• Crescendo
  • Posts
  • Problem: Your Pricing Isn't Predictable

Problem: Your Pricing Isn't Predictable

How to fix price predictability without messing up your price metric

Welcome back to Crescendo Insights, where we provide a bite-sized piece of monetization strategy each week.

If you’re not a subscriber yet, hit the button below to keep getting these emails delivered to your inbox.

The BLUF (Bottom Line Up Front)

  1. An AI company misinterpreted customer feedback and made a big mistake

  2. Don’t confuse price metric and price structure

  3. 5 ways to make your price structure more predictable

Case Study: AI Pricing

Crescendo recently worked with an AI unicorn to build out their first pricing structure. As is the case with many VC-backed companies, monetization had been an afterthought: after product, after brand, definitely after hiring top talent. We were brought in to help craft a strategy that made sense.

Like most AI companies, this company was also fixated on cost. AI costs real money, and while the costs aren’t as extreme as with non-software products1 , most AI founders nonetheless care a lot about costs.

If you’re curious about our thoughts on cost and monetizing AI, check out this post: How to Price Generative AI

TL;DR - we don’t like to worry about cost. But back to the story.

Through our typical methods of assessing price metric (Theory, Data Analysis, Customer Interviews, and Quant Studies), we had determined that usage-based pricing was the right fit. A user-based price metric would under-monetize the power users and over-monetize the lite users. Pricing per user would inevitably lead to poor revenue retention and churn.

Poor Market Feedback

We made our recommendation, the company launched the new pricing, made tons of money, and we all went home happy2 .

Just kidding…

After soft-launching the pricing3 , customers started complaining.

Ever since you switched to usage-based pricing, my bill has been jumping all over the place! I can’t handle it; go back to charging for users plz.

A cranky customer

They also brought on a new marketing executive from a large public company, who said something like

Large enterprise customers prefer seat-based pricing. Usage-based pricing is too unpredictable.

A curmudgeon

This caused our intrepid company to about-face:

  • They reverted to user-based pricing

  • They did not increase rates (as they initially planned)

  • Their growth continued to go up and to the right like a rocket ship

  • They raised a big round of funding from famous investors with a stratospheric valuation

Big success right? We don’t think so.

Don’t solve structural problems with metric

Remember this graph from last week?

One of the most common mistakes that companies make is not typing the diagnosis to the right prescription: most commonly this means they solve a structure problem with metric. What do I mean?

Diagnosing a metric problem vs. a structure problem

Metric is how your price scales with value. It is your primary price differentiator, enabling your company to charge small customers a different price from large ones. If you have one of the problems below, you have a price metric problem:

  • Low price differentiation (between largest and smallest customers)

  • Poor net dollar or net revenue retention, especially when coupled with great product engagement

  • Similarly - great NPS + poor NDR

  • Failure to grow customer footprint, especially as customers grow

Structure is how the metric aligns with volume. It’s the shape of the curve, rather than the quantity sold. If you have one of the problems below, you have a price structure problem.

  • Customers complain about price consistency or predictability

  • Low % of recurring revenue (aka low revenue predictability)

  • Large discretionary discounts, especially at high volumes

  • High transaction costs: unnaturally long sales cycle, constant customer conversations, poor sales / billing infrastructure

How to make your price more “predictable” without touching price metric

When customers are complaining about price predictability, that is a structural problem, not a metric problem. Users are not “more predictable” than usage. How do you make a usage-based (or any metric) more predictable?

Widen the volume tiers

With wider tiers, customers are less likely to jump to a new tier as their usage fluctuates.

Flatten the volume tiers

Flat tiers (e.g. $50,000 if you are between 100 and 200 users) is a more consistent structure than variable tiers (e.g. $500 per user if you are between 100 and 200 users).

Rely on a minimum or flat fee

Raising the minimum fee will make less of your price depend on the quantity. Note - this will also lower your price differentiation, which you may not want to do.

Use price caps; not overage fees

Remember how we all loved getting overage fees in the early 2000’s when we would text someone overseas4 ? Yeah I don’t either. Remove overage fees (by instituting a cap, grace period, or buffer) to make your price more predictable.

On the left, your price stays the same if you go above your volume tier cap (presumably renegotiated at renewal). On the right, we charge an overage fee when you go over your tier limit.

Let the customer price their own risk

My favorite method. Utilize a model where customers can pre-commit more quantity in exchange for a more predictable price.

On the left, there are 3 plans to choose from. If you want more predictable prices, you can raise the flat fee. On the right, you have no such choice - your plan is determined by your volume.

What happens when you mess this up

The problem with fixing structure problems with metric is that metric is WAY harder to change later on, and tends to lead to a myriad of stalled growth issues.

  1. You won’t be able to raise rates to true willingness-to-pay. When your price does not align with value, you’ll be far away from optimal price level.

  2. You’ll have very high downsell risk. Customers will try to optimize their price by reducing the sub-optimal price metric, causing net revenue loss.

  3. Customers will be frustrated. When price doesn’t track well with value, customers (and your reputation) will be in jeopardy as you try to justify your prices.

We will see what happens with our AI company. It is still early days, but both of us at Crescendo have seen what can go wrong when companies get their price metric wrong.

Appendix: Metric vs. Structure

The Elements of Price Metric

The Elements of Price Structure

Get in touch

Crescendo works with medium-sized software companies to improve their pricing, packaging, and promotion strategies. If you’d like to book a quick consult, reach out at [email protected] or schedule time via the button below.

1  Most of the products out there

2  With fatter wallets than before

3  Something we always recommend

4  Or remember when we would get yelled at for doing that?

Reply

or to participate.