Top 7 SaaS Pricing Mistakes That Quietly Cap Your Revenue
You do not have a traffic problem. You have a pricing page that is charging your best customers the least.
Most founders set a price once, flinch, and never touch it again. That single flinch costs more than a quarter of marketing, and unlike marketing it is free to fix.
None of what follows is about charging more for its own sake. Each one is a place where the amount somebody pays has come loose from the value they get — which is bad in both directions, because it also means the customers getting the most from you are often paying the least.
1. Charging per seat when the value is not per seat
A team of two running forty projects pays you less than a team of forty running one. If your product's work scales with projects, documents, API calls or transactions, per-seat pricing inverts the relationship between value and price.
Seat pricing is popular because it is easy to explain and easy to forecast. That is a real advantage and it is why it is worth keeping when the value genuinely is per person — a communication tool, a design tool, anything where one more human means one more set of work.
The test is simple: name your heaviest customer and your lightest, then compare what they pay. If the heavy one pays less, your metric is wrong.
2. One plan, priced for the median
The median customer is nobody. A single plan prices out everybody smaller and leaves money on the table with everybody larger, and the larger ones were willing to pay for the certainty of a bigger plan.
Two or three tiers is usually enough. The point is not to maximise choice — too many options stall the decision — but to let a customer self-select by how much they need, which is information you cannot otherwise get before they sign up.
3. Never raising prices
The price you set when the product did one thing is still the price it charges now that it does nine. Every feature shipped since then has widened the gap between what people get and what they pay.
The fear is real but it is usually misplaced: founders imagine the backlash from existing customers, then protect against it by never raising anything, including for people who have not signed up yet.
Raise it for new customers first. Grandfather the existing ones for the first few increases. You will find the risk was almost entirely in the anticipation, and the revenue was in the customers you had not met.
4. No annual option
An annual plan does three things at once: it collects a year of cash now, it removes twelve chances to churn, and it tells you which customers are committed. A ten to twenty percent discount buys all three.
Founders skip this because it feels like discounting. It is closer to the opposite — you are being paid sooner for a customer who was going to stay anyway, and you find out who those are.
5. Pricing against competitors instead of against the alternative
Your competitor's price is a number they guessed, possibly years ago, possibly while funded differently from you. Matching it imports their guess into your business.
The number that matters is what your customer does today without you. If somebody is paying a contractor four hundred a month, or burning six hours a week on a spreadsheet, that is the real comparison — and it is almost always a larger number than the nearest tool's price page.
This is also why the interviews matter more than the competitive research. Only a customer can tell you what the alternative costs them.
6. A free plan that solves the problem completely
Free works when free users produce something: content that gets indexed, invitations that bring colleagues, integrations that make the product stickier inside a company. That is a marketing channel with a hosting bill.
Free does not work when it simply solves the problem for your smallest customers. Then it is a support queue, a hosting bill and a conversion rate you cannot move, because nobody upgrades from a plan that already does what they need.
The honest test: what does a free user do that makes a paying one more likely? If the answer is nothing, you have a discount you cannot withdraw.
7. Hiding the price
"Contact us for pricing" is right when deals are genuinely negotiated — procurement, security review, a real salesperson. For a self-serve product it costs you every visitor who was comparing options at eleven at night and will not fill in a form to find out whether you are affordable.
The ones who leave are disproportionately the ones who would have bought without a conversation, which is the cheapest revenue you have.
What to change this week
One thing: find your heaviest and lightest customers and compare what they pay. If the relationship is backwards, your pricing metric is the problem and nothing else on this list will matter until it changes.
If it is the right way round, raise your prices for new customers by the amount that makes you slightly uncomfortable, and leave everybody else alone. Then tell other founders what happened — pricing is the thing everybody guesses at privately and almost nobody reports back on.
Frequently asked questions
How often should a SaaS raise its prices?
Once a year is a reasonable default, applied to new customers first. The decision that matters is not the frequency but whether you ever do it at all — most founders set a price once, flinch, and never revisit it.
Should I grandfather existing customers when I raise prices?
For the first few increases, yes. It removes almost all the risk from the decision and costs you very little, because the revenue is in the customers you have not signed yet. Grandfathering forever eventually becomes its own problem.
What is the most common SaaS pricing mistake?
Charging per seat when the value is not per seat. A team of two running forty projects pays less than a team of forty running one, which inverts the relationship between what a customer gets and what they pay.
Is a free plan worth it?
Only when free users produce something — content, invitations, integrations — that brings paying ones. A free plan that simply solves the problem for your smallest customers is a support cost with no return.