diagnostics.

3. Valuation model — pre-revenue path

Pre-revenue assets cannot be valued on multiplicative ARR-based logic. Instead, we use an additive base + user-value model with demand-signal bonuses.

Range halving (×0.6 / ×1.4) calibrated to the 90% confidence interval observed across approximately 120 indie pre-revenue sales 2024–2026.

4. Asset-type multipliers

Asset-type multipliers represent the months of MRR each business type fetches in 2026, before factor adjustments. Bounds reflect typical low/high range observed in arms-length transactions.

5. Factor weights

Each factor is a multiplier applied to the score. Weights are calibrated against actual closed-deal data and reviewed quarterly. Default value of 1.0 means no adjustment; values below 1.0 represent discounts; values above 1.0 represent premiums.

Revenue trend (last 90 days)

Monthly churn

Customer concentration

Handover readiness

Target audience

Business stage

Revenue model

Margin profile

Customer count

Team size

Legal entity

Transferable assets

Urgency / timeline

Traffic-source diversity

Industry weight

Industry weights tuned to 2026 acquirer signals from public M&A coverage and reviewed quarterly

6. Buyer-archetype classification

Cases are classified into one of six buyer archetypes via a deterministic decision tree operating on valuation range, stage, audience, revenue type, churn, margin, transferable assets, and motivation. The classifier returns an archetype identifier, an array of determinants (the specific signals that drove the pick, each with a weight), and a strength score reflecting confidence in the classification.

Archetype catalogue

Decision rules (in priority order)

1. Pre-revenue path → Exploring Entrepreneur. All cases on the pre-revenue path default to this archetype, regardless of range. Determinants: isPrerev, optionally augmented by demand and codeQuality signals.
2. High ≥ $1M, mature stage, registered entity → Institutional Investor. Override to Strategic Acquirer if motivation is cash-out / opportunistic AND IP / contracts are transferring (signals strategic-acquisition framing rather than pure financial deal).
3. High ≥ $250K AND (enterprise audience OR substantial+ user base) → Strategic Acquirer. Strengthened if motivation is pivoting AND audience-asset is transferring (IP, list, socials).
4. High ≥ $50K AND clean retention (lt5/m510) AND established/mature stage AND recurring/usage revenue AND healthy margin → Portfolio Acquirer.
5. B2B audience AND recurring/usage revenue AND high ≥ $30K → Portfolio Acquirer. Lower-bar fallback for clear B2B SaaS portfolio targets.
6. High < $25K → Independent Operator. Strengthened if solo team with ≤ 2 transferable assets (clean indie deal).
7. Otherwise → First-Time Acquirer. Mid-range cases without strategic or portfolio markers.

Determinants

Every classification produces an array of determinants documenting the signals that drove the pick. Each determinant carries a key, label, value, weight (0.0–1.0 relative importance), and a full-sentence note explaining the signal. Determinants are sorted by weight and exposed in the result for transparency.

7. Confidence model

Confidence is computed as a weighted sum of three components, scaled to a 0.0–1.0 score with a categorical label.

Required fields differ by path: 12 fields for revenue path, 10 for pre-revenue path. Flag count comes from the combo-flag detector (Section 8). Archetype strength is set by the classifier based on how cleanly the case fits the matched archetype's defining rules.

8. Combo-flag detection

A separate pass detects logically incoherent answer combinations and applies penalty multipliers to the score. Each flag carries a key, human-readable label, penalty (multiplicative), and a reason explaining the contradiction.

9. Verdict thresholds

The verdict is derived from the clamped score (revenue path) or from the high estimate (pre-revenue path). Five tiers map to three colors (go / caution / stop).

Revenue path

Pre-revenue path

10. Marketplace-fit scoring

Marketplace recommendations are produced by a three-layer scoring pipeline operating across a pool of 18 marketplaces. The top-scoring platform is surfaced as the primary recommendation; the next three appear as alternates for transparency.

Layer 1: per-platform fit rules

Each marketplace has 5–7 hardcoded rules that award (or deduct) points based on case answers. Rules cover hard disqualifications (e.g. premium brokers reject pre-revenue), range fit, asset-type alignment, audience match, urgency compatibility, and operational cleanliness.

Layer 2: archetype bias

A small adjustment (5–15 points) layered on top of fit-rule scores, based on the matched buyer archetype. Bias is intentionally modest — it acts as a tiebreaker between platforms with similar fit scores rather than overriding strong fit signals.

Layer 3: sort and select

All 18 platforms are scored, sorted descending by total, and the top result is returned as the primary recommendation. The next three appear as alternates. The marketplace pool is curated to cover the full range of indie-business sale venues from auction formats through premium brokers.

11. Limitations & reproducibility

Known limitations

• Pre-revenue model under-weights AI proprietary fine-tunes. Roadmap: add a moat-classification question Q3 2026.
• Geographic distribution not modeled. US-focused vs global-default has approximately ±20% delta in observed sale prices.
• Customer NPS / community engagement signals absent. Qualitative buyer-side signals not yet captured.
• Platform fees / commissions not factored into "what you'd net". Output represents gross sale price.
• Industry weights require quarterly review. Categorical premiums shift with macro conditions.

Reproducibility

The valuation, archetype, confidence, and platform-fit logic is deterministic. Given an identical answer set, the pipeline always produces an identical synthesis object. This makes the tool suitable for benchmarking, regression testing, and integration into other valuation workflows. See the API documentation for integration details.

A self-test fixture suite is included in the codebase, covering 15 representative cases across all six archetypes plus edge cases (incomplete inputs, all-flags-trigger, ecom-solo-no-false-flag, cascade-mode viewport, mutation safety, etc.). The test runner is exposed via window.__willisellitSelfTest() in the browser console.

12. References

• Acquire.com — public sold-listings transaction data, 2024–2026 (~480 deals)
• Empire Flippers — quarterly transaction reports including aggregate multiples by category
• Microns sale archive — small-SaaS-focused transactions, ~120 deals
• FE International — 2026 multiples summary report
• SaaS Capital — quarterly indie-SaaS valuation benchmarks
• Indie Hackers exit threads — qualitative signals on concentration risk and founder dependency
• TechCrunch / The Information M&A coverage — public deal disclosures for industry-weight calibration

If you have transaction data that disagrees with these multipliers, contact the editorial team. Methodology is reviewed and revised quarterly.

Last revised 2026-05 · v0.6 · willisellit.com is an independent valuation tool. No affiliation with any specific marketplace is implied or claimed.