Why renewals happen by default
Nobody decides to keep a tool that nobody uses. It happens because the decision has no date in anyone's calendar. The renewal date is visible, the notice date is in the contract, and the review, if there is one, starts when the invoice arrives. The marketing technology hub puts it bluntly in its answer on how often to review the stack: reviews that happen only when a renewal invoice arrives tend to produce a renewal rather than a decision. The tool categories hub lists "a renewal arrives and nobody knows whether to sign" among its common problems and traces it to a missing decision record and success measure.
A renewal review has four possible outcomes, and each one needs something different from the review.
| Outcome | When it fits | What it needs before the notice date |
|---|---|---|
| Renew | The tool meets its requirements, is used and is moving its success measure | Evidence of use and results; any terms worth improving |
| Renegotiate | The tool is worth keeping but the size, scope or terms are wrong | A negotiation list drawn from use, results and incidents |
| Replace | Another option would meet the current requirements better, counting the cost of moving | An alternative scored on the same criteria; an evaluation plan |
| Retire | The job is no longer needed, or is done elsewhere | A retirement plan and notice given in time |
The choosing and evaluating tools guide says to read the decision record at renewal and lists the terms worth negotiating at signing, from the price at renewal to what happens at the end of the term by default. This tutorial is the review that reads it. An agency on a retainer renews in much the same way, and the tutorial on reviewing an agency's scope and fees before renewal runs the equivalent review for a marketing agency, from the scope against the work done to the fee model.
Before you start
Gather what the tool was bought to do and what it has done since. The decision record from its selection, which the marketing tool evaluation template holds if it was used, with the requirements, the success measures and the known gaps. The business case, if there was one. The contract or terms accepted, with the renewal date, the notice date and what happens by default at the end of the term. The adoption review the marketing tool onboarding SOP records, which the SOP names as the starting point for deciding whether to keep a tool. Invoices for the current term. And the tool's own figures on users, usage and anything it reports about its results.
Agree who decides. The tool owner leads the review; the budget holder decides; whoever owns data protection is consulted if anything about the tool's data has changed. If none of these people can be found before the notice date, the first job is to find them.
The steps
Build the renewal calendar from notice dates, not renewal dates
For every tool with a contract, record three dates: the renewal date, the date by which notice must be given, and the date the review must start to finish before the notice date. Work the start date back from the notice date, leaving enough time for the outcome that takes longest: if replacing the tool is plausible, that is an evaluation of an alternative. Record what the terms say happens if nothing is done. The guide to cutting the marketing budget treats annual tool contracts as commitments that change only at a notice date; the calendar is how those dates stop being surprises. Give each entry an owner and put the review start in their calendar, not only in a spreadsheet. The marketing stack management playbook gives the calendar a monthly reading, so each review starts in time alongside the request, access and retirement routines that run around it.
Gather the decision record and the original requirements
Read the decision record from the purchase: what the tool was chosen to do, which requirements were essential, what it scored, what gaps were known and when it was due for review. Read the business case for what it was expected to change; the tutorial on writing a business case for a marketing investment shows what a complete one contains, including how the result was to be proved. Where no record exists, reconstruct the requirements from the order form, the onboarding records and the people who chose it, and note the gap. Then check whether the requirements still describe the business: use cases change, and a tool can be judged against a job the business no longer has. The designing a marketing stack guide keeps the use-case list that current requirements are written from.
Measure use: active users, live use cases and adoption
Count who uses the tool and for what. Compare paid seats with active users, and the use cases it was bought for with the ones that are live. A tool bought for several use cases and running one of them is a different renewal from one running all of them. The tutorial on reviewing marketing tool licences and seats covers the comparison of seats, contacts, usage and modules in detail; for the renewal, the output is a short statement of what is used, by whom and for which use cases, and what is paid for but not used.
Read the success measures against the baseline
Take the KPIA metric tied directly to an objective. the decision record named and read it against the BaselineAn estimate of what a metric would have done without a given change, campaign or promotion, against which the effect is measured. it was meant to improve on. Look at the trend over the term, not one month. Be careful with small numbers: a rate built on a few dozen events can swing widely with no change in performance, and the guide to reading small numbers explains how to read the count behind a rate and put a range around it before calling a result. Be careful too with attribution: if other things changed at the same time, the tool's contribution may be smaller or larger than the measure suggests. Write down what the evidence supports and how confident it is.
List incidents, support problems and integration workarounds
List what went wrong during the term: outages that affected campaigns, data that arrived late or wrong, support requests and how they were handled, and the workarounds the team built to make the tool fit, such as a manual export, a mapping to maintain or a sync someone checks each week. The choosing guide counts workarounds as part of total cost, and they are often the largest cost the invoice does not show. Ask the people who run the tool; they know the workarounds that never reached a ticket. Outages are easier to weigh when each was written up at the time, as the guide to preparing for a marketing tool outage asks, with its window, the bookings or enquiries lost and the hours spent on fallbacks.
Check what has changed in data handling and terms since signing
Compare the tool today with the tool that was reviewed at onboarding. Has it gained a module that handles different data, a new integration, a change in where data is processed or a move from a browser tag to a server-side connection? The onboarding SOP treats each of those as a major upgrade that needs its full procedure, including the data and privacy review, so route any of them there rather than judging them here; whatever Personal DataAny information relating to an identified or identifiable living person, such as a name, email address, location data or an online identifier. questions they raise belong to whoever owns data protection. Note any change to the supplier's terms during the term. For an AI tool, re-ask the questions the guide to evaluating AI tools sets on data, training and retention terms and on model changes, because those are the ones most likely to have moved.
Re-score the incumbent against current requirements and any alternative
Score the incumbent on the current requirements, using the same criteria and weights you would use for any alternative. Apply the essential requirements as a gate before the weighted sum, as the weighted scoring matrix and the choosing guide both set out: a tool that fails an essential requirement does not stay because it scores well on the rest. Where an alternative is worth considering, the MarTech categories guide helps find the category to look in, and the choosing guide's method applies in full if replacement becomes likely. Count total cost on the same basis for each: for the incumbent, the licence plus the workarounds; for an alternative, the licence plus migration, integration, training and the period of running both.
Decide renew, renegotiate, replace or retire
Make the decision with the budget holder, against the evidence. Renew if the tool meets the requirements, is used and is moving its measure. Renegotiate if it is worth keeping but the size, the scope or the terms are wrong. Replace if an alternative scores better once the cost of moving is counted, and start its evaluation and onboarding through the choosing guide and the onboarding SOP. Retire if the job has gone or is done elsewhere, following the tutorial on retiring a marketing tool. Whatever the outcome, the notice date governs it: if the review cannot finish in time, decide with the budget holder whether to give notice to keep the exit open, knowing that the terms decide what that notice commits the business to.
Negotiate from the evidence and record the outcome
Turn the review into a negotiation list. Seats, tiers or modules that are paid for but not used. Incidents and support problems that cost the team time. Terms the choosing guide lists as worth negotiating that the current contract lacks or handles poorly: the price at renewal and any cap on increases, notice periods, what happens by default at the end of the term, data export on exit and deletion with confirmation. Negotiate with the evidence, not with a target figure from elsewhere. Then record the outcome in the decision record: what was decided, on what evidence, the terms agreed, the success measures for the next term and the date of the next review, and put the next notice date in the calendar.
Renewal questions at a glance
The steps produce answers to a short set of questions. Where the answers are poor, the right-hand column shows what usually follows.
| Question | Where the evidence comes from | If the answer is poor |
|---|---|---|
| When must notice be given, and what happens if nothing is done? | The contract or terms accepted | Start the review now, or decide whether to give notice to keep the exit open |
| What was the tool bought to do? | The decision record; the business case | Reconstruct the requirements and note the missing record |
| Are those still the business's requirements? | The current use-case list | Re-score against the current requirements, not the old ones |
| Who uses it, and for which use cases? | The tool's user list and usage; the licence review | Renegotiate the size, or retire if no live use case remains |
| Has the success measure moved against the baseline? | The measure named in the decision record | Read the counts with care; if it has not moved, ask why before renewing |
| What has gone wrong, and what workarounds exist? | Incident and support records; the people who run the tool | Add the cost of workarounds to total cost; raise service problems |
| Has anything changed in data handling or terms? | The supplier's notices; the tool's settings | Route major changes to the onboarding SOP's review |
| Would an alternative do better once the cost of moving is counted? | A scoring on the same criteria | Plan the replacement in time, or renegotiate |
A furniture retailer reviews its recommendation tool (illustrative)
Fluvorix Furniture reviews its recommendation tool before renewal
Fluvorix Furniture, an invented online furniture retailer, checks its renewal calendar and finds that its product recommendation tool renews in 12 weeks, with notice due 8 weeks before renewal: 4 weeks away. The decision record from the purchase two years earlier lists 4 use cases (recommendations on product pages, in the basket, in post-purchase emails and on the home page) and a success measure: the share of online revenue from recommended products, against a baseline of 6.0%.
The review finds 2 of the 4 use cases live; the email and home page placements were never built. The share of revenue from recommendations averaged 6.3% over the last 12 months, within the range the team would expect from month-to-month variation on its order counts, so the evidence for an improvement is weak. Support tickets show 3 incidents where recommendations showed out-of-stock items for several days, and the team maintains a nightly stock export to work around a feed the tool reads only once a day. Nothing about the tool's data handling has changed.
Re-scored on the current requirements, the tool passes every essential requirement and scores 3.4 out of 5. One alternative in the same category scores 3.7 but would need a new integration and a period of running both, which the team cannot fit before peak season. The marketing lead and the budget holder decide to renegotiate rather than renew by default: they ask for the smaller tier that matches the 2 live use cases, a fix or a credit for the stock feed, and a review date in 6 months with the email placement built and measured in between. The outcome, the new terms and the next notice date go into the decision record.
Common mistakes
- Working from the renewal date. The notice date comes first; a review that starts at the invoice can only renew.
- Judging the tool against old requirements. Use cases change; re-score against what the business needs now.
- Reading a flat measure as success or failure on small counts. Look at the count behind the rate and the range around it before deciding.
- Ignoring the workarounds. A manual export someone runs every week is part of what the tool costs.
- Scoring the alternative and not the incumbent. Compare both on the same criteria, with the cost of moving counted.
- Treating retirement as failure. Retiring a tool whose job has gone is a good outcome; renewing it by default is not.
- Not recording the outcome. Without a decision record, the next renewal starts from nothing again.
