Why do buyers keep software they rate poorly?

Interview the same buyers again at least nine months later and, of the tools they rated 4 out of 10 or lower, about 44% are confirmed still in use, 18% confirmed replaced or dropped, and 38% never come up. Where the outcome is known, roughly seven in ten were still there. Poorly rated tools were confirmed gone more than four times as often as tools rated 9 or 10, but a low score was far from a guarantee of departure. Buyers who say why one stays most often say it still does a narrow job cheaply, or that leaving costs more than staying.

Based on verified interviews with the marketing, ecommerce, data and technology leaders who decide whether to keep a tool or replace it, at DTC and enterprise brands. Buyers are anonymized before publication; vendor names and views are reported as given. No vendor paid to appear or could edit this page.

44%
Of the tools a buyer rated 1 to 4 out of 10, the share confirmed still in use when Alium interviewed the same buyer again, nine months or more later.
18% vs 4%
The share confirmed gone at the second interview: tools first rated 1 to 4, against tools first rated 9 or 10.
4 in 10
Of the poorly rated tools whose buyer gave a reason for keeping it, the share where it does a narrow job well enough or costs little. Reasons overlap; this is not a share of all poorly rated tools.

Do buyers replace software they rate poorly?

Some do, but a low rating does not reliably say what happens next. Among the people Alium interviewed twice at the same company, at least nine months apart, a tool they had rated 1 to 4 out of 10 was confirmed gone at the second interview about 18% of the time, confirmed still in use about 44% of the time, and not mentioned the remaining 38%. Tools rated 9 or 10 were confirmed gone only about 4% of the time, so confirmed departure is more than four times as common among the lowest-rated. Among the poorly rated tools whose outcome is known, about 71% were still there. That leaves out the 38% that did not come up, so it does not mean seven in ten of all poorly rated tools stayed. These are outcomes observed among these buyers, not churn rates for any vendor or a prediction for any one tool.

Confirmed gone Still there Not mentioned

Each row follows every tool rated in that band by a buyer Alium interviewed again at the same company at least nine months later. Shares are of all the tools first rated in the band, including those not mentioned the second time. Still there means the buyer confirmed the tool was in use; confirmed gone means the buyer reported it replaced or removed; not mentioned means the outcome could not be determined. Each row covers several dozen tools or more, pooled across buyers, and one buyer can rate several tools, so rows are not counts of separate companies. Shares are rounded.

What the three outcomes mean

The three outcomes come from what the second interview shows, and they do not overlap.

Still there
The buyer confirms the tool is still in use at the second interview, in the stack they describe or by talking about using it. That includes tools being migrated off or considered for replacement, which the next section separates out. It measures continued use at the follow-up, not a decision to keep the product for good.
Confirmed gone
The buyer says the tool was replaced or dropped since the first interview.
Not mentioned
The tool did not come up. An interview covers what the buyer chose to talk about, not a full inventory, so these tools may still be there or may be gone; nothing on this page counts them either way.

Every share here is a share of the tools that re-interviewed buyers rated, not of all buyers or all companies. It is not a churn rate for any vendor or category, and the people Alium happened to interview twice are not a random sample of the market.

Still there is not the same as staying

Confirmed in use at the second interview is not the same as a decision to keep a tool. About a quarter of the poorly rated tools still there were tied to a move away, and the evidence is not equally strong across them. Roughly half of that quarter were already mid-migration: a content system being moved site by site onto something new, a commerce platform midway through a move. The rest were being shopped for or had a replacement planned, like a front-office system whose replacement kept slipping. A migration underway is much stronger evidence of departure than an evaluation. For the other three-quarters no move away was identified, which still does not show that anyone decided to keep them for good.

A marketing director at a multi-site healthcare services company shows how long on the way out can last. At the first interview, the front-office system its clinics run on was not due for replacement before the end of the following year, and even then only the preparation would start: “This is basically almost like an open heart operation for our business.” At the second interview, eleven months on, it was still there, and the date had moved: “It keeps shifting.” The decision sits with the global group: “We keep being shifted out. I don't have any control of that.”

One product accounts for a good part of this group. Google Analytics 4, which is free and is the default for web analytics, makes up about a fifth of the poorly rated tools still in use at the second interview, and none of its low ratings ended in a confirmed departure; some buyers had added another analytics tool alongside it instead. Leave it out and the share confirmed still in use drops from 44% to about 39%. So the finding is partly one product, but continued use of poorly rated tools is not limited to it, and these interviews do not show that its price or its default status is what kept it in place.

This is what buyers in general found. What fits your stack?A free tech stack review, built from hundreds of verified buyer interviews.
Get my stack review →

Why do buyers keep a tool they rate poorly?

This evidence is separate from the follow-ups above. It comes from what buyers say, in a single interview, about tools they rate 1 to 4 and still run, so it describes their explanations for continued use, not which tools were later replaced or kept. For about half of those tools the buyer does not say why it stays, which is a missing explanation rather than evidence that there is no reason. The shares below are of the tools whose buyer gives at least one reason; a tool can carry several, so they add to more than 100%.

Reasons buyers give for still running a tool they rate 1 to 4 out of 10, by family. Each share is of the poorly rated tools with at least one stated reason, counting those whose buyer gave that one. Reasons are not exclusive, and they record what buyers said, not measured causes.
Why it is still there Share In one buyer's words
It does the narrow job it is kept for, or costs little 38%“If it was a significant cost, I probably would have turned it off by now.”
Leaving costs more than staying: the migration work, what depends on it, the rest of a vendor's suite 32%“another large project with a limited benefit versus just staying on where we're at”
Not yet: a contract still running, or other projects first 27%“I have another year of contract.”
Not the speaker's decision 22%“It sucks, but it's what I was handed and I just deal with it.”
Nothing better on the market 10%

It still does the job it is kept for. The largest family is not a complaint about lock-in at all. Buyers keep a tool for one narrow use and stop expecting more of it: a session-recording tool already on its cheapest plan, which its chief digital officer would have switched off if it cost more; a free analytics tool kept as a fallback; a reviews platform kept only because, as one ecommerce leader puts it, “we are only using for syndication of reviews to, you know, various partners”. The rating measures the whole product. The decision to keep it often rests on one job.

Leaving costs more than staying. About a third describe what an exit would take: years of reporting history to move, a migration they only recently finished, other systems that read from the tool. A manufacturer's digital marketing director would like to fold a small training platform into the main commerce site and has not, for the reason quoted in the table above. The cost here is usually the work and the dependencies, not the vendor.

Not yet. About a quarter are waiting for a contract to end, or have something bigger to finish first. An ecommerce director at a DTC fashion brand, unhappy with the loyalty tool it moved to, is staying for the year left on the contract. A nonprofit's ecommerce head, unhappy with a testing tool, is not looking: “just from a timing perspective, it doesn't really make sense.”

With a plan to replace both the commerce and email platforms, this buyer renewed the suite contracts anyway, for between one and three years each, “because I wanted some elbow room” to bring in a new identity provider first, and started on a data strategy before choosing what comes next. Senior VP, IT · specialty retailer · May 2025

Not the speaker's decision. About a fifth say the call is someone else's: IT, a central marketing-technology team, the parent company, or a predecessor who signed the contract. Sometimes the people who share a tool cannot agree.

Unhappy with a work-management tool, the team explored replacements and could not settle on one: “Our director wanted to change, but other directors, like from the creative team, didn't want to.” They gave the incumbent another try. Associate Director, Digital Marketing · real-estate developer · Apr 2026

The counter-current: the suite is one of the reasons buyers give least

Being tied to the rest of a vendor's products, or to an enterprise agreement, sounds like the obvious explanation. It comes up for fewer than one in ten of the poorly rated tools with a stated reason, against about a third whose buyer describes the work or cost of leaving, the broader family the suite tie sits inside. Migration work and the systems that depend on a tool come from how it is used rather than from the vendor's bundle. Waiting out a contract, the other obvious candidate, is about one in eight, part of the quarter in the not yet family. These are how often each explanation is given, not a ranking of real causes: a buyer may name only some of their reasons.

What this means if you own a tool you rate poorly

Our recommendations, drawn from these accounts rather than from a process buyers follow. A low rating on its own does not settle whether a tool should be replaced. Four questions help establish what value it still delivers, what leaving would take, when a change could happen and who can authorize it:

Want a review of your own stack?

Get my review →

What not to read into this

That low ratings do not matter. Tools first rated 1 to 4 were confirmed gone more than four times as often as tools rated 9 or 10. That is an association between first satisfaction and confirmed departure in these repeat interviews; it does not show that the rating caused the decision, and it does not predict what any one account will do. Some of the poorly rated tools still in use were being migrated off or considered for replacement, and the long-term outcome of the rest is unknown.

That a tool still in place got better. Poorly rated tools that were rated again the second time tended to come back higher, and tools first rated 9 or 10 tended to come back lower. That is consistent with regression toward the mean as well as with real changes in experience or judgment, and the interviews cannot separate the two. A higher second rating is not, on its own, evidence that the software improved.

That these are churn rates. The shares are of tools rated by people Alium interviewed twice, a group that is neither a random sample of buyers nor of any vendor's customers. And about a third of tools did not come up the second time, so neither still there nor confirmed gone is a ceiling.

That the reasons explain the outcomes. The reasons come from what buyers say about tools they rate poorly in a single interview. They describe why buyers say a tool stays, not which reasons predict that it will.

Common questions

Do companies replace software they are unhappy with?

Often not within a year. Among buyers Alium interviewed twice, at least nine months apart, a tool they had rated 4 out of 10 or lower was confirmed still in use at the second interview about 44% of the time, confirmed gone about 18% of the time, and not mentioned the remaining 38%. Where the outcome is known, roughly seven in ten were still there. A low rating still matters: tools rated 9 or 10 were confirmed gone only about 4% of the time, so the poorly rated ones were confirmed gone more than four times as often. That is an association in these interviews, not a churn rate or a prediction for any one tool.

Why do companies keep software they don't like?

When buyers say why a tool they rate poorly is still in use, the answers fall into a few families. Of the tools whose buyer gives at least one reason, about four in ten are kept because they do a narrow job well enough or cost little. About a third are held by the cost of leaving: the migration work, the systems and partners that depend on them, or the rest of a vendor's suite. About a quarter wait on timing, because a contract is still running or other projects come first. About a fifth are not the speaker's decision. Many carry more than one reason. For about half of the poorly rated tools still in use, the buyer gives no reason, which is a missing explanation rather than proof there is none.

Is vendor lock-in the main reason buyers stay with software they dislike?

Not in the sense of a vendor's suite. Fewer than one in ten of the poorly rated tools with a stated reason are held by the rest of the vendor's products or an enterprise agreement, against about a third whose buyer describes the work or cost of leaving, such as migrating years of data, rebuilding what was built on the tool, or the other systems and partners that depend on it. Those are lock-in of a kind, but they come from how the tool is used, not from the vendor's bundle. These are how often buyers give each explanation, not a ranking of real causes.

Does a low rating mean a tool is about to be replaced?

Not reliably. In these interviews about one in five tools rated 4 or lower was confirmed gone nine months or more later, against about one in twenty-five rated 9 or 10, and another 38% of the poorly rated tools did not come up. A tool still in use may also be on its way out: about a quarter of the poorly rated tools still there at the second interview were tied to a move away, roughly half of them already mid-migration and the rest being shopped for or slated for replacement. A low rating is associated with more confirmed departures, not a prediction for any one tool.

What should you do with a tool your team rates poorly?

Our recommendation, not a practice buyers describe, is to treat keeping it as a decision rather than a default, because a low rating alone does not settle whether it should go. Identify the work that still depends on it and whether something you own or a credible alternative does that work at comparable quality and cost; map what leaving would take, including migration, training, running both and the risk of disruption, and compare it with the cost of staying; confirm renewal dates, notice periods and deadlines; and name who has the authority to approve a change. A tool that survives those four questions is being kept for a reason; one that does not may be kept by default.

For the rating at which buyers start looking at alternatives, see when buyers start shopping; for the tools that travel with a person to their next company, do marketing tools follow people; for the push to run fewer tools, the great consolidation; and for how long a platform move actually takes, ecommerce replatforming timelines.

A tech stack review, powered by the buyers who've been there

What matters most to you, a shortlist for your stack, and the next steps to take — built from hundreds of verified buyer interviews.

Get my stack review

Free · no password · your answers are anonymized before they inform research like this.

Want to come back to this?We'll email you a link to this page, so it's easy to find later or forward to your team.
Methodology. Alium conducts verified interviews with software buyers — the marketing, ecommerce, data and technology leaders who select and operate these platforms. The outcome figures on this page come from the people Alium interviewed twice at the same company, with the second interview at least nine months after the first (typically about eleven). Every tool a buyer rated in the first interview is classed by what the second interview shows: still there (it is in the stack the buyer describes, or named as in use), confirmed gone (the buyer says it was replaced or dropped), or not mentioned (it did not come up). Tools rated 1 to 4 out of 10 were also read by hand in both interviews, and the hand reading matched the automated classification within two points; the same reading identified the still-there tools tied to a migration underway, a planned replacement or active shopping at the second interview. The reasons are a separate, single-interview analysis: each unit is one tool rated 1 to 4 in one interview, drawn as a random sample from interviews from 2025 onward whose buyer discussed changing tools at all, and coded from the buyer's own words into the families shown. The half with no stated reason is a share of the sampled tools still in use at that interview, including those being evaluated or replaced; the family shares are of the sampled tools with at least one stated reason, a tool can carry more than one, and the contract figure is part of the not-yet family on the same base. Duplicate product records are merged before counting. Buyer identities are verified at interview time and anonymized before publication; vendor names are reported as given. No vendor paid to appear or was able to edit this page.