TL;DR
Asking customers about willingness to pay produces inaccurate results because software is an experienced good that customers can’t properly evaluate beforehand.
To read the original Forbes article, click here.
Forbes: To Land On The Right Pricing, Software Leaders Must Look Not Just Outward, But Inward As Well
Many software companies rely too heavily on simplistic outward-facing strategies to determine pricing, such as basing prices on competitors or asking customers about willingness to pay.
As I’ve previously written, basing pricing solely on competitors is flawed because what works for one company may not work for another. In addition, competitors likely based their own pricing on faulty assumptions.
One of the more egregious assumptions is that directly asking customers what they are willing to pay results in accurate answers.
Willingness to Pay is Complicated
Software is an experienced good: customers don’t know what they are willing to pay until they use it and fully deploy it into their workflows for maximum return.
Joel Spolsky, the founder of Stack Overflow & Trello, figured out the flaws of the willingness to pay concept years ago: You can get “radically different answers” from one day to the next “when you ask people how much they’re willing to pay for something.”
Consider two of the examples he used:
- Ask “some aircraft designers how much they would pay and they sort of think $99 would be a maximum price, even though aircraft designers regularly use software that costs on the order of $3000 a month without being aware of it, because someone else does the purchasing.”
- Give focus group participants Starbucks drinks during a break, and a “side conversation about paying $4 for a cup of coffee” might ensue, putting everyone in a “real frugal mood when you ask them about their willingness to pay.”
Van Westendorp Price Sensitivity Meter (PSM) Inaccuracies
Van Westendorp‘s Price Sensitivity Meter (PSM) has been gaining popularity in the blogosphere, particularly among product managers, despite concerns about the approach that have existed for decades. It’s even been exposed as a feature in some survey platforms.
A critique published in the trade press two decades ago put the model’s popularity down to two reasons:
- The “ease and low cost of asking four simple questions.”
- The “intuitive appeal of the questions, which sound to marketers like the things they want to know.”
That same critique held the Van Westendorp Price Sensitivity Meter to be “used far more widely, and its results taken more literally, than they should be.” This technique introduces powerful biases that skew results. That’s because each of the 4 questions that make up the PSM are all variants of “How much would you pay for this?”
It also noted that the questions invite “lowball answers” along with the observation that “a surprisingly high number of respondents give four answers that don’t show the expected relationships (e.g., their ‘too inexpensive’ price is higher than their ‘inexpensive’ price)…as many as 20% of respondents may give inconsistent figures.”
It is shocking, and concerning, how many B2B software companies use (or are planning to use) Van Westendorp Price Sensitivity Meter (PSM) to develop their pricing.
Peer-reviewed work has since tested what that criticism predicted, and it lands alongside the broader case that willingness-to-pay surveys fail B2B software companies. When the meter’s four intersections are re-derived across resampled sets of the same respondents, the band they produce moves. The acceptable range it returns can span several editions’ worth of price, which is another way of saying it cannot tell you what to charge. The intersections are not stable points at all. They are wide overlapping regions that only look like points because you drew them once.
The fair reading is not that the instrument is worthless. The same peer-reviewed work argues it gives useful directional guidance once those raw intersections are replaced with proper statistical modeling, and it demonstrates the narrower band that produces. But notice what that fix does and does not do. Modeling the responses more carefully tightens the estimate without changing what was measured, which is still what people said about a price they were not being asked to pay. What replaces it is evidence from real transactions, then continuous monetization to move deliberately from where those transactions sit. A more precise estimate of a stated intention is still an estimate of a stated intention.
Is Your Pricing Built on Survey Responses Nobody Actually Meant?
Van Westendorp’s four price-point questions reveal stated preference, not purchase behavior — and your licensing, packaging, and pricing decisions deserve a harder test than hypothetical answers.
Willingness to Pay: The Biggest Contributor
In the game of oversimplifying the problem with techniques like Van Westendorp PSM, these types of approaches ignore one of the largest contributors of them all:
The salesperson’s impact on the prospect’s willingness to pay.
To understand this, you must get your hands dirty. By looking inward, not just outward, you will better understand the market realities contained right inside of your own invoice transaction and usage data.
You can read our full article on Forbes: To Land On The Right Pricing, Software Leaders Must Look Not Just Outward, But Inward As Well
If you are weighing a survey because you have no transaction history to read yet, that is the case worth talking through rather than surveying around. Describe your situation to a pricing expert and a reply will come back on what evidence you actually have available, and what it can and cannot tell you about price.
