How Do You Price a SaaS Product?

Setting a software price is one of the most uncomfortable decisions of a launch: too low and you cannot live off your product; too high and nobody tries it. And contrary to a common belief, there is no formula: benchmarks give landmarks, never a number.

We ourselves changed our grids several times. VoxCut Android went from €11.99 to €14.99 as a one-time payment; VidScope reworked four tiers; VectorPop is €39 for life. This article shares the method that served us, rather than a promise of an ideal price. It complements our analysis of why a SaaS can have users but no sales.

The answer depends on frequency of use and on what you cost to run. A tool used every week that relies on a paid online service (servers, an AI API, storage) justifies a subscription: your costs recur, so should your revenue. A tool used three times a year that runs entirely on the user's computer lends itself poorly to a subscription: nobody wants to pay monthly for an occasional need.

That reasoning led us to a one-time payment for VectorPop, InOneShot and Nyctale: the maths is simple for the user, and 'no subscription' is a real differentiator against competing services billed monthly. Our article on the real cost of a custom tool versus SaaS over three years details that calculation from the buyer's side.

Conversely, VidScope, which uses AI at each analysis, mixes pay-per-use tiers, a monthly subscription and a lifetime purchase capped by a monthly quota. Each model matches a profile: the occasional user, the regular one, the one who wants to pay once.

Start from value, not from your costs. Ask what the customer saves or gains: how much time, money, errors? A tool that spares a professional billing €50 an hour one hour of manual work a week earns them more than €2,500 a year, and a €39 price then looks trivial. Conversely, a convenience tool for an individual must stay close to the price of a coffee or a meal.

Then look at the alternatives, including free and manual ones. A small program's real competitor is often the spreadsheet, open-source software or work done by hand. If your price is several times what the alternative costs without an obvious time gain, it will be hard to defend.

Finally, get real potential customers talking. Ask five people at what price they would say 'that is too expensive' and at what price they would doubt the quality. The range between the two is your ground, and it is almost always narrower than you imagine.

Think about how the price is presented too. To individuals, show the price including taxes, which is what they compare. To professionals, the price excluding tax is the norm. A price shown the wrong way can look 20% dearer or cheaper than reality, and create a misunderstanding at payment time.

Free exists to let people see the value before being asked to pay, but it must be calibrated. If the free version covers the whole need, nobody will upgrade; if it is too limited, nobody sees the point. On VectorPop, vectorization, presets and preview stay unlimited, and only export is counted, with five exports included to test real files: the user sees the result, and pays when they want to take it away regularly or in high quality.

A time-limited trial suits products whose value appears over time, such as a management tool. Freemium suits occasional-use products where everyone should be able to test without commitment. A demo, finally, mostly suits software sold to businesses, which want to see before buying.

Whatever the mechanism, when you offer to charge matters more than the design of the offer. Field experience converges: proposing purchase right after the user got a concrete result converts better than showing it on opening. Show the value first, ask afterwards.

The fewer tiers the better. Three or four offers are plenty, with a recommended one highlighted. Each tier must answer an identifiable profile, not an artificial variation in quantity. A grid that forces three re-reads to understand what you get drives people away as much as a price that is too high.

Be consistent everywhere. A price that differs between the home page, the promotional banner and the checkout page destroys trust in seconds. We have had the case: a banner announced a launch offer at €79 while the pricing card showed €149. The visitor does not know which is true, and chooses not to buy.

Beware of permanent discounts too. A launch offer must have a real limit, in number or date, otherwise it simply becomes the price, and buyers know it. Conversely, a price rise announced in advance, from a given date, creates a legitimate sense of urgency.

A readability tip: bring the price down to a meaningful unit. For a €39 lifetime tool, comparing with the cost of an equivalent monthly subscription over two years makes the saving obvious without needing to boast about it. The comparison must stay honest: cite real competitors, with their real, verified prices.

You change a price when you have data, not when you doubt. With no sales, lowering is tempting, but it is not always the right answer: if the purchase path is broken or the audience poorly targeted, a lower price will fix nothing. Check these two points before touching the number.

If buy clicks are many but payments rare, look at price and payment: hidden fees, extra steps, unexpected currency. If visitors do not even click on the offer, the problem is more in the promise or positioning. Change one parameter at a time to know what acts.

Finally, once the price is validated, do not change it without reason. A carefully reworked grid beats a series of nervous adjustments: it gives customers a stable message and you a stable frame for measuring.

The first was letting a price move several times in the same day. During the VoxCut Android decision, the figure swung between €14.99 and €19.99 before settling, after a debate that nearly led to an inconsistency between the annual and lifetime offers. The price finally chosen, €14.99 one-time, was the right one. But deciding fast, with no time to step back, is a risk: write down the options, sleep on it, decide the next day.

The second is the inconsistency between pages, mentioned above. It seems anecdotal, but it blurs trust at the most sensitive moment. After a price change, always walk through every page where it appears: home, price grid, banner, checkout page, app-store listing.

The third was opening the free tier too widely. On VoxCut, a free version with unlimited exports had been considered, before we questioned it the same day: it opened the floodgates and left no reason to go paid. Free should show the value, not deliver it entirely. That is what led us to VectorPop's free export quota.

Setting a software price means choosing a model consistent with frequency of use, anchoring the number in customer value, calibrating the free tier, then keeping a simple and stable grid. Market benchmarks help, but only your own data will tell you whether you are right.

If you are preparing to launch software or a SaaS, La Fabrik Numérique can help frame the offer as well as the technical build. Let us talk from the contact page, or first read our article on the difference between a showcase website and a web app.