Learn why price is the strongest SaaS growth lever, how AI is breaking per-seat pricing, and three practical ways to fix leakage and move toward value-based pricing.
Session 4 of the Flashpoint Venture Scout Academy was a guest session on SaaS pricing, run by Krzysztof Szyszkiewicz of Valueships, a consultancy that works only on pricing for technology companies. He covered why price is the strongest growth lever, how AI is breaking old pricing models, and three things any software company can do about it. It's written for scouts, angels and operators who want to read a startup's pricing with a sharper eye.
Because a small change in price moves profit more than the same change in anything else. In the benchmark Krzysztof shared, a 1% price improvement lifts profit by about 12%, compared with roughly 3.5% for a 1% gain in volume and about 2.5% for a 1% cut in fixed costs.
The other levers are getting harder to pull. AI compute makes costs volatile, and sales gets tougher as inboxes fill with automated outreach. Yet almost no startup has anyone whose job is pricing.
Across Valueships' projects, the average price went up in about 90% of cases. If a company prices off costs, competitors or gut feeling, it's probably too cheap.
Companies price on cost, on competitors, on context or on value. Value-based pricing captures the most, and it's also the hardest to do well.
You charge a share of the value the customer gets, and willingness to pay rises with how much value the customer actually perceives. So the job has two parts: create value, then make sure the customer understands it.
Krzysztof compared it to a car manual: almost nobody reads it, so few people know everything the car can do. Software customers likewise perceive less value than the product delivers, and onboarding and communication close that gap.
He then walked through a "price waterfall" from value to cash:
Each step points to a different problem. Low conversion or tough negotiations usually mean buyers don't see the value. Good conversion with weak margins means the list price is too low. Pricing on par with rivals but winning less? Look at discounting.
For the fastest return, work right to left: start with discounts and existing customers, then list price, and only then perceived value. Customers notice a price change right away, but a change in messaging takes much longer to land.
AI is breaking per-seat pricing, adding real costs to every unit sold, threatening exposed incumbents and making hidden prices easy to guess. Most software companies still charge per user or per access, which is the wrong metric for a market moving this fast.
If a smaller team plus AI agents can produce the same output, a company billing per seat shrinks as its customers cut headcount. Krzysztof sees more companies testing other metrics, even if few have replaced seats as their main one.
Classic software worked like a railway: expensive to build, then each extra trip cost almost nothing. AI features carry a real cost per use, and vendors can't fully predict customer usage or what model providers will charge next.
He mapped this on two axes: how far AI can automate the user's task, and how easily AI can get into the workflow. Companies high on both, such as marketing tools without sensitive data, should plan their defense now.
A buyer can now ask an AI assistant what an enterprise license usually costs before filling in the "contact sales" form. Whether that estimate is right or wrong, the buyer is now anchored to it. Browser plug-ins that crowdsource quoted prices push the same way, so companies need to own the pricing story and prove the value behind the number.
Krzysztof's response to all four trends is to treat pricing as a core skill you exercise often. One benchmark he cited: the 500 largest product companies changed their pricing about 3.6 times in a year, mostly through small moves like shifting a feature between plans or testing a price with a slice of users. Lovable, Clay and Salesforce all change or test pricing regularly.
List every customer with their standard price, what they actually pay and when they started. The gap between standard and actual price is your leakage, and most of it usually has no good reason behind it.
Krzysztof's process is simple enough to do in about 15 minutes:
The point is to know how much churn you can afford. In his experience, break-even churn sits far above what actually happens after a well-run increase, which makes this the easiest revenue to capture before a next funding round.
Context-based pricing means adjusting your baseline price to the buyer's situation and your own. It matters most in sales-led companies, where every deal is quoted.
Choose a value metric closer to the outcome the customer pays for, and test it on an add-on or with usage limits before touching the core product. Switching the main pricing metric is the hardest pricing change a company can make.
The options run from least to most tied to value: seats and access, then usage or credits, then work and output, then success. An SEO tool, for example, could charge per user, per article optimized or per ranking achieved. The closer the metric is to the result, the larger the share of value a vendor can capture.
Public examples he cited include Chargeflow, which takes 25% of each chargeback it wins back, and Salesforce's Agentforce, launched at $2 per conversation. Credits work for complex products with many use cases. For a point solution, he called them a lazy choice when an output metric is available.
His two ways to test:
For an early startup, he suggested a pricing page with two or three value metrics plus a small ad budget to see which draws clicks. Below-benchmark click-through is an early warning about the idea.
Pricing tells you a lot about a startup quickly. Ask what the company charges for, who owns pricing, when prices last changed and how far realized prices sit below list.
A founder who prices on cost or by copying competitors, bills per seat in a market AI is automating, and hasn't touched prices in years probably has upside left. A founder who tests pricing often and ties it to customer outcomes is usually closer to the customer, too.
Those observations help separate a good product with weak monetization from a business that knows what it's worth.
In Krzysztof's experience, pricing usually sits best in product, since product owns value delivery and hears customer feedback. The ideal is a part-time pricing owner who works across sales, customer success, product and finance.
Some may, so the goal is to know how much churn you can afford before you act. In practice, a well-explained increase tied to the features customers use rarely causes churn anywhere near the break-even level.
Credits suit complex products with many different use cases. For a point solution, a metric tied to output, such as resolutions or workflows completed, usually captures more value and is easier for buyers to understand.
Build a pricing page with two or three different value metrics and run a small ad campaign to see which one gets clicks. If click-through rates sit below normal benchmarks, revisit the idea or the value message.
Krzysztof thinks the market is heading toward outcome-based and success-based pricing, but over years rather than months. Another path he described is AI tailoring software so closely to each customer that SaaS starts to look like a service.
Scouts send us companies from their network, join Scout Academy sessions like this one for free, and earn when we invest.
Apply to become a Scout
