The pattern is common enough to be predictable. A renewal notice lands, the number is 12 to 20 percent higher than last year, and the platform it renews is the one your team spends its day working around. The renewal date is somewhere between two and six months out. Nobody wants to sign, and nobody has a replacement ready. This guide is about what you can actually do inside that window, and, just as importantly, what you cannot.
None of what follows is legal advice. Contract language varies enormously between agreements, and the specific clauses discussed here are things to put in front of your own counsel rather than conclusions to act on directly.
Read the contract before you read the quote
Most companies react to the price and never open the agreement. That is backwards. The renewal price is the negotiable part. The clauses governing whether you are permitted to say no are usually not, and they typically bite months before the date on the invoice.
Three provisions decide how much room you have. The auto-renewal clause determines whether the contract rolls forward on its own. The notice period sets how far in advance written non-renewal must be delivered: 60 to 90 days is common, and some agreements require it in a specific form, to a specific address. The uplift or escalation clause caps, or fails to cap, the annual increase. Ask your counsel to confirm all three in writing, along with any minimum seat commitment that survives the term.
Serving notice is reversible in most agreements and it does not commit you to leaving. It converts an automatic outcome into a decision you still own. Confirm the mechanics with counsel, but as a matter of practice, serving notice early and negotiating afterwards is a materially better position than negotiating while the clock runs out.
What you actually have, and what you do not
It helps to be honest about the asymmetry. The vendor knows your seat count, your module mix, your support ticket history, and roughly how embedded the system is. They have seen this conversation thousands of times. You have seen it once every three years.
What you have
- Timing. Vendors have quarter and fiscal-year targets, and concessions are easier to obtain near them than in the middle of a quarter.
- Seat count. Reducing licensed users to those who genuinely use the system is often the largest single lever, and it is defensible with usage data.
- Module scope. Modules bought for a project that never landed are frequently still being paid for.
- Term length. Vendors want multi-year commitments and will trade discount for duration. That trade is not always in your favor.
- A credible alternative, but only a real one.
What you do not have
- A cheap exit. The switching cost is real, the vendor knows it, and pretending otherwise damages your credibility in the room.
- Time to build a replacement inside a renewal window, if the window is measured in weeks.
- A bluff worth making. "We are evaluating alternatives" is only useful if it is true and you can name what you are evaluating.
- Much movement on list price itself. Concessions usually arrive as term length, seat count, module scope, or professional services credits rather than a lower per-seat rate.
That last point about the bluff matters more than it sounds. Vendors can tell the difference between a customer who has scoped an alternative and one who is posturing. If you say you are evaluating options, expect to be asked which ones, and expect the answer to be tested. An evaluation that is genuinely underway (a named alternative, a scoped bridge, a partner engaged) changes the conversation. An empty threat costs you the next three years of it.
What can be built before a renewal date
This is where most planning goes wrong. The instinct under deadline pressure is to promise a replacement and start building, then discover in week nine that financial reporting cannot move in time. The better move is to separate the workflows that are actively costing you from the ones that are merely annoying, and move only the first group.
| Scope | Feasible before the date? | Why |
|---|---|---|
| Core dispatch, scheduling, and field updates | Yes, roughly six weeks | Operationally urgent, structurally simple, and the data set is mostly current rather than historical. |
| Job completion to invoice handoff | Yes, with an accounting integration | The invoice can be generated outside the ERP and pushed into accounting; the ledger itself does not have to move. |
| Customer records and job history | Partly | Active records migrate cleanly. Full history is a separate, longer exercise and rarely needs to block a cutover. |
| General ledger, financial close, statutory reporting | No | Audit continuity, period close, and accountant familiarity make this the wrong thing to rush. |
| A purpose-built replacement platform | No | Six to twenty-four months, phased. Attempting it against a renewal date produces a cutover nobody can roll back. |
A trades services business facing a six-figure ERP renewal moved core dispatch, field updates, and invoicing onto a bridge stack in six weeks, avoiding roughly $100,000 in renewal cost while the real platform was designed around how the work actually ran.
In that engagement the bridge was built on Airtable, Stacker, and Make.com, and it was intentionally temporary. Its job was to beat the renewal date and to buy observation time. The team used the bridge period to watch how the business actually operated, separate genuine requirements from habits the old system had imposed, and clean up data quality before any of it reached a purpose-built platform. The renewal saving was the headline. The better outcome was designing the replacement against evidence instead of against a deadline.
A timeline keyed to months before renewal
How much room you have is almost entirely a function of when you start. The table below is the sequence we would run, working backwards from the renewal date.
| Time to renewal date | What to do now | What is still on the table |
|---|---|---|
| 6+ months | Have counsel confirm the auto-renewal clause, notice window, uplift cap, and any surviving seat minimum. Pull actual usage data by user and module. | Everything, including starting a phased platform replacement this year. |
| 4–6 months | Decide whether to serve written notice. Begin a real evaluation of alternatives, and scope what a bridge would have to cover. | Bridge build, competitive evaluation, and a renewal negotiated from a position you can defend. |
| 3 months | Serve notice if the window is closing. Fix the bridge scope: which workflows move, which data comes with them, who signs off that it works. | A six-to-ten week bridge for core operational workflows. Not a platform replacement. |
| 6–10 weeks | Build and validate the bridge in parallel with the live system. Keep the vendor conversation open. A working alternative improves the terms you are offered. | Non-renewal on core workflows, or a short bridging term at a reduced seat count. |
| Under 30 days | Stop trying to build. Negotiate the shortest available term at the smallest defensible seat count and plan the exit properly afterwards. | A one-year or month-to-month extension. Multi-year should be off the table. |
The bridging renewal is usually the right answer
There is a middle option that gets overlooked because it feels like a loss. You renew (briefly, at a reduced footprint), and you use the bought time to do the replacement properly. Twelve months at half the seats and none of the modules you stopped using is frequently cheaper than the discounted three-year deal, and it does not mortgage the decision.
- 1
Produce a usage report by named user and by module. Licensed but inactive seats are the strongest and least arguable reduction you can ask for.
- 2
Ask for the shortest term the vendor will write. If they will only quote multi-year, ask what a one-year term costs so the premium for flexibility is explicit rather than implied.
- 3
Price the multi-year discount against the cost of being stuck. A 20 percent saving on years two and three is not a saving if you intend to be off the platform in year two.
- 4
Have counsel check what the renewal does to the notice clause, the uplift cap, and any termination-for-convenience right. Short terms sometimes come with worse exit language.
- 5
Get any professional services credits, data export commitments, and support obligations written into the agreement rather than described in email.
When the honest answer is to renew and stay
Not every renewal you dislike is a system you should replace. If the complaints are about configuration, training, or a bad implementation rather than about the platform's structural fit, replacing it will reproduce the same problems in more expensive software. The signals that justify leaving are specific: per-seat pricing that keeps necessary users out of the system, workflows the platform cannot represent so people run them in spreadsheets, integrations that break in ways nobody can monitor, and a customization bill that rises every year while flexibility falls.
If those are absent, negotiate hard, right-size the seats, and keep what you have. A renewal you resent is not by itself a business case. The business case is the five-year total cost including the labor spent working around the system. If that number favors staying, stay.
Common questions
- Can we cancel an ERP renewal after it auto-renews?
- That depends entirely on your agreement, and it is a question for your counsel rather than a general rule. In many enterprise contracts the renewal term becomes binding once the notice window passes, and mid-term exits are limited to specific breach conditions. This is why the notice date, not the renewal date, is the deadline worth tracking.
- Does serving non-renewal notice damage the vendor relationship?
- In our experience it is treated as routine procurement practice, not as an insult. Notice is usually withdrawable and it is often the only way to keep the option of leaving alive while terms are still being discussed. Confirm the withdrawal mechanics with counsel before you rely on it.
- How much can we realistically move off the ERP in six weeks?
- Core operational workflows (dispatch, scheduling, mobile field updates, customer records, notifications, and the handoff into invoicing) are achievable in roughly six weeks on a bridge stack, and we have delivered exactly that scope against a renewal date. Financial close, general ledger, and statutory reporting are not, and should not be attempted on that timeline. Active data migrates in the window; full history usually follows later.
- Is a no-code bridge stack a real solution or just a delay tactic?
- It is a real solution for a defined period with a defined exit. It removes the immediate cost and unblocks the operation, and it produces something more valuable than a delay: evidence of how the business actually works, gathered from working software rather than from requirements workshops. It becomes a problem only if nobody decides what replaces it.
- Should we tell the vendor we are evaluating alternatives?
- Only if it is true and you can name what you are evaluating. Vendors test the claim, and an evaluation that turns out to be hypothetical costs you credibility for the rest of the negotiation and the next one. If the evaluation is genuinely underway, saying so plainly is reasonable and usually changes what is offered.
- What if the renewal date is three weeks away?
- Stop planning a build. Negotiate the shortest term at the smallest defensible seat count, get the terms reviewed, sign, and then run the replacement decision on a sensible timeline. Three weeks is enough to improve a contract; it is not enough to move an operation.