On 11 March 2026 Clay replaced its single credit with two meters: Data Credits for third-party data, with marketplace prices cut 50% to 90% and no charge for failed lookups, and Actions for the platform work Clay does. Its heaviest customers had been sold credits at or below breakeven. The lesson: price pass-through costs near cost and charge for your own value separately.
Key takeaways
- If one unit is paying for both your costs and your value, split it.
- Clay charges near cost for resold data and charges separately for the platform work it does.
- No result, no fee: Clay stopped charging for failed data lookups.
- Clay said it expected about 10% lower revenue from the change, and made it anyway.
- Publishing the reasoning turned a risky change into a credibility win.
In The Subscription Playbook I wrote about Leadfeeder, a company that sold leads on a subscription.
It charged $63 a month for up to 100 leads, rising to $599 a month for 3,001 to 5,000. I liked it because the price grew with the customer.
Clay sells into the same world: go-to-market data for sales teams. And for years it did something similar, with one credit that bought everything.
Then, in March 2026, it tore that model up.
What Was Wrong With Clay's Old Credit System?
Clay is a go-to-market data and automation platform. Customers bought credits and spent them on data from more than 150 vendors, without signing separate contracts with each one.
Simple. Convenient. And quietly broken.
One credit was doing two jobs:
- Paying for the data Clay bought from vendors (a cost Clay passed through).
- Paying for Clay's platform: the enrichments, AI tasks and automations that make Clay valuable.
Clay's internal memo, which it later published, put it bluntly. It had priced itself "like a data product" when it had become "an end-to-end GTM platform."
The memo admitted two more things:
- Its most-scaled customers were being sold credits "at or below breakeven."
- Customers only upgraded when they ran out of credits. So upgrades were about running out, not about getting more value.
What Did Clay Change on 11 March 2026?
Clay called it "the most significant change to Clay's pricing since we launched."
| Before (to March 2026) | After (from 11 March 2026) |
|---|---|
| One credit for data and platform work | Two meters: Data Credits and Actions |
| Platform markup baked into data prices | Marketplace data cost cut 50% to 90% |
| Failed lookups (no data found) charged | Failed lookups free |
| Starter, Explorer and Pro plans | Launch and Growth plans |
| Top-up credits at a 50% premium | Top-up premium cut to 30% |
Actions are the new unit for the work Clay does: enrichments, AI tasks, API calls and pushes to other tools. They default to 4 to 5 Actions per Data Credit, start below a cent each, and Clay says 90% of customers never exceed their allocation.
Token-heavy AI models are passed through "at cost with no markup." Customers who bring their own data provider keys skip Data Credits but still use Actions.
The Numbers: What Did the Split Cost Clay?
Clay said it expected the repricing to cut revenue by about 10%. It also admitted its 2022 Pro tier had been underpriced and run at a loss.
Legacy customers had until 10 April 2026 to switch.
That's a company choosing a smaller, healthier number over a bigger, shakier one. Not many businesses are brave enough to do that in public.
Find the metric that grows with your customers
Score candidate value metrics, separate cost from value, and pick the one customers will accept.
Why Did Clay Split Its Credits?
Because blended units hide where your money comes from.
In the book I said your usage metric "must be value-based, scalable, flexible, feasible, and predictable." A single credit that mixes cost and value fails the first test. Customers can't tell whether they're paying for data (which they could buy elsewhere) or for Clay (which they can't).
Splitting the meters solves three problems at once:
- Competitive data. With the markup gone, Clay's data looks fair next to buying it directly. That should pull more data purchases through its marketplace.
- Paid platform value. Clay earns on its software even when customers bring their own data. Before, those customers paid almost nothing.
- Healthier upgrades. Customers upgrade because they're doing more work in Clay, not just because they ran out of credits.
That's exactly what a good value metric should do.
My take
The same trap exists in plenty of small businesses. Picture a builder who marks up materials heavily and charges a modest labor rate. Then a client buys their own tiles, and the builder's margin disappears. Clay's fix is the same one I'd give the builder: charge close to cost for the stuff anyone can buy, and charge properly for the expertise only you have.
How Does This Fit the Book's Frameworks?
Two connections stand out.
Gross margin. In the moats chapter I said high gross margin is one of the things buyers and investors pay most for. Selling credits at breakeven to your biggest customers is a slow leak in that moat. Clay plugged it. Our gross margin guide shows how to check your own.
Trust and churn. "No result, no fee" is a promise customers remember. Charging for a lookup that found nothing always felt unfair. Removing it costs Clay a little revenue and buys a lot of goodwill.
What You Can Steal From Clay
- Separate pass-through costs from your own value. Materials, third-party fees, licenses or freight on one line. Your expertise and time on another. Show both on the invoice.
- Price the commodity near cost. If customers can buy it elsewhere, a fat markup just invites them to.
- Charge properly for what only you do. That's where your margin should come from.
- Stop charging for failures. No result, no fee. It builds trust fast.
- Publish the reasoning behind a price change. Clay's open memo turned a risky change into a credibility win. Our guide on how to raise prices covers how to write that message.
- Give a switching window. Clay gave legacy customers about a month. For a bigger change, I'd give longer.
When Is Unbundling the Wrong Move?
When customers value simplicity more than fairness.
Two meters are harder to explain than one. If your customers are small, busy and buy on gut feel, a single all-in price may sell better even if it's less precise. Clay's buyers are sales and operations teams who live in spreadsheets. They can handle two meters. A café's customers can't.
Test it before you commit. The Value Metric Finder scores candidate metrics, and the Price Increase Simulator shows what a pricing change does to revenue. For more on credit design, read our guide to credit-based pricing or compare current plans in the Clay pricing teardown.
Frequently asked questions
How much does Clay cost in 2026?
As of October 2026, Clay's free plan includes 100 Data Credits and 500 Actions a month. Launch starts at $185 a month ($167 billed annually) with 2,500 Data Credits and 15,000 Actions. Growth starts at $495 a month ($446 annually) with 6,000 Data Credits and 40,000 Actions plus CRM integrations. Enterprise is custom. Check clay.com/pricing for current figures.
What is the difference between Data Credits and Actions?
Data Credits pay for information bought from third-party providers through Clay's marketplace, such as an email address or company details. They start from about $0.05 each. Actions pay for the work Clay's platform does: enrichments, AI tasks, API calls and pushing data to other tools. Actions start below a cent each, and Clay says 90% of customers never exceed their allocation.
Can you bring your own data provider to Clay?
Yes. If you connect your own API keys for a data provider, you skip Data Credits for that provider but still use Actions for the work Clay does. That's the point of the split: Clay earns on its software even when customers don't buy data through its marketplace.
Did existing Clay customers have to switch?
Legacy customers were given a window, until 10 April 2026, to switch to the new plans. Grandfathering and a clear switching window are standard good practice when you change the unit customers pay for, because it gives them time to compare old and new bills before anything changes.
Why would a company cut prices on purpose?
Because the old price was in the wrong place. Clay's marked-up data made it look expensive next to buying data directly, while its platform was underpriced. Cutting data prices and charging for platform work makes Clay competitive on the commodity and profitable on what's unique. It's a short-term revenue hit for a healthier long-term model.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- Clay, The thinking behind our new pricing: our internal memo (Mar 2026)
- Clay, Introducing Clay's new pricing (11 Mar 2026)
- Clay, pricing page (checked 27 Oct 2026)
- Landbase, Clay Pricing 2026: Plans, Costs, and What You Pay (secondary, 2026)
- Kareem Amin (CEO), LinkedIn pricing update (Mar 2026)
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.