What to actually do when your SaaS vendor raises prices

A vendor raises your bill 18% at renewal. Your instinct is to cancel. That instinct is wrong more often than it is right, and acting on it costs teams real money every year. Here is a decision procedure that has held up across a few hundred tracked price changes.

The first thing to establish: is this a real increase or a packaging change?

Not every price rise shows up as a bigger number. Vendors have become fluent in moving the goalposts without touching the headline price, and these are the shapes that actually cost you money:

• The seat minimum. A per-seat tool goes from "billed on active users" to "billed on assigned users, minimum 5." Your bill rises while the sticker price stays identical. We record this as a limit change, not a price change, because that is what it is.
• The included allowance. Storage drops from 100 GB to 20 GB. Overage is then billed at a rate that makes the original tier unviable. Your effective unit cost rises sharply past a threshold you will not notice until the invoice.
• The annual reset. The price was always "$20/month if billed annually", and the annual commitment quietly resets to the new higher number at renewal. No single line item changed on the pricing page.
• Feature deprecation. A capability moves from your plan to an add-on. The price is unchanged and the tool is materially less useful at that price.

All four of these have appeared in the 21 changes recorded on this site, and only the first one looks like a price increase on the page. Before you decide anything, pull the vendor's pricing page from before your renewal and put it next to today's.

Work out your real unit cost before reacting

Price increases are usually defended per seat. Per-seat arguments fall apart when you look at what a seat actually costs over the year. Run this:

real annual cost = (subscription + overage + add-ons + migration labour) / people who actually use it

Three examples from real categories we track:

• A design tool at $45/seat/month for 12 licences, where 4 people use it weekly and 2 never log in. Real cost: $2,160/yr across 10 real users, or $216 per real user per year. The vendor's sticker price is $540 per person.
• An error-tracking service at $26/100k events where a single misbehaving cron job generates 80% of the volume. Removing that one job cuts the bill by 78% and costs nothing. Real cost after the fix: $58/mo.
• A project tool at $10/user/month where 9 of 12 seats belong to people who join one project a year. Replacing seats with a shared guest model costs $30/month and loses nothing.

Most teams discover they are paying a large premium for seats, events or storage they do not need, and that the price increase is not the largest line item on the invoice. Optimise before you renegotiate. A vendor negotiating against an optimised bill is a different conversation from a vendor negotiating against an inflated one.

Negotiate. Almost nobody does, and it works more often than expected.

For annual contracts above roughly $1,000, the price on the public page is a starting position rather than a fixed quantity. The things that move a vendor, roughly in order of leverage:

1. Multi-year commitment. Two or three years up front is the single most reliable lever, because it converts a renewal risk into banked revenue.
2. Being an early reference. Case study rights, a logo on the pricing page, a review. Worth real money to a vendor in the early trajectory, near-worthless to a large one.
3. Seat reduction at renewal. Reducing 20% of seats is frequently exchanged for a 10% discount, because it is painless for the vendor in the current fiscal quarter and painful for you to re-provision later.
4. Annual prepayment. Paid up front in exchange for the discount already offered. Free if you have the cash.

Ask explicitly, and put the ask in writing. "We are renewing on the 14th. The new number is $X. We can do $Y on a two-year term. Can you meet that?" is a stronger opening than describing how much the increase hurts your business.

When walking away is genuinely the right call

Some increases are not negotiable and not worth absorbing. Cancel or migrate when:

• The increase is more than 40% year over year. That is a repricing, not an adjustment, and it will keep happening.
• The vendor has removed a capability you rely on and replaced it with a materially more expensive add-on.
• You are under 10% of their customer base and your account manager has gone quiet. You have no leverage and no roadmap influence.
• The tool is one of many you hold for the same job. Redundancy is not waste; it is the thing that makes a price increase survivable.

And when you leave, take the data with you and check the export before you cancel. We have seen teams cancel, discover the export excludes history, and pay again to recover it.

Build the watch list, not the spreadsheet

The reason teams get surprised is that nobody owns this. Pricing is reviewed when it is invoiced, which is after the decision has effectively been made.

The durable fix is small: subscribe to changes for the vendors you actually depend on, and review the list once a quarter. Our full change feed and per-vendor pages exist for exactly this, and every entry links back to the source page so you can confirm it rather than trust it.