# A 3,940-product SaaS census finds flat platform pricing is more common than pure per-seat billing

PulseSignal’s September 2 snapshot of 3,940 monitored SaaS pricing pages finds flat/platform pricing is the largest company-level model among products with numeric paid prices, while pure per-user pricing remains 22% and the median cheapest qualifying paid plan remains $29 per month.

The useful finding is not that one pricing model has 'won.' It is that observable SaaS pricing is much more heterogeneous than the default per-seat narrative suggests—and that the answer changes sharply with methodology. PulseSignal’s living public-price census keeps moving, and its latest disclosed extraction audit improved to 90%, but the data should still be treated as directional packaging evidence rather than a universal market truth.

- Status: Active
- Published: 2026-08-30T17:01:50+12:00
- Updated: 2026-09-05T10:01:45+12:00
- Categories: SaaS, Indie Business, Pricing & Billing, Product Economics
- Tags: pricing models, PulseSignal, SaaS economics, SaaS pricing
- Canonical HTML: https://beyondthe.news/dossiers/pulsesignal-saas-pricing-census-flat-platform-per-user-entry-price-2026

## Latest material change

Refreshes PulseSignal's living census to the September 2 snapshot: Flat/platform 811, Per-user 418, Mixed 218, Other per-unit 204, Not labeled 119, Usage-based 111; USD entry-price sample 1,573; disclosed extraction audit improves from 80% to 90%. Broad conclusion and $29 median remain unchanged.

## What changed

PulseSignal’s living 2026 pricing census has refreshed again, now to a September 2 snapshot covering 3,940 monitored SaaS tools. The priced-company denominator remains 1,881, but several classifications moved: Flat / platform is now 811 companies (43%), Per-user 418 (22%), Mixed 218 (12%), Other per-unit 204 (11%), Not labeled 119 (6%) and Usage-based 111 (6%). The USD entry-price sample increased from 1,572 to 1,573 companies while the median cheapest paid plan remains $29 per month. PulseSignal’s latest disclosed weekly plan-level audit also improved from 53 of 66 readings accurate (80%) to 55 of 61 (90%). The broad conclusion remains unchanged, but the live census is materially more current and its disclosed extraction-quality signal has improved.

## Why it matters

Pricing architecture determines how revenue scales with customer growth, team size and product usage. Founders often inherit assumptions such as 'B2B SaaS is per-seat' or 'AI has made usage pricing dominant' without checking the market they actually compete in. PulseSignal's census gives builders a large observable benchmark for public pricing structure and entry points, while its category breakdown is more actionable than one global headline. The methodological caveat remains material: it measures published numeric pricing, not negotiated contracts. The improved 90% weekly extraction audit makes the latest snapshot more credible than the prior 80% audit, but it still does not make individual classifications or small percentage differences exact.

## Flat/platform remains the largest public-pricing classification

Among the 1,881 companies with a numeric paid plan in the September 2 snapshot, PulseSignal classifies 43% as Flat / platform and 22% as pure Per-user. The counts moved to 811 Flat / platform, 418 Per-user, 218 Mixed, 204 Other per-unit, 119 Not labeled and 111 Usage-based. A company is labeled Mixed when its extracted plans span multiple pricing families, so the 22% figure does not mean only 22% of SaaS products ever charge for seats; it means only 22% are classified as purely per-user at company level.

## Usage-based pricing still looks much smaller under this methodology

Only 111 companies, or 6% of the priced-company sample, land in PulseSignal’s pure Usage-based category. That differs sharply from procurement-oriented research such as Vertice, which reports consumption-based pricing as the largest contract model in its Q2 2026 dataset. The apparent contradiction is methodological: PulseSignal classifies the public architecture of a company’s numeric paid plans, while contract datasets can classify how purchased software spend is actually charged.

## The median public entry price is still $29

Across 1,573 companies with a qualifying USD cheapest paid plan, the median monthly-equivalent entry price remains $29. PulseSignal reports $39 for Flat / platform, $19 for Per-user, $19 for Mixed, $35 for Other per-unit, $49 for Not labeled and $25 for Usage-based. Entry price is not total cost: it does not normalize included seats, usage allowances, overages, feature gates or annual commitments.

## Category-level structure remains more useful than the global average

PulseSignal publishes category slices only where at least 15 priced companies are present. The September 2 table spans 35 qualifying categories and still shows substantial variation: project management is led by Per-user at 51%, SMB marketing automation by Flat / platform at 81%, product analytics by Flat / platform at 52%, and team messaging by Per-user at 75%. The practical lesson for founders is unchanged: benchmark pricing architecture against adjacent substitutes and buyer expectations rather than an all-SaaS average.

## The disclosed extraction audit improved materially

PulseSignal says its latest completed weekly plan-level audit, dated August 30, marked 55 of 61 readings accurate against live vendor pages, or 90%. The prior disclosed audit in BTN’s dossier was 53 of 66, or 80%. That is a meaningful improvement in the quality signal, but the audit still samples plan-level readings rather than validating every company-level pricing-model classification.

## Independent datasets still point in the same broad direction, not to the same exact number

Toolradar's separate 2026 dataset of roughly 2,575 paid tools finds explicit per-seat wording in only a minority of tools under its conservative text-match method, while Vertice’s procurement data emphasizes consumption-based contract structures. These datasets are not directly comparable, but together they reinforce the need to inspect methodology before converting one pricing share into a market rule.

## Key details

- PulseSignal’s September 2, 2026 snapshot monitors 3,940 SaaS tools.
- 1,881 companies with a numeric-priced paid plan are included in the pricing-model share calculation.
- Flat / platform: 811 companies (43%).
- Pure Per-user: 418 companies (22%).
- Mixed: 218 companies (12%).
- Other per-unit: 204 companies (11%).
- Not labeled: 119 companies (6%).
- Usage-based: 111 companies (6%).
- 1,573 companies contribute to the USD entry-price analysis; the median cheapest paid plan remains $29/month.
- Current model-level entry medians are $39 Flat/platform, $19 Per-user, $19 Mixed, $35 Other per-unit, $49 Not labeled and $25 Usage-based.
- PulseSignal’s latest disclosed weekly plan-level audit found 55 of 61 sampled readings accurate against live pricing pages (90%), up from the prior disclosed 80% audit.
- The census measures observable public numeric pricing and excludes quote-only/free-only companies from model shares.
- Separate datasets using different definitions produce very different pricing-model shares, so methodology materially affects the result.

## Builder takeaways

- Benchmark pricing against your product's category before choosing a meter; the global SaaS average hides major differences between CRM, project management, analytics, marketing and infrastructure tools.
- If you are considering per-seat pricing because it feels 'standard,' inspect whether direct competitors are actually pure per-user products or use flat, mixed or usage-linked structures.
- Treat the $29 median as a public entry-price reference, not a target. Recalculate competitor pricing for your expected customer size and usage profile before using it in packaging decisions.
- When comparing pricing research, read the denominator and classification method. Public price pages, signed contracts and tier-text pattern matching answer different questions and should not be blended as though they were one market share.
- Use PulseSignal's category rows and linked vendor records as leads, then verify current pricing directly on the competitors that matter to your product.
- Even with the improved 90% plan-level audit, do not optimize around tiny percentage differences in the census; the broad pattern is more defensible than fine-grained rank ordering.

## What to watch

- Whether PulseSignal’s extraction-audit accuracy remains near or above 90% as the dataset and classifier mature.
- Changes in the Flat / platform, Per-user, Mixed and Usage-based shares as the living census continues to refresh.
- Whether category-level pricing models converge or become more hybrid over the next year.
- Independent datasets using inspectable public pricing pages that reproduce or challenge PulseSignal’s roughly 43% Flat / platform result.
- How public self-serve pricing diverges from contract-level purchasing data as more SaaS companies hide enterprise pricing behind sales.

## Uncertainties

- PulseSignal is a pricing-monitoring vendor and the census is derived from its own automated extraction system rather than an independently audited market database.
- The latest disclosed plan-level audit found 90% of a 61-reading sample accurate, so individual classifications and small aggregate differences can still contain extraction error.
- Only companies with readable numeric paid pricing enter the model-share denominator; quote-only vendors may have systematically different pricing models.
- Company-level classifications are mutually exclusive and can hide the fact that a Mixed company uses per-seat pricing on one plan.
- The $29 entry-price median excludes non-USD observations rather than currency-converting them.
- Comparisons with Toolradar, Vertice and other research are directional because each dataset defines and samples pricing models differently.

## Sources

- [How SaaS Is Priced: The 2026 Census](https://getpulsesignal.com/insights/saas-pricing-models) — PulseSignal · primary_dataset · 2026-08-29T00:00:00+12:00. Primary living census; September 2 snapshot provides current denominators, model shares, category tables, entry-price medians, extraction method and the August 30 90% plan-level audit.
- [SaaS Pricing Models 2026: Patterns Across 2,575 Paid Tools](https://toolradar.com/reports/saas-pricing-models-2026) — Toolradar Research · independent_dataset · 2026-05-15T00:00:00+12:00. Separate public-pricing dataset used only as methodological context; finds explicit per-seat wording is a minority under its conservative text-match approach.
- [SaaS pricing models](https://www.vertice.one/insights/saas-pricing-models) — Vertice · independent_procurement_data · 2026-09-05T20:26:00+12:00. Contrasting procurement-oriented dataset showing consumption-based pricing leading in Q2 2026; useful for demonstrating that different sampling/classification methods answer different questions.