- Crescendo
- Posts
- Cursor's Pricing Backlash
Cursor's Pricing Backlash
Dissecting a pricing rollout through the lens of best practices

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)
The PR backlash from Cursor’s recent pricing change could have been lessened by following a few best practices.
The team at Cursor rolled out pricing immediately, to all customers: we typically recommend a phased rollout to 4 segments in a specific order.
The communication at Cursor was too focused on costs and what the new pricing would do for the company, rather than how it would help users. We offer a different framework with a best-in-class example.
What Happened at Cursor?
A few weeks ago, the AI code writing company Cursor announced a pricing change. The public reaction to this pricing change has been both immense and negative. I don’t normally cover “current events” on this newsletter, but I think that Cursor can serve as an example in how we think about rolling out pricing changes at Crescendo.
Let’s dive in.
But first, let’s assume…
Before we go anywhere, we need to make some assumptions about Cursor since it was not Crescendo (surprise!) that helped them with their pricing.
Key assumptions we need:
Cursor’s bad PR is not balanced with significant business upside. There is always a chance that despite the public outcry, they are making gobs and gobs of money from sticky customers. I’ve been through negative public reactions before, but it helps when you’re doubling revenue in a month.
Cursor’s new pricing is actually misaligned to value. Here we have to take angry users at their word, which isn’t always the right way to think about it.
Cursor needed to “rip the Band-Aid off”. Cursor changed prices in a VERY risky fashion. We should assume that they needed to do that, rather than use a slower, less risky approach.
Cursor had good advisors. I don’t know whether Cursor performed their pricing analysis in house or if they had advisors. I also don’t know who their advisors were. I hesitate to play armchair quarterback when I was not in the room advising on pricing.
Price Metric Shift = Open Heart Surgery
So what did Cursor actually change. Fundamentally, the company changed price metric from a capacity-based metric to a usage-based one.
Old Cursor (specifically the Pro plan): 500 fast responses + unlimited slow ones
New Cursor: $20 of “usage credits” + overage fees
We at Crescendo like to say that performing a price metric shift is like doing open heart surgery. There are some times when you absolutely must do it, but it’s risky and painful.
Best Practices for Performing a Metric Shift
Ensure the metric aligns with willingness-to-pay. For more on that, see these posts about how to pick the right metric. For example, it’s not clear to me that users of Cursor get more value out of the tool when they query with larger queries (as opposed to just more queries).
Build a Winners & Losers model to ensure no users are getting wild price hikes and/or cannibalization. If some users will be getting wild price hikes, you’ll need to proactively reach out to those users with discounts and grace periods.
Don’t rationalize a price increase with costs. Cost-plus pricing is almost always misaligned to value. Find a metric that tracks with willingness-to-pay, not costs.
Screenshot of Cursor’s blog
Ensure the metric is easily understood. Cursor switched to a “token” based model, very popular these days with agentic products. Unfortunately, tokens are often poorly communicable in the eyes of customers, which is why we rarely recommend them.
Rollout Best Practices
Even if Cursor was right about their new price metric (aka, it does align with value and the limits were well placed), the manner in which they rolled out the new prices could be improved.
The first way to de-risk a rollout is to do a phased rollout across 4 customer segments, in a specific order that minimizes backlash.
The second de-risking strategy involves clear, value-focused communication. We’ll cover these two strategies in turn.
Ideal Order of Price Rollouts
We like to preach that every price rollout should address the following 4 customer segments, in this order, but the speed in which you go through them is up to the company’s risk tolerance.
New Customers. New customers are the easiest to rollout new pricing to because they haven’t experienced anything else and will not be anchored to a previous model.
Upsells. Customers who self select as needing to upgrade their plan are the least likely to churn. This is strategy is commonly found in the telecommunications industry; as soon you need to upgrade your plan, you forfeit your legacy prices. The same can be said for accessing new features. Customers who want access to the latest and greatest features should be forced to purchase on the new model.
Legacy. Customers on legacy prices are often where the largest pricing opportunity exists. The problem is that these customers also will receive the highest price increase, by definition. We have found that the best way to handle these customers is through an open negotiation.
Listen customer, our list prices are a full 10x what you’re paying today. I’m not expecting you to go to list price all in one go, but we need to work something out here…
Brand New. Customers that bought the day before the price change may feel like a rug is being pulled out from under their feet. These customers are the last to migrate to a new pricing plan, and you may need to uphold their legacy pricing for some time.
Clear, Value-Focused Communication
Lastly we need to make sure we’ve communicated WHY we are raising prices to our customers. Here are some good reasons to raise prices:
It will enable us to invest in product development
We’ll kick out the riff raff so the product performs better / is more exclusive
The price is now more aligned to the value you receive
Here are some bad reasons to raise prices:
We (read: our investors) want more money
We have high costs
The existing structure wasn’t good for us (e.g. unpredictable)
That said, almost everyone mentions those last 3 instead of the first 3. Let’s dissect Cursor’s price increase notification
We're also happy to report we're rolling out changes this week to make our Pro plan more generous. By default, the Pro plan will move from request limits to compute limits; all users will get at least $20 of model inference at API prices per month. Concurrent with this change, we're rolling out unlimited access to the 'Auto' model and lifting all limits on tool calls. Existing users can choose to stay with the "500 request limit" method if they prefer.
Overall, not bad. Pro is more generous (sounds like a price decrease). Unlimited access in “Auto” mode. “Lifting limits”. What could be wrong about this?
2 things:
It doesn’t explain the key issue - moving from “request limits to compute limits”
It’s simply inaccurate. The Pro plan became far more expensive, not more generous.
The team then released the following statement, explaining why they moved to compute limits.
New models can spend more tokens per request on longer-horizon tasks. Though most users' costs have stayed fairly constant, the hardest requests cost an order of magnitude more than simple ones. API-based pricing is the best way to reflect that.
Notice anything? I bolded it for you…It’s cost-plus pricing!! Bad bad bad! Where is the value to users? Something like “your bill will now be more reflective of how you use Cursor” or “those who get the most value out of Cursor will now be paying the most”. Even something like “users will now have more control over their budget” would be value aligned and true (I think).
A World Class Price Increase Notification - Netflix
Let’s show you what great looks like. Netflix is now a master at raising prices, ever since their 2011 debacle when they lost 37% of their stock price and 600k subscribers, they have become true masters at raising prices.
A good price change notification has the following elements:
Thanks for being a customer
Past value delivered
Future value promised
New price (not % change)
Link to broader FAQ
Invitation to continue partnership
Check out this Netflix notification from 2022. Textbook perfect.

Odds and Ends and Making Amends
To their credit, Cursor recovered from their price change well (at least from a PR perspective). Part of that was due to their swift explanation published on their blog. But most likely, it was due to the credits they offered users who had unexpected price hikes. Obviously, they could have avoided this by identifying those users ahead of time, but at least they offered refunds eventually.
There are so many things that go into rolling out a new pricing model, and to be honest, we only covered the public facing ones here. Don’t forget about the rest of the teams behind the scenes: legal, ops, CX, and billing. They also need to be brought into the fold early.
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.
Reply