How to Research a New Software Tool Before You Sign Up: My Real Checklist

The short version: Before you sign up for any software tool, check its cancellation process before you check its features, read the one-star reviews on G2 or Capterra for patterns rather than the star rating itself, and book a demo but treat everything the salesperson tells you as marketing, not fact. Do those three things and you will avoid roughly 80% of the bad software decisions I see business owners make.

Why most people research tools backwards

Here is what usually happens. Someone sees a tool on LinkedIn, watches a slick two-minute demo video, checks the pricing page, and signs up. The whole process takes twenty minutes. Then three months later they are stuck in a twelve-month contract with a tool that does 60% of what they needed and a support team that takes four days to answer an email.

I have done this. Twice, badly enough that it cost me real money and real time. So this is not theory. This is the process I built after getting burned, and I still use it every single time I add a new tool to my stack, whether it is a £15 a month app or a £400 a month platform.

The seven checks I run before I sign up for anything

This takes me about 45 minutes for a tool I am serious about. Not twenty minutes, not three days. Forty-five minutes, done in one sitting, with a notes doc open.

• Read the cancellation policy first. Not the pricing page. The cancellation or refund policy. If I have to search their help centre for more than two minutes to find out how to cancel, that tells me something about how they treat customers who want to leave.
• Search "[tool name] complaints" and "[tool name] alternative" on Google. Not the tool's own site. This surfaces Reddit threads, forum posts, and comparison articles that the company did not write themselves.
• Sort G2 or Capterra reviews by lowest rating, not highest. The five-star reviews all sound similar. The one and two-star reviews tell you the actual failure pattern, whether that is billing issues, support that disappears after the sale, or a feature that quietly stopped working after an update.
• Check when the company last raised prices. Use the Wayback Machine on their pricing page. If a tool has raised prices three times in eighteen months, that pattern tends to continue.
• Look at their status page or uptime history. Most SaaS companies now have a public status page (status.toolname.com is a common pattern). If it shows six outages in the last quarter, factor that in.
• Check the team on LinkedIn. Has the company had a visible round of layoffs in the last year? Has the founder gone quiet? A shrinking team often means shrinking support quality even before the product changes.
• Book the demo, but bring your own questions. Ask directly: "What happens if I need to cancel mid-contract?" and "What is your average first-response time on support tickets?" Watch how quickly they answer versus how quickly they redirect.

None of this is complicated. It is just slower than clicking sign up, and most people are not willing to spend 45 minutes on something they can pay for in 30 seconds. That gap is exactly why so many businesses end up locked into tools they regret.

The part about reviews nobody wants to say out loud

Here is the uncomfortable bit. Review platforms are not neutral. G2 and Capterra both run on a model where software companies actively push customers toward leaving five-star reviews, often with a gift card incentive attached, right after a successful onboarding call when the customer is at their most optimistic and has used the product the least. I have been on the receiving end of this myself, as a customer being asked to leave a review for a £50 Amazon voucher within two weeks of signing up, long before I had any real sense of whether the tool would hold up.

That does not make review sites useless. It means the star rating is close to meaningless and the pattern in the text is where the truth lives. A tool sitting at 4.6 stars can still have a consistent thread of "billing kept charging after I cancelled" running through its one-star reviews, and that thread matters more than the average.

The other thing nobody says plainly: the sales demo is not there to help you decide. It is there to remove your objections. That is the salesperson's job, and they are usually good at it, which is exactly why you should never let a demo be your only research step. If a decision is made in that 30-minute call and nowhere else, the tool won.

A real example: the tool I signed up for without checking any of this

In 2023, I moved my client scheduling and intake process onto a new booking and CRM tool that a fellow consultant had recommended. The demo was strong. The salesperson was warm, quick to answer, and showed me exactly the workflow I wanted. I signed up for an annual plan because it was 20% cheaper than monthly, paid roughly £900 upfront, and moved three months of client data across.

Two months in, the tool changed its automation limits without warning, which broke the exact workflow I had been sold on in the demo. When I emailed support, the first response took six days. When I asked about a partial refund given the change in functionality, I was told annual plans were non-refundable, full stop, and pointed to a line in the terms I had not read closely enough. I ended up eating the remaining ten months of the contract while running a second, cheaper tool alongside it for the actual work.

Had I done the checklist above, two things would have stopped me. First, a Reddit thread from four months earlier already had three other users describing the same automation change and the same support delay. Second, the cancellation policy, which I never read until I needed it, would have told me the annual plan was locked with no exceptions. Both pieces of information existed and were public before I signed up. I just did not look.

What to do during the free trial

Most people treat a free trial as a chance to poke around for ten minutes and see if the interface looks nice. That is not a trial, that is a glance. A real trial has a job to do, which is stress-testing the specific task you bought the tool for.

Before the trial starts, write down the three tasks the tool has to nail. Not "I want better productivity," but something specific, like "it needs to send an automated follow-up email 48 hours after a form submission and log the response in a shared inbox." Then, in the trial period, try to build that exact workflow, including the messy edge cases, like what happens if two people submit the form at the same time, or what happens if the automation fails silently.

This is where choosing the right productivity tool for your business tends to go wrong for most teams. They test the happy path only, everything works, they roll it out to the whole team, and the failure only shows up three weeks later on a Friday afternoon when nobody set up the fallback.

Also test the support channel during the trial, not after you have paid. Send one real question through their chat or email during the trial week and time how long it takes to get a useful answer. That single data point tells you more about what life with this tool will be like than the entire demo call.

The cancellation test

This is the step almost nobody does, and it is the single most useful thing in this whole process. Before you enter your card details, go find the actual cancellation flow. Most tools let you see this without being a paying customer, either through a public help article or by starting the signup process and looking at account settings before confirming payment.

What you are checking for: is cancellation self-serve, or does it require emailing support and waiting for a human to action it? Is there a notice period, like 30 days before renewal, that you need to know about now rather than discovering it the week after you get auto-charged? Does the tool make it as easy to leave as it made it to join?

Companies that hide the cancel button three menus deep, or route it through a "chat with us before you go" retention flow, are telling you something about their business model. They are betting on friction, not satisfaction, to keep your subscription. That is worth knowing before you are the one stuck navigating that flow at 11pm trying to stop a renewal charge.

Where this matters more than usual: AI tools and live chat platforms

Two categories deserve extra scrutiny right now. AI tools are shipping features and changing pricing tiers monthly in 2026, so a comparison you read in January can be out of date by March. If you are weighing up options, it is worth reading a proper task-by-task comparison rather than a generic "best AI tools" listicle, which is why I put together a real breakdown of which AI tools suit different business tasks instead of just ranking them by popularity.

Live chat platforms are the other category where the demo and the reality often diverge, usually around response routing and after-hours coverage, so it is worth running through a proper comparison of live chat companies before you commit to one, rather than picking whichever rep called you back first.

If you are choosing a tool that touches assessments, training records, or anything with a compliance angle, the stakes are higher again, and it is worth reading the wider debate on whether AI tools belong in assessments and training before you assume every vendor's compliance claims are accurate. Vendors in this space are especially prone to overstating what their tool has been "approved" for.

Even something as simple as picking a scheduling tool for social media benefits from the same checklist, which is why the questions I'd ask before choosing a scheduling tool for a small team are laid out in this breakdown for beginners rather than assuming the biggest name is automatically the right one.

When it is worth paying someone else to do this research

If you are choosing a tool that will touch your whole team, cost more than a few hundred pounds a month, or lock you into an annual contract, the 45 minutes above is the minimum, not the maximum. For anything AI-related in particular, where the market shifts fast and the vendor claims are hardest to verify without technical knowledge, it is often worth bringing in outside eyes. This is one of the things I do with clients directly through AI consulting for small businesses, where the point is not to sell you a tool but to run the research above before you commit budget to something that may not fit how your team works.

The cost of getting this wrong is rarely just the subscription fee. It is the migration time, the retrained team, and the three months spent limping along on a tool that half-works before someone finally admits it needs to be replaced.

A closely related walkthrough: How Do You Research a New Software Tool Before Signing Up?.

Frequently asked questions

How long should I spend researching a software tool before signing up?

For anything under about £30 a month with a monthly rolling contract, 15 to 20 minutes of checking reviews and the cancellation policy is enough. For anything with an annual contract, a team rollout, or a cost above roughly £100 a month, budget a full 45 minutes to an hour, including a demo call and a trial run of your specific workflow, before you pay.

Are G2 and Capterra reviews trustworthy?

Treat the star rating with caution, since many companies incentivise five-star reviews shortly after onboarding, but the written patterns in the one and two-star reviews are useful, especially when several unrelated reviewers describe the same problem, such as slow support or billing that continues after cancellation.

What is the single most important thing to check before signing up for any tool?

Find the cancellation process before you enter your card details. If it is hard to find, requires emailing a human, or has a hidden notice period, that tells you more about how the company will treat you than anything in the demo.

Should I trust a sales demo when deciding on a new tool?

Use it as one input, not the deciding factor. A demo is built to remove your objections, so pair it with independent reviews, a trial run of your actual use case, and a real support interaction before you commit any budget.