Here’s the dirty secret about pricing: it’s often made up.

There are many sciences that help teams make smarter pricing decisions, such as Van Westendorp, conjoint, win-loss price sensitivity, and competitive price-to-value comparisons. I’ve done them all.

But pricing is not purely a science. There’s no one equation that gives you the perfect number. As a pricing professional, I probably shouldn’t admit this—but pricing is as much art as science.

Buying and selling are human activities. People don’t pull out a calculator to decide if a number is “fair.” They decide subjectively. I’ve interviewed over five hundred software buyers. I can count on one hand the number who calculated a hard ROI when making their purchase decision.

At the end of the day, a price comes down to whether both sides agree to it. That’s it.

Pricing decisions are made with much more instinct than you might expect, even at the largest companies. Take OpenAI as an example. On a podcast, the Head of ChatGPT explained how they set the $20 monthly per-user price point:

Think about that. One of the fastest-growing companies in history. Worth half a trillion dollars. Nearly $12 billion in revenue already. And their price? An educated guess based on a Discord survey. Two years later, it’s still $20.

That number has anchored the whole market. Cursor, for example, copied it directly. Billions of dollars in revenue and industry pricing norms are downstream of one gut call. What if it had been $21? Or $22.50? That tiny change could have shifted profitability across the ecosystem.

I’ve seen this up close. A client’s data-supported price came in at $23/user. The CEO overruled it because “$19 just feels better.” And that was that.

Once you recognize the art of pricing, you see it everywhere. The NBA players union splits revenue 50/50 with the league. Why 50/50? The odds they each contribute exactly half the value are near zero. It could just as easily be 53/47. But 50/50 feels right.

The mistake isn’t acknowledging gut feel—it’s pretending it doesn’t matter. 

Two ways we can apply this to make better decisions. First is to meld science with psychology.

Use research to establish a range, then use intuition to choose within it. If your willingness-to-pay data says $78–$84, pick $79—it feels natural. If competitive analysis says three features should be add-ons, but you know one makes or breaks sales conversations, include it in the base package.

Gut feel breaks ties. Data sets the boundaries.

The second way we can apply this don’t assume competitors are any smarter with their pricing. We tend to admire larger rivals and imagine their prices come from deep logic. They’ve been super successful and have made tons of money, surely they know someone we don’t about pricing, the thinking goes.

The truth is the opposite. Usually, they’re winging it too—if anything, weighed down by fear of changing what already “works.”

Intuition and gut often rule even at the largest companies. Sometimes, these are accompanied by science, but often not. We end up with the pointing Spider-Man meme, with everyone afraid to move, assuming that the other party has it all figured out.

So don’t be afraid to challenge conventional wisdom. Research plus psychology can be your differentiator.

The best pricing isn’t purely scientific. It’s science, tempered by instinct.

The two biggest barriers to usage pricing models have traditionally been complexity and tooling. I have to wrap my head around tokens? How is a unit defined? The fear is that customers and sales reps will shut down the moment they hear about credit rollover policies.

That complexity barrier is fading fast. OpenAI and other AI-native companies are introducing infrastructure-style pricing to consumers and SMBs alike. And the truth is, usage pricing doesn’t have to be complicated.

Credits are consumers' way of making sense of value without constantly thinking about dollars and cents. Lovable, for example, charges you based on the type of action you take. It hasn’t slowed them down at all.

Even ClassPass—my gym membership—uses a credit system to book classes. If consumers are comfortable using credits to reserve pilates, your business users will be just fine.

Of course, credit systems do take more work to set up than unlimited or fixed-use models. You have to define policies, set values for different actions, and explain the model to customers. But buyers will accept that complexity—as long as you’re delivering clear value.

Welcome to the era of credits.

“Intuition is nothing more and nothing less than recognition.” — Herbert Simon (Nobel Laureate in Decision Theory)

Have questions or feedback? Just reply to this. I read every email.