Your first price isn’t just a price. It’s the beginning of your commercial model. Why pricing is the commercial decision technology companies find hardest to get right, and hardest to act on.<\/em><\/p>\n\n\n\n The foundational evidence on price as a profit lever is old, and it is worth being clear about what it is.<\/p>\n\n\n\n McKinsey’s best-known analysis took the average economics of an S&P 1500 company and found that a 1% price rise, with volumes held steady, would lift operating profit by 8%. That is roughly half as much again as a 1% cut in variable costs, and more than three times the effect of a 1% rise in volume. An earlier 1992 version, on a different sample of companies, put the figure at around 11%. Neither is an experiment. Critics have long pointed out that it is a what-if calculation run on existing margins, not a test of what happens when prices actually change. The key assumption is in the premise: volume holds.<\/p>\n\n\n\n So the finding hasn’t been “replicated” in the scientific sense, because it doesn’t need to be. It is arithmetic, and it holds wherever operating margins are thin. For a pre-profit technology company the percentage is meaningless, but the logic is sharper, not weaker. Every point of price that holds drops straight into contribution margin, and from there into burn rate and runway. The real question is never the arithmetic. It is whether the volume assumption survives contact with customers.<\/p>\n\n\n\n The newer evidence is about exactly that question. In a 2025 survey of 240 software companies, three in four had changed their pricing in the previous year. Among those that did, the median reported uplift in revenue growth was 10%, and only 3% said the change had hurt growth. These are self-reported outcomes from companies that chose to change, so they show association, not proof. But they point the same way as the arithmetic.<\/p>\n\n\n\n Investors have long treated this as the core of business quality. Asked by the US Financial Crisis Inquiry Commission how he assessed a business, Warren Buffett said: “The single most important decision in evaluating a business is pricing power.” A company that can raise prices without losing customers to a competitor, he argued, has a very good business.<\/p>\n\n\n\n Most companies don’t believe they have that power, and the belief has been remarkably stable over time. Simon-Kucher’s 2011 Global Pricing Study surveyed 3,904 respondents. Only 35% judged that their company had enough pricing power to achieve the right price, and companies reporting low pricing power also reported EBITDA margins around a quarter lower than those reporting high pricing power (roughly 12% against 16%). That is an association between two self-assessments, not evidence that one causes the other. It is also a sample weighted towards large firms: 41% of respondents came from companies with sales above \u20ac1bn.<\/p>\n\n\n\n The finding that companies struggle to realise price increases has held across successive waves of the same study:<\/p>\n\n\n\n The same 2025 study found that only 40% of companies rank pricing as their top profit lever.<\/p>\n\n\n\n So pricing matters enormously, the evidence for that is consistent, and most companies still report that they don’t capture it. That gap is worth examining, because the reasons are structural rather than a matter of effort or intelligence. They apply with particular force to technology companies moving from validation into commercial growth.<\/p>\n\n\n\n If companies simply mispriced at random, some would be too high and some too low, and the market would correct both. Practitioner evidence suggests otherwise.<\/p>\n\n\n\n McKinsey’s pricing practice reported that in its experience, 80 to 90% of poorly chosen prices are too low. That is the judgement of experienced advisers across many engagements, not a measured dataset, and we have not found a large independent study that tests it directly. But the reasoning behind it is sound, and it explains why the asymmetry matters. A price that is too high won’t sell, which is relatively easy to fix by lowering it. A price that is too low forgoes revenue and profit and also fixes the product’s value position at a low level, and once prices reach the market they are difficult or even impossible to raise.<\/p>\n\n\n\n This is why a first price is not just a price. The launch price, net of discounts, becomes the market’s first reference point for what the maker believes the product is worth, and it tells the market more about that than any sales pitch. For technology companies, the first price is usually set under pressure: a pilot, a design partner, an enterprise customer whose logo matters more than their invoice. It is rarely thought of as a strategic decision. It becomes one anyway.<\/p>\n\n\n\n Underpricing can also damage credibility, not just margin. McKinsey describes a company whose data-management system claimed to save large firms hundreds of millions a year, but which launched the core software at an enterprise licence under $100,000 to win share quickly. Buyers didn’t take the claims seriously, because a product with that impact should have been priced alongside ERP systems costing $1 million or more. A low price did not remove a barrier to purchase. It created a new one: disbelief.<\/p>\n\n\n\n The broader record on new products is poor, though the data needs careful reading. In Simon-Kucher’s 2014 study of more than 1,600 professionals, respondents reported that 72% of their new products failed to meet profit targets, and a quarter said not one had. That figure is self-reported and says nothing about why products fell short. When Simon-Kucher itself discusses it, the causes it lists include feature overload, poor monetisation, and products nobody asked for. The firm’s partners have built a strong argument that monetisation is where much of that failure is decided, and it is worth taking seriously. But it is an argument informed by their practice, not a finding the survey establishes.<\/p>\n\n\n\n Most founders we speak to have done serious customer discovery. Some have had hundreds of conversations. They understand the problem, the buyer and the use case in detail. It’s reasonable for them to assume those conversations have also told them what the market will pay.<\/p>\n\n\n\n The research suggests they usually haven’t, for three reasons.<\/p>\n\n\n\n People overstate what they would pay when nothing is at stake.<\/strong> Economists call this hypothetical bias, and it is one of the most studied effects in valuation research. A meta-analysis of 28 studies that compared stated and actual payment using the same method found a median ratio of stated to actual value of 1.35, with a heavily skewed distribution. An earlier meta-analysis found that, on average, people overstated their preferences by a factor of about three in hypothetical settings.<\/p>\n\n\n\n Most of that work concerns environmental and public goods. The most relevant modern evidence comes from a 2020 marketing meta-analysis of 77 studies of private goods, which put the average gap at 21%. It also found that the gap was larger for higher-value products and specialty goods, and that indirect methods such as conjoint overstated real willingness to pay more than direct questions did. Those are consumer studies, and we know of no equivalent meta-analysis for B2B technology. But high-value, specialised purchases are exactly the category where the bias is largest.<\/p>\n\n\n\n A discovery conversation is hypothetical by definition. The product may not exist yet, and nobody is signing anything.<\/p>\n\n\n\n People understate when they think the answer will be used.<\/strong> The same literature describes an opposite effect. When respondents believe their answer can influence what they will later be charged, they have a reason to shade it. One study of university parking fees found staff rejecting options that were clearly in their interest because they expected the fee structure to be revised. We are not aware of research measuring this effect in B2B customer discovery. But the incentive is obvious: a prospective customer asked about price by the company that will later quote them has every reason to aim low. This is the founder’s specific difficulty. The same conversation that inflates interest may deflate price, and there is no way of knowing from the inside which effect is winning.<\/p>\n\n\n\n How you ask changes the answer.<\/strong> McKinsey describes a controls manufacturer that developed a new high-pressure steam valve for nuclear plants. Its first research described the technical benefits and asked customers to compare it with an existing valve, and most said a 20 to 25% premium was justified. When the company redid the research around the customer’s own economics, starting from the cost of maintenance shutdowns the valve could reduce, customers named a figure several times the price of the existing valve. Same product, same customers, very different answer. The academic evidence is mixed on which method is least biased: one meta-analysis favours choice-based approaches, another finds indirect methods overshoot more. What it agrees on is that the method matters.<\/p>\n\n\n\n None of this means customer conversations are wasted. They are essential for understanding the problem, the buyer and the context. They are simply the wrong instrument for measuring price, in the same way that a good interview is the wrong instrument for measuring demand at scale.<\/p>\n\n\n\n There is a second, less comfortable reason pricing is hard to do internally: the people best placed to understand the product are the least well placed to price it neutrally.<\/p>\n\n\n\n McKinsey’s review of new product pricing notes that many suppliers rely too heavily on their internal perceptions, which can unintentionally skew their market research, so that the work ends up confirming the benefits the developers already believe in, or the anecdotes the sales team has brought back.<\/p>\n\n\n\n Founders carry several anchors into any pricing decision, usually without noticing:<\/p>\n\n\n\n Each anchor is a legitimate input. None of them measures what a customer would pay for the value delivered, and together they pull in one direction: down. That is consistent with the practitioner view that most mispricing is underpricing.<\/p>\n\n\n\n If pricing is this important and this difficult, you might expect companies to assign it clearly. The evidence says most don’t.<\/p>\n\n\n\n OpenView’s survey of more than 1,000 SaaS executives found that among expansion-stage companies with $1\u201320M ARR, 55% had nobody whose job description included pricing, and for the rest it tended to be a small part of someone’s role rather than a dedicated focus. More than two-thirds said the CEO or leadership team owned pricing decisions. Companies without a dedicated pricing resource also rarely researched customer value and pricing, or A\/B tested price changes.<\/p>\n\n\n\n That survey is from 2017, but newer data shows the same pattern. In the 2025 survey of 240 software companies, the founder or CEO was the primary pricing owner in 71% of companies under $1M ARR and 69% of those at $1\u20135M. At $5\u201320M ARR, founder ownership fell to 38%, and 21% said nobody owned pricing. The report’s author called that band pricing’s no-man’s land. The samples in each revenue band are small, so treat the percentages as indicative. The shape is the point: pricing sits with the founder until the founder is too stretched to hold it, and then, for a while, it sits with no one.<\/p>\n\n\n\n In deeptech and hardware, in our experience, the commercial team is often the founder plus one or two people, stretched across fundraising, partnerships, pilots and sales.<\/p>\n\n\n\n When pricing belongs to everyone at leadership level, it belongs in practice to no one. That has a consequence that is easy to miss.<\/p>\n\n\n\n Earlier this year, we completed a market entry study for a technology company weighing up a new market. The recommendation was no-go, and the company accepted it. That is what a well-framed entry decision is supposed to produce: a clear answer, including a clear no.<\/p>\n\n\n\n But the study also found substantial pricing headroom in a smaller, commercially attractive segment, and identified pricing as the obvious next step. The company changed nothing.<\/p>\n\n\n\n That isn’t a criticism of them. It’s a pattern, and it illustrates something the pricing data points towards. A market entry question has a natural owner, a decision date, and a cost of inaction that is easy to see. A pricing opportunity has none of those. Nobody is waiting for the answer. No distributor contract or launch budget forces the question. And changing a price that currently works feels like a risk, while leaving it alone feels like nothing at all, even when the evidence suggests leaving it alone is the costly option. Behavioural economists have a name for this preference for the current state of affairs: status quo bias.<\/p>\n\n\n\n This is the second sense in which pricing is hard to do internally. The difficulty isn’t only getting a reliable answer. It’s that the answer, once found, has nowhere to land.<\/p>\n\n\n\n In many industries, pricing means setting a number. For technology companies, and especially for deeptech and hardware-enabled businesses, the harder decisions sit underneath the number:<\/p>\n\n\n\n In software, these choices are moving quickly. In the 2025 survey of 240 software companies, 41% named hybrid pricing (a subscription combined with usage) as their primary model, against 27% who said it had been their model a year earlier. Seat-based pricing fell from 21% to 15% over the same period. It is a survey of one newsletter’s readership, and the year-earlier figures rely on respondents’ recall. But the direction is clear, and it means the value-metric question is live even for companies that thought they had settled it.<\/p>\n\n\n\n The same product can also require a different model in a different market. In one pricing project we worked on, a safety instrumentation company sold hardware and a renewable software licence separately in the UK and Europe. US buyers wouldn’t buy that way. They expected hardware and software bundled together, with updates included, and they placed a materially higher value on the bundle. Different buying conventions required different packaging, and the packaging changed what customers would pay. The commercial model changed because buyer behaviour changed.<\/p>\n\n\n\n These are not decisions a company makes once it has enough customers to justify a pricing strategy. They are made, often by default, in the first pilots and the first contracts. Each one shapes what customers expect, what investors model and what the first commercial hire inherits.<\/p>\n\n\n\n When we set prices for our own services, we did what we recommend: we tested willingness to pay rather than trusting our own sense of what was fair. Four things happened, and each one echoes the evidence above.<\/p>\n\n\n\n The method had to fit the sample.<\/strong> Our first research design produced incoherent results at the sample size we could realistically reach, so we moved to a simpler sequential price test.<\/p>\n\n\n\n Demand didn’t follow our assumptions.<\/strong> The offer we expected to lead was not the one buyers valued most. Our instinct had been an anchor.<\/p>\n\n\n\n The cheapest offer lost money.<\/strong> Once we counted the fixed cost of scoping, kick-off and presentation that every engagement carries, the smallest one didn’t cover itself. The real price floor was higher than it looked.<\/p>\n\n\n\n Launch pricing nearly became the model.<\/strong> Prices set low “for now” were on their way to becoming the reference point for everything we sold.<\/p>\n\n\n\n We are experienced at this, and we still needed the data to overrule our own judgement. That is the point: being close to a product makes pricing it harder, not easier.<\/p>\n\n\n\n For an early-stage technology company, the practical conclusion is not that every founder needs a formal pricing strategy before their first customer. It’s that the pricing decisions being made now are not provisional, even when they feel it. The questions worth asking are straightforward:<\/p>\n\n\n\n For a scaling company with an established price, the conclusion is different but related. Years of data on realised price increases suggest that the hardest pricing question isn’t what to charge a new customer. It’s how to change what existing customers pay, without asking them directly and signalling the change before you’ve decided to make it.<\/p>\n\n\n\n In both cases, the pattern in the evidence is consistent. Pricing is the lever with the largest arithmetical effect on profit. It is also the one most exposed to biased evidence, internal anchors and unclear ownership, and the one whose findings are most likely to sit unused. The companies that handle it well treat price as a commercial decision from the start, not a number to revisit once there are enough customers to justify it.<\/p>\n\n\n\n We have separated foundational evidence from newer evidence.<\/p>\n\n\n\n Foundational evidence.<\/strong> McKinsey’s profit-lever analysis (1992, 2003) is arithmetic applied to average company margins. It shows how sensitive profit is to price if volume holds. It is not an experiment, and its size depends on margins.<\/p>\n\n\n\n Newer survey evidence.<\/strong> Simon-Kucher’s Global Pricing Studies (2011, 2014, 2025) and the 2025 Growth Unhinged\/Tremont survey are self-reported, and several are weighted towards larger companies. Where they link pricing to profit or growth, they show association, not cause. The finding that companies struggle to realise price increases has been consistent across every Simon-Kucher wave we checked.<\/p>\n\n\n\n Academic evidence.<\/strong> The work on hypothetical bias uses experimental comparisons of stated and actual payment. It mostly concerns public and consumer goods; we found no equivalent meta-analysis for B2B technology.<\/p>\n\n\n\n What we left out.<\/strong> Several widely repeated statistics about founder underpricing (for example, that most SaaS companies are underpriced by 30\u201350%, or that 81% of founders believe they underpriced) could not be traced to a primary dataset, and we have not used them. We found no large independent study, old or new, directly measuring how often new B2B products are underpriced. The claim rests on practitioner experience and on the mechanisms described above.<\/p>\n\n\n\n Sources:<\/em> Marn and Rosiello, “Managing Price, Gaining Profit”, Harvard Business Review<\/em> (1992); Marn, Roegner and Zawada, “The Power of Pricing” and “Pricing New Products”, McKinsey Quarterly<\/em> (2003); Simon-Kucher, Global Pricing Study (2011, 2014, 2025); Ramanujam and Tacke, Monetizing Innovation<\/em> (Wiley, 2016); List and Gallet, Environmental and Resource Economics<\/em> (2001); Murphy et al., Environmental and Resource Economics<\/em> (2005); Schmidt and Bijmolt, Journal of the Academy of Marketing Science<\/em> (2020); Lipscomb and Koford, Applied Economics Letters<\/em> (2011); Samuelson and Zeckhauser, “Status Quo Bias in Decision Making”, Journal of Risk and Uncertainty<\/em> (1988); OpenView, SaaS pricing survey (2017); Poyar, The 2025 State of B2B Monetization<\/em> (Growth Unhinged\/Tremont, 2025); Warren Buffett, interview with the Financial Crisis Inquiry Commission (2010).<\/p>\n\n\n\n <\/p>\n","protected":false},"excerpt":{"rendered":" Your first price isn’t just a price. It’s the beginning of your commercial model. Why pricing is the commercial decision technology companies find hardest to get right, and hardest to act on. The lever everyone agrees on The foundational evidence on price as a profit lever is old, and it is worth being clear about […]<\/p>\n","protected":false},"author":1,"featured_media":554,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[7],"tags":[],"class_list":["post-553","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-pricing"],"blocksy_meta":{"styles_descriptor":{"styles":{"desktop":"","tablet":"","mobile":""},"google_fonts":[],"version":7}},"_links":{"self":[{"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/posts\/553","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=553"}],"version-history":[{"count":1,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/posts\/553\/revisions"}],"predecessor-version":[{"id":555,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/posts\/553\/revisions\/555"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=\/wp\/v2\/media\/554"}],"wp:attachment":[{"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=553"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=553"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lsadvisory.co.uk\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=553"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}The lever everyone agrees on<\/h3>\n\n\n\n
\n
The mistake runs one way<\/h3>\n\n\n\n
Why you can’t simply ask<\/h3>\n\n\n\n
Why being close makes it harder<\/h3>\n\n\n\n
\n
Nobody owns it<\/h3>\n\n\n\n
The finding nobody acts on<\/h3>\n\n\n\n
For technology companies, the price is the model<\/h3>\n\n\n\n
\n
We ran this on ourselves<\/h3>\n\n\n\n
What this means<\/h3>\n\n\n\n
\n
About the data<\/h3>\n\n\n\n
\n\n\n\n<\/h3>\n\n\n\n