Your pricing page is a hypothesis
Most B2B SaaS pricing was set by the founders, early, under pressure, with a competitor's pricing page open in another tab. Then the company grew, the product got better, the customer base changed, and the pricing stayed exactly where it was. What started as a guess quietly became policy.
We found $2M in annual pricing upside for one client, and the uncomfortable part is where we found it: in data they already had. It sat in their systems for two years. Nobody had looked, because pricing belonged to nobody. Sales wanted it lower, finance wanted it higher, and the product team had usage data that could have settled the argument but was never invited to it.
The join nobody runs
Here is the actual work, and it's less glamorous than "pricing strategy" sounds. You take billing data (who pays what, on which plan, since when) and product usage data (who does what, how often, at what volume), and you join them on the account. Most companies have never run this join. The two datasets live in different tools, owned by different teams, and each side assumes the other has it handled.
The first time you look at the joined table, a few patterns show up almost every time:
- A handful of your heaviest users are on your oldest, cheapest plans. They adopted early, their usage grew tenfold, and their bill never noticed.
- The features you charge a premium for aren't the ones your best accounts actually lean on. The thing they'd genuinely pay more for is bundled somewhere as an afterthought.
- Discounts cluster around whoever negotiated, not around any measure of value. Two nearly identical accounts pay meaningfully different prices because one of them asked.
None of this tells you the right price. It tells you where your current pricing and your delivered value have drifted apart, which is where the money is.
Test it like you'd test anything else
The reason pricing stays untouched for years is fear, and the fear is reasonable. Change prices carelessly and you churn the customers you most want to keep. The answer is to stop treating pricing as one big irreversible decision and start treating it as a stream of small testable ones.
New signups are the safest place to start. Existing customers keep their terms; new cohorts see the new structure. You learn what the market bears without breaking a single promise. Grandfathering isn't a cop-out, it's what makes the experiment ethically and commercially safe to run.
For the client above, we built an A/B testing framework so pricing changes ran as controlled experiments on new cohorts instead of company-wide bets. A pricing engine followed, recommending plans from usage rather than from negotiation stamina. Monthly revenue went from $420K to $755K in a year. The full story, including what the diagnostic looked like, is in the case study.
Where to start this quarter
Run the join. Billing to usage, by account. If your data team can't produce that table inside a week, that's a finding in itself, and arguably a more urgent one than pricing.
Then sort by usage descending and read the top 50 rows with someone from finance in the room. In our experience the first pricing conversation worth having falls out of that meeting on its own.
If you'd rather have someone who has done this before sitting in that meeting, we should talk. Revenue analytics is one of our six core capabilities, and pricing work is where it pays back fastest.