Your price change will be judged in the twelve weeks before you announce it, not the day you do. Which accounts renew inside the notice window, and what are your reps paid to close them? Both answers exist. Neither has been asked out loud.
Three weeks after the announcement, the renewal conversations go quiet. The new price was modeled, approved and shipped on schedule, and the accounts that matter most are asking for a call instead of signing. Nothing on the invoice changed. A set of internal decisions that should have happened before the announcement did not, and the renewal calendar is where they surface.
Every vendor your customer buys from is already raising prices. Vertice's own index puts SaaS inflation at 13.2% for 2026, with per-employee spend reaching $9,100 at the end of 2025, up from $8,700 a year earlier. Your change is not arriving in a quiet market, and the accounts you worry about have already sat through somebody else's increase this year. I have watched three of these go wrong, and each post-mortem named the model. The model was fine.
The Model List Is Already Solved
Every pricing model in use is a variation on four decisions: what you count, how you package it, where the fences sit, and what the entry price is. Per-seat counts people. Usage-based counts consumption. Tiered packaging bundles features behind fences. Freemium moves the entry price to zero. There is no fifth category, and if you are still choosing between them you are at an earlier stage than this article.
We have written the taxonomy already. If you need the enumeration, read penetration pricing and come back. What matters here is what the taxonomy cannot tell you. A pricing model describes a steady state, and your business is never in one. The moment you move from the old state to the new one you are running a change, and changes have a calendar, an affected population and internal incentives designed for the previous price. A model choice with no migration plan is not a strategy. It is a preference.
Decide What You Are Actually Changing
Three different changes get called "pricing" in the same meeting, and they carry completely different blast radii.
Price level, packaging, or metric: three changes, three blast radii
A price level change keeps the unit identical and moves the number. It is the bluntest and in some ways the cleanest, because nothing about the customer's daily use of the product shifts. The exposure concentrates in one place: every account renewing after the effective date.
A packaging change moves the fences. Features get regrouped, limits redrawn, and the price of a bundle may not change at all. Fewer accounts are affected, but unevenly. A customer who fit comfortably inside a tier last quarter may now be one seat over a limit they never knew existed. Those accounts do not call on announcement day. They call at renewal.
A metric change moves what you count, and it is the slowest and most dangerous of the three because it invalidates your own forecasting. Stop counting seats, start counting events, and your historical revenue per account stops predicting anything. Your expansion model, churn model and capacity plan all change meaning on the same day.
Change | What the customer experiences | Blast radius | What breaks first |
|---|---|---|---|
Price level | A different number on the same invoice line | Every account renewing after the effective date | The discount conversation |
Packaging | A different bundle, sometimes at the same total | Accounts sitting near a fence | Segment logic |
Metric | A different quantity being counted | Every account, plus your forecast | Your revenue baseline |
That last column is the one people skip. A price-level change does not break your product. It breaks the conversation your reps have been having for three years, because the fastest way to close a renewal under a new number is a concession.
The question that sorts them
Before any of this becomes a plan, answer one question in writing: who pays more at the next renewal, and do they know why?
If you cannot name the accounts, you are not ready to announce. If you can name them but cannot state a reason each one would accept, you are ready to announce something you will spend two quarters explaining.
This is where your segment work pays off. Tiers are a segmentation instrument before they are a billing instrument, and if you have not decided which segments you are willing to lose, the change decides for you. The mechanics are in segmentation, targeting and positioning; you need one output from it, a named list of accounts with a reason for each. If the wider plan is still open, our product marketing strategy covers where pricing sits beside positioning and launch.
The Grandfathering Decision
The guides ranking highest for this query stop at which model to pick. Sales compensation is missing from them, and none of them runs a change end to end. Grandfathering is the partial exception: one guide gives it a single sentence, and another a paragraph recommending a twelve-month window with sixty days' notice. That is real coverage and it deserves the credit.
What none of them does is size the cohort before promising anything, or run the arithmetic that tells you which version you can afford.
Time-boxed versus permanent: the retention math on both
Permanent grandfathering feels generous and behaves like a liability. You keep every account and the goodwill, and you permanently lower the contract value of a cohort that grows with your base. Worse, permanence is discoverable: once a prospect learns your website price is negotiable if you wait, your rate card stops being a rate card.
Time-boxed grandfathering moves the question from "do I lose them" to "when". That sounds softer, and in a sense it is, but it hands you the one thing a hard cut never gives: a date you can prepare for. The window becomes the product. Make it long enough to cover a full renewal cycle for your median contract, defined in months tied to renewal dates.
Neither is free, and the honest comparison is retained revenue. Run it with your own placeholders, because the worked examples circulating online are illustrations rather than evidence. Take a cohort of N accounts at your average contract value, annualize it, then apply three scenarios: the fraction that stays under permanent grandfathering, under a time-box, and under a hard cut. Add two years, not one. Permanent grandfathering almost always wins year one and loses year two.
The grandfathering decision takes longer than the model decision every time. Not because the math is hard. Because the math keeps disagreeing with what you want to believe about how you treat customers.
How to size the cohort before you promise anything
Three cuts, all of which you can pull this week.
Renewal dates inside the window. If a third of your base renews in the same quarter, that quarter is your entire project, and nothing else on the calendar gets to compete with it.
Usage distribution. Plot every account against the limit that is changing. You are looking for the cluster sitting just under the line, because that is where the surprise lives. Accounts at ninety-five percent of a limit will read your announcement as a bill.
Discount depth. Who is already off list price, and by how much. A price increase applied to an account that is thirty percent discounted is not a price increase. It is a negotiation, conducted by whoever is most afraid of losing the logo.
Once those three sit on one page, you can promise something.
Modeling Churn Before You Commit
You cannot test a price change the way you test a landing page. There is no clean control group when the announcement is public and every customer receives it at once.
Seventy-nine percent of IT leaders met a price increase at renewal in the past twelve months, and only 38% treat renewals as a cost-reduction exercise. Those figures come from Zylo's 2026 SaaS Management Index, vendor research run on its own customer base, so read them as directional. The asymmetry is what matters: your customer has been through this already, and most of them still walk in unprepared.
The three inputs you already have
Renewal dates. Start here, with a reframe that changes the calendar. Most SaaS contracts renew automatically, commonly cited at roughly eighty percent of agreements, and the modal non-renewal notice window is sixty days. Those figures are a contract-management vendor compiling SaaStr and Gartner material, so treat them as widely reported benchmarks rather than verified fact. The useful part survives the caveat: in an auto-renewing contract the notice deadline is the real constraint, not the announcement date. A team that plans around the announce date will miss the window and leave a slice of the cohort on the old terms by default.
Usage distribution and discount depth. The same cut you used for cohort sizing tells you who is under-billed today, which is where the revenue actually is. Discount depth then corrects it: modeled lift from a list-price increase is fiction if half your base is off list.
Reading the outcome as a KPI series
Do not judge the change on one revenue number. A single month of bookings can look fine while the leading indicators say the cohort is deteriorating. Watch the series: renewal rate by cohort, expansion among accounts that crossed the new fence, support tickets tagged to billing, and the exception rate your reps are granting. That last one is the earliest honest signal, and it is the one most teams never chart.
The discipline is the same one that makes any metric useful: it has to change a decision. A metric that cannot tell you to stop, reprice again or grandfather a segment belongs in a dashboard. We wrote about that test in the marketing KPIs that change a decision; the reporting trade-offs are in marketing analytics tools.
The Communication Sequence
The order in which people hear about a price change determines how much of it you spend the next quarter defending.
Who hears it first
Not the customer, and not sales alone. The order is: the people who take the call, then the customer, then the market.
Customer success and support get the briefed script and the exception policy before sales does. Sales gets it before the customer does, with one instruction: the exception path is real, documented, and not theirs to improvise. Anything less and your announcement is followed by a wave of discounts that quietly undo the change.
Then the customer, in a channel you control, on a date you chose. Then the market: your pricing page, your sales deck, and the first prospect who asks whether the old price is still available.
What to write in the notice, in order
Five things, in this sequence, because the reader's questions arrive in this sequence.
- What changes. The number, not the philosophy behind it.
- When it changes. The date the new amount first appears on an invoice, not the date you decided.
- What stays the same. Their plan, their limits, their support arrangement. Name them, because silence reads as "everything".
- What they can do about it. Real options: annual prepay, a different tier, a scheduled call. Options are not exceptions, and no rep should be allowed to blur that line.
- Who to talk to. A named person, not a support queue.
One migration timeline, run backward from the notice deadline rather than forward from the announcement:
Window | What has to be finished | Owner |
|---|---|---|
T-120 days | Cohort sized: renewal dates, usage distribution, discount depth | Finance, RevOps |
T-90 days | Three scenarios modeled; grandfathering version chosen; exception policy written | CEO, Finance |
T-75 days | Comp plan rewritten and signed; CS and support scripts drafted | Sales ops |
T-60 days | Notice goes out, the contractual deadline in auto-renewing agreements | Legal, CS |
T-30 days | Renewal conversations begin with the new number visible | CS |
T-0 | New price applies to new and renewing business | Billing |
T+30 days | Exception log reviewed; first read of the KPI series | Finance |
T+90 days | Cohort retention compared against the model; next cohort decision | Leadership |
If your announcement date sits inside the T-60 line, you are already late. Move the effective date rather than compressing the sequence.
Resetting Sales Comp So the Change Survives
This section decides whether any of the above survives contact with a quota, and it has the least published guidance behind it. I am reasoning from incentives rather than citing a study, because I could not find a credible one and I would rather say that than invent a number.
A rep paid a percentage of booked contract value has no reason to care what price you announced. Their job is to close the renewal, and the cheapest way to close one under a new number is a concession. If the commission plan still pays on the old price, you have built a machine that produces exceptions at the rate your customers ask for them.
Three changes fix it, and all three are unglamorous.
Pay on the new price from the announcement date, not the effective date. Otherwise every renewal closed in the gap is worth less to the rep than it is to you, and they will price accordingly.
Split credit across the transition. A rep who closes an account at the new price during the window should not lose it to a colleague who waited. Transition periods reward whoever moves last unless the plan says otherwise.
Make the exception path expensive. Not forbidden. Documented, approved above the rep, and visible in a log reviewed monthly. Once exceptions cost something politically, how many of them there are tells you something true about your price rather than about your reps.
One more thing. If quota is set on last year's average contract value, the price increase does the rep's job for them and your best team this year is the one that did the least. Reset quota on the new price when you reset the plan.
The Question to Put to Your Leadership Team
On a day when the price is not on the agenda, ask one thing: if we announced a fifteen percent increase tomorrow, could you name every account it touches, say which ones renew inside the notice window, and tell me what our reps are paid for closing them? If any of those three is a shrug, you have a sequence problem, and it is cheaper to fix now than after the emails go out.
Frequently Asked Questions
How often should a SaaS company change pricing?
Less often than your product changes, and more often than most companies actually do. The binding constraint is not customer tolerance but your capacity to run the change well: a cohort sized, a comp plan reset, a notice sequence executed. Without that, you cannot run two changes a year regardless of what the market will bear. The exception is an emergency, a cost structure that moved under you or a metric that no longer reflects the product. Then the timeline compresses whether you are ready or not.
Should existing customers be grandfathered?
Sometimes, and almost never permanently. Grandfathering is a retention instrument with a price, and the price is a permanently lower contract value for a cohort that grows. Time-box it against renewal dates, size the cohort before announcing, and be explicit about what happens when the window closes. Permanent grandfathering is defensible in exactly one situation: a small, closed set of accounts you cannot afford to lose and cannot afford to negotiate with.
How do you test a price change without losing customers?
You do not test a public price change on your existing base, and anyone who says they did is describing a different kind of test. What you can test is new business: quote the new price to a cohort of new customers while the old price runs, and read revenue per trial rather than conversion alone, because a price that halves conversion and doubles revenue is not obviously worse. For the installed base you model instead of test.
Name the Accounts Before You Announce
Build the cohort list before the model is signed off: every account the change touches, which of them renew inside the notice window, and what your reps are paid for closing them. Allable keeps that list, the pricing decision and the messaging in one project, so the three stay in step through the announcement. The 7-day trial is at studio.allable.ai.
