Pricing Model Picker: Per-Seat vs Usage vs Hybrid, Honestly

Pricing model choice is one of the highest-leverage decisions a solo founder makes, and one of the most reversed. The four options (flat-monthly, per-seat, usage-based, hybrid) are not interchangeable and the "right" answer depends on the value metric, the churn profile, and the operational cost of metering. This article walks the picker on a concrete heavy-usage scenario and names when each model is the honest choice.

1. The four pricing models the picker scores

The picker scores four models. Flat monthly: one price per account regardless of seats or usage (the picker prices it at the five-seat equivalent of the competitor seat price). Simple, predictable, fails when account size varies materially. Per-seat: account pays per user. It remains the standard primary model in B2B SaaS — as an illustrative range, a plurality of SaaS under $10M ARR use per-seat as primary^[1]. Usage-based: account pays per metered unit (API calls, documents, GB, tokens). Usage-based penetration has risen materially among newer SaaS companies over recent years^[3].

Hybrid: combines per-seat for baseline access with usage-based overage for variable consumption. The picker bundles a fixed number of included units per seat and meters everything above that as overage. Common in AI-tooling SaaS where seats unlock the product and tokens or generations meter actual value. Paddle's pricing benchmarks^[2] show hybrid models in 18% of AI-tooling SaaS launched in 2024 vs 6% of pre-2020 SaaS, a clear directional shift.

Each model has a value-metric assumption underneath. Per-seat assumes value scales with team size. Usage-based assumes value scales with consumption. Hybrid assumes both — baseline access has value, and consumption has additional value. The wrong-value-metric error is the most common pricing mistake; the five numbers before pricing SaaS article covers the diagnostic.

2. Worked scenario: a heavy-usage document SaaS

Inputs: 8 users per account, 500 metered units per user per month, $0.04 gross margin per unit, $30 competitor per-seat price, $0.12 competitor usage price, and 4% monthly account churn. This is a document- or generation-heavy product where each account runs 4,000 units a month — the regime where the seat-vs-usage question actually bites. Running the picker on exactly these inputs:

The picker returns the four per-account-per-month revenue projections (each already scaled by the 0.96 retained fraction): flat monthly $144, per-seat $230.40, usage-based $460.80, and hybrid (seat + overage) $599.04. The recommended model is hybrid, at $599.04 per account per month — it captures both the seat base and the heavy consumption tail that the other three models each leave on one axis.

3. Why hybrid beats per-seat here

The hybrid number decomposes cleanly. Each account holds 8 seats at $30, a $240 base. The picker includes 100 units per seat (800 units), so 3,200 of the 4,000 units bill as overage at $0.12, adding $384. The pre-churn total is $624; the 0.96 retained fraction brings it to $599.04. Per-seat alone bills only the $240 base (×0.96 = $230.40) and gives away the entire 4,000-unit consumption tail — which is exactly the value the customer is buying.

This is the honest core of the seat-vs-usage debate. In a heavy-usage product, headcount is a weak proxy for value. Two accounts with the same 8 seats can differ 10x in units processed; per-seat charges them identically and either underprices the heavy user or overprices the light one. Usage-based ($460.80) fixes the proxy but throws away the predictable seat revenue. Hybrid keeps both: the seat fee covers baseline access and predictable infrastructure, the overage charge tracks the variable consumption that drives cost and value.

The trade-off: hybrid is harder to communicate. Customers want to know "what will this cost me." Per-seat answers cleanly. Usage-based answers with a calculator. Hybrid requires both. The packaging implication: build a pricing-page-as-calculator if the hybrid model is non-trivial. The AI solopreneur stack covers the supporting infrastructure.

4. Per-seat pricing, honestly

Per-seat wins when team size is the customer's natural unit of value. Three product categories where this is clean: collaboration tools (Slack, Notion), CRM (Hubspot, Salesforce), and design tools (Figma). Three categories where per-seat is awkward: AI-tooling (one seat can drive 1,000x the API spend of another), data products (the seat does not consume the data, the team does), and consumption-heavy infrastructure (CDN, monitoring). The worked scenario is squarely in the awkward column — per-seat lands dead last among the non-trivial models at $230.40 because the value is in the 4,000 units, not the 8 seats.

The hidden cost of per-seat: customers actively suppress seat count. The pattern is universal — one shared login, three people using it. This is not a customer-success failure, it is a pricing-model failure. The customer cannot rationalize paying for 3 seats when 1 seat does the work. Usage-based or hybrid pricing eliminates this anti-pattern because consumption is metered, not access.

As an illustrative range^[1], median per-seat pricing commonly runs $20 to $50 for SaaS under $1M ARR, climbing to $80 to $200 for products serving larger enterprise teams. The $30/seat input in the worked scenario is in the typical solo-SaaS range.

5. Usage-based pricing, honestly

Usage-based wins when consumption is the natural unit of value and customers can predict their consumption within a 2x range. In the worked scenario it doubles per-seat ($460.80 vs $230.40) precisely because 4,000 units per account is the real value being delivered. Three good usage metrics: API calls, processed documents, generated outputs. Three bad usage metrics: stored data (consumption flows in but never out, customers hate the bill creep), elapsed compute time (impossible to predict), and "active users" (already a seat metric in disguise).

The biggest objection to usage-based pricing from buyers is unpredictability. A B2B buyer needs to budget a line item; usage that swings 5x between months breaks the budget process. The solution is either a committed-spend floor (customer pays for X units whether they use them or not) or a usage-cap with overage protection (customer's bill is bounded above by a known number). Hybrid is one such floor: the seat base is the committed spend and overage is the variable tail.

Paddle's research^[5] shows that pure usage-based pricing has 30% to 50% higher first-year revenue volatility than per-seat for the same product. Hybrid smooths this materially. The annual contract pricing article covers how to convert volatile usage into stable revenue via annual commitments.

6. How churn changes the answer

The picker returns a retained fraction of 0.96 at 4% monthly account churn. This input scales every displayed revenue projection by the same factor, so it does not move the relative ordering — hybrid stays on top whether churn is 3% or 8%. The churn input changes the magnitude of every number, not which model wins.

Where churn changes the recommendation is at the extremes, in lifetime terms the single-month picker does not surface. At 1% monthly churn (B2B with annual contracts and high switching cost), per-seat pricing accumulates value over years and can beat usage-based on lifetime revenue per account. At 12% monthly churn (consumer SaaS or self-serve SMB), only usage-based pricing — which collects value in the early months before churn — generates positive unit economics. The model picker captures the single-month snapshot; the lifetime read is downstream of it.

The SaaS churn cohort vs aggregate article covers how to measure the input accurately. Founders who use a poor churn input get a poor pricing recommendation; the rest of the math is downstream.

7. Switching pricing models without breaking trust

Pricing-model switches are common (most SaaS does it at least once) and expensive when done badly. Three rules.

1. Grandfather existing customers for at least 12 months. Anyone on a contract or expectation of stable pricing should keep the old model until renewal. Forcing a switch mid-contract is a churn event.
2. Switch only one variable at a time. If you are moving from flat-monthly to per-seat, do not also change the price level. Customers can absorb one change; they cannot absorb three at once.
3. Communicate the value-metric reasoning, not the financial reasoning. "We are switching to per-usage because it tracks how much value you get from the product" lands. "We are switching to per-usage because the average revenue per customer is too low" repels.

The methodology behind the picker's revenue projections is documented at the Pricing Model Picker methodology page^[4], including the retention-adjusted formula and the included-units-per-seat bundling.

Frequently asked questions

Is per-seat pricing always better than flat-monthly?

Not always, but in the worked scenario per-seat ($230.40 per account) doubles flat-monthly ($144). Per-seat beats flat-monthly whenever accounts hold more than five seats, because the picker's flat-monthly model is pinned to a five-seat-equivalent price. Per-seat wins when account team sizes are large and vary; flat-monthly wins when teams are predictably small and price sensitivity is high.

When should a solo founder use usage-based pricing?

When the value-metric is clear, easy to meter, and aligned with customer perception of value (API calls, processed documents, generated outputs). In the worked scenario, usage-based ($460.80) doubles per-seat ($230.40) because each account consumes 4,000 metered units. Avoid usage-based when customers cannot predict their consumption, when metering infrastructure adds operational burden, or when the value-metric does not map to customer outcome.

Why does hybrid pricing win the worked scenario?

Hybrid (seat + overage) at $599.04/mo per account beats the others because it captures both the baseline access value (via the eight seats at $30) and the variable-value tail (via overage on the 3,200 units beyond the 800 included). It collects the per-seat base AND the consumption tail that pure per-seat leaves on the table.

How does account churn affect the recommendation?

Churn scales every model's revenue by the same retained fraction (0.96 at 4% monthly churn), so it does not change the ordering — hybrid stays on top. Where churn flips the recommendation is at the extremes: at very low churn, per-seat accumulates lifetime value; at very high churn, only usage-based — which captures value early — generates positive unit economics.