SaaS Pricing Models: What 225 Live Pricing Pages Actually Use

Most guides to SaaS pricing models present seven or eight options as if the market were evenly split between them. It isn’t. Our directory ingests 225 live pricing pages, and two models account for roughly three-quarters of everything that actually ships. The interesting question isn’t “what are my options.” It’s why the market keeps converging on the same two answers, and whether your product is one of the exceptions.

Here’s the real distribution, straight from the directory: 41% of products price per seat, 28% are usage-based, 9% hide behind a custom quote, 8% charge a flat rate, and 7% run a hybrid. Everything else you’ve read about (freemium, tiered, per-feature) is a layer on top of one of these five, not a separate model.

The five SaaS pricing models, by the numbers

Per-seat: the default for a reason (41%)

Per-seat wins because it’s legible. A buyer can estimate their bill in their head, finance can forecast it, and expansion revenue arrives automatically as the customer’s team grows. The failure mode is just as well known: seats get shared, viewers get counted like editors, and AI features break the model entirely when one seat suddenly does the work of four. If your product’s value doesn’t scale with headcount, per-seat quietly caps your revenue at the customer’s org chart.

Usage-based: growing fast, harder to run (28%)

Usage-based pricing aligns price with value better than anything else on this list, when the metric is right. The catch is operational: customers fear surprise bills, finance teams hate variable line items, and you need real metering infrastructure before you can send an invoice. We covered a live example in the Linear vs Jira teardown, and you can see the current tier structure on Linear’s pricing page in the directory; the products that make usage work almost always pair it with a generous included allowance and a hard spend cap.

Custom quote: the enterprise toll booth (9%)

Nearly one in ten products in the directory publishes no price at all. That’s a deliberate choice, not a gap: custom pricing maximizes extraction from large buyers and keeps competitors guessing. It also filters out everyone who won’t sit through a sales call. Unless your ACV justifies a salesperson on every deal, this model costs you more pipeline than it protects. The tell is in your own funnel: if inbound prospects ask for a price and vanish when told to book a call, the model is taxing pipeline you’ll never see in a CRM report.

Flat-rate: rare, and usually a statement (8%)

Flat-rate is the contrarian pick: one price, everything included. It sells simplicity and “no surprise bills,” a real wedge when your category leader just added sliders and add-ons. The trade-off is a hard ceiling on expansion revenue; flat-rate products grow by adding customers, not by growing inside accounts.

Hybrid: powerful, complex, easy to fumble (7%)

Hybrids (a platform fee plus usage, or seats plus consumption credits) capture value on two axes at once. They’re also the easiest model to over-engineer. If a prospect can’t estimate their bill within a minute of landing on your pricing page, the model is costing you deals you’ll never see in a lost-deal report.

What the starting prices tell you

Across the 120 products with a published entry price, the median starting tier is $21.50 per month, the floor is $3, and the ceiling is $299. Just over half the directory (56%) offers a free plan, and about the same share offer a free trial. In other words: the market’s default motion is “start free or cheap, expand later,” which is a packaging decision layered on top of whichever of the five models you run. Freemium isn’t a pricing model; it’s an acquisition strategy. We broke down when each works in freemium vs free trial.

The $3-to-$299 spread on entry prices is worth reading as a positioning map, too. The products at the floor are buying distribution and betting everything on expansion; a $299 entry tier is a qualification filter doing the work a sales team would otherwise do. Your entry price tells the market who you’re for before a single word of copy loads, and it should be chosen with that job in mind, not just unit economics.

How to choose yours

Match the model to how your value scales, not to what your favorite company does. If value scales with people collaborating, per-seat is honest. If it scales with consumption (API calls, checks, contacts, events), usage-based is honest, and per-seat will either overcharge small-heavy users or undercharge big ones. If your buyer is an enterprise with procurement and your ACV clears five figures, custom pricing stops being cowardice and starts being strategy. Flat-rate is for products with narrow, predictable cost curves and a positioning story built on simplicity.

And if you’re early stage: pick the boring, legible option in your category and spend your creativity on packaging instead. We’ve made the longer version of that argument in our SaaS pricing strategy guide: the model matters less than the value metric underneath it.

The legibility test: run it before you ship

Every model above lives or dies on whether a buyer can estimate their own bill, so test exactly that before a new pricing page ships. Put the draft in front of a handful of prospects or friendly customers and ask one question: what would this cost your team? Don’t explain the tiers and don’t guide the math. You’re scoring three things: whether they can name the metric you charge on, whether their estimate lands anywhere near what you would actually invoice, and whether the guessing visibly makes them nervous.

The results map straight onto the failure modes above. Prospects who can’t name the metric are telling you your hybrid is over-engineered. Prospects who guess low are the seat-sharing problem in embryo; they’re already planning to consolidate licenses. Prospects who guess high and wince are usage-based billing anxiety arriving before the first invoice, which is your cue to lead with the included allowance and the spend cap instead of burying them in a footnote. The whole exercise costs an afternoon, and it catches model-level mistakes that A/B testing the page never will, because a conversion test can’t tell “wrong headline” from “wrong model.”

Switching models later, without burning the base

Most teams don’t get the model right on the first pass, so it’s worth knowing the migration mechanics before you need them. The sequencing that works: new customers first, existing customers later or never. Launch the new model for new signups, grandfather the installed base, and run both in parallel while you learn where the new model’s revenue actually lands. Forced migrations are where the horror stories come from: they take your most loyal customers, the ones who’ve been around long enough to sit on old pricing, and convert them into your loudest critics with a single email.

Two rules make a migration survivable. First, grandfather generously but put a boundary on it: an open-ended “legacy plan” is a liability that compounds, while a defined transition window is a courtesy with an end date. Second, translate the change into each customer’s actual bill before they have to do it themselves. Any migration announcement that doesn’t say “here is what you, specifically, will pay” generates a support ticket per recipient. The customers who churn during a model change are rarely churning over the new price; they’re churning over the week they spent not knowing what the new price would be.

The one move worth making this week

Pull up the pricing pages of your five closest competitors and bucket each into one of the five saas pricing models above. If everyone in your category runs the same model, note what buyers complain about (check their reviews: billing gripes are gold) and ask whether the opposite model is a wedge. That’s a 30-minute exercise that occasionally produces a repositioning insight worth more than a quarter of feature work.

Outmano tracks the live pricing pages behind these numbers: 225 SaaS products and counting, refreshed weekly, with change alerts when a competitor moves. Browse the directory at outmano.com/tools.

Frequently Asked Questions

What are the main SaaS pricing models?

Five models cover essentially the whole market: per-seat, usage-based, flat-rate, hybrid, and custom (quote-only) pricing. In our directory of 225 SaaS products, per-seat is the most common at 41%, followed by usage-based at 28%. Freemium and tiering aren’t separate models: they’re packaging layers on top of one of the five.

Which pricing model is best for an early-stage SaaS?

The one your buyer can understand without a calculator: usually per-seat or a simple flat rate, whichever matches how your value scales. Early on, pricing legibility converts better than pricing cleverness. You can migrate to usage-based or hybrid later, once you know which metric actually correlates with the value customers get.

Is usage-based pricing better than per-seat?

Neither is better in the abstract; they price different things. Usage-based wins when consumption, not headcount, drives value: infrastructure, APIs, messaging. Per-seat wins for collaboration tools where each added user genuinely gets value. The worst outcome is picking usage-based for the growth story and then fighting billing anxiety in every sales cycle.

How many pricing tiers should a SaaS have?

Three is the well-worn default: an entry tier to reduce friction, a middle tier where most revenue lands, and a high anchor. In our data, most products with parsed pricing run two to four public tiers plus an enterprise contact-sales option. More than four public tiers usually signals a segmentation problem, not a pricing strategy.

How often do SaaS companies change their pricing model?

Full model changes (like flat-rate to usage-based) are rare but increasingly common; we’ve detected several major ones in the directory this year alone, including monitoring tools moving to usage-based sliders. Price and packaging tweaks are far more frequent, often quarterly. If you haven’t revisited pricing in 18 months, you’re overdue.