Byron Sharp's finding, formalised by Dale Harrison, that roughly 5% of category buyers are actively in-market at any given moment is not a rule of thumb. It is a mathematical consequence of penetration and purchase frequency. Enter three numbers for your category and see the calculation run.
National statistics offices (free, always the primary source):
For a specific segment (e.g., "urban households with kids under 12"), apply demographic filters at the same source.
B2B Business registers (free):
Sales-tool account counts (paid, higher fidelity):
For enterprise-only segments (Fortune 500, DAX 40): use published lists directly. This is often only a few hundred accounts — a very different game from B2C.
Best sources (paid — the gold standard):
Free proxies:
If you have no external data: use your own POS data — your brand's buying households ÷ TAM = your penetration. Divide by your market share estimate to back into category penetration. Flag as estimate.
B2B Category penetration is rarely available syndicated. Best sources: (1) your CRM — customer accounts ÷ addressable accounts at the category level (you + top competitors combined); (2) Gartner / Forrester / IDC adoption studies; (3) LinkedIn B2B Institute research on horizontal categories.
Same panel providers as penetration (GfK, Kantar, NielsenIQ, Circana). Ask for "average annual purchase frequency among category buyers."
Free proxy: your own repeat-buyer data. If 40% of your buyers return within 6 weeks, W ≈ 8–9 in that segment.
Category benchmarks (rule of thumb):
If W is below ~2, switch to B2B mode — long-cycle categories are more honestly modelled with contract-length and evaluation-window inputs than with fractional annual frequencies.
Your own commercial data: average contract length or average purchase interval per account. If you sell 3-year enterprise SaaS deals, cycle = 36 months. If you sell annual renewals, cycle = 12 months.
Industry benchmarks:
Sources: Gartner Peer Insights, Capterra, G2 buyer reports; your own CRM close-won data; industry-specific analyst studies.
The window during which an account is actively in-market: from first serious inbound/outbound touch through to decision. Not the whole cycle — just the active-consideration period.
Your own commercial data: your sales-cycle length from opportunity created to closed-won/lost. This is the strongest signal you have.
Industry benchmarks:
Reference: Gartner's B2B Buyer Journey research; John Dawes (Ehrenberg-Bass) for the LinkedIn B2B Institute has published cycle-length data for several horizontal B2B categories. The eval window ÷ cycle length ratio is what determines the in-market share.
The 95/5 rule is derived from the NBD-Dirichlet model, which assumes a stationary category — one where buyer population and purchase frequency are stable.
Stationary — penetration has moved ±2pp over the last 3 years, purchase frequency is stable, no cohort of first-time buyers entering.
Growing — penetration is climbing >3pp/year OR a significant new-buyer cohort is entering the category for the first time. In a growing category, the in-market pool can reach 25–35% (Harrison), because first-time buyers are all simultaneously "in market" forming their initial preferences.
How to check for your category:
B2B Growing = a new category (early adoption phase, LLM tooling in 2023–25 for example) OR a category with regulatory disruption forcing simultaneous evaluations across the base (GDPR triggered this for privacy tech in 2018).