Google Analytics used to be great. For business owners, there was no better tool for measuring engagement on your website, figuring out which products resonated with users and making data-driven decisions about where to go next.

Then, one day, everything changed. Google released GA4, replacing its earlier model with a totally different one that users — including me — frankly hated. “GA4 reporting is a dumpster fire that will kill us all,” complained one Reddit user, summing up the sentiment of many. Unlike its predecessors, GA4 is complex, frustrating and seemingly designed for a poor user experience.

I get it. Products need updates to stay relevant and beat the competition. But Google did what no company should ever do — a complete rewrite of a successful tool. In their effort to reinvent, they broke something that already worked incredibly well. It’s a mistake I’ve seen before, and one I’ve made myself. Here’s what to do instead.

The pressure to reinvent

Leaders face a lot of pressure to innovate constantly. AI is rapidly altering the landscape of what’s possible, and no one wants to be left behind. Every founder is familiar with the uneasy feeling when a competitor launches a stylish new feature that suddenly makes their own offerings look a little dated. The temptation to tear it all down and start over can be intense.

Reinvention sounds bold, even sexy. It calls to mind Madonna, eternally a pop icon precisely because of her ability to stay fresh and relevant even after decades in the spotlight. According to an analysis among CEOs by PwC, 45% of respondents said they doubted their company would continue to exist in a decade if they failed to change or reinvent their business models.

Of course, there are many strong and compelling reasons for reinvention — markets shift, technology evolves and customer expectations never stop rising. But too often, it’s done for the sake of shaking things up, rather than to solve a real problem or create genuine value for the people who already love what you’ve built.

The perils of a total rewrite

GA4 may be a perfect object lesson in “if it ain’t broke, don’t fix it,” but there are plenty of others. Chartbeat is another example of a product our teams loved that one day, for no apparent reason, was suddenly replaced by a zombie-fied version of itself that didn’t suit our needs at all. It lacked so many of the core functionalities of the old Chartbeat that we had to stop using it altogether.

Such an abrupt switch feels like a betrayal to users. You may think you’re doing them a favor, but in reality, you’re throwing a wrench in a process that may have worked really well for them.

To be fair, I’ve been guilty of springing a major rewrite on my users, too. In earlier releases of our Form Builder, we’ve switched up key functions and added splashy new features without warning, thinking our customers would be surprised and delighted by what seemed to us to be obvious improvements. They were surprised, all right — and not pleased at all. We lost tons of customers, and we learned our lesson.

The right way to reinvent

Our customer exodus taught me several valuable lessons about the right way to reinvent a product.

First, look at the data. Is your product doing well? Are people using it? Consider what changes you want to make and your motivation for making them. Vanity is not a good reason — for example, if you’re trying to shoehorn AI into operations where there’s no reason to think they’re needed, stop. Change for the sake of change almost always does more harm than good.

Next, test before you leap. When we built the next version of our Form Builder, we approached it differently. Instead of suddenly releasing an unrecognizable update that shocked our users, we worked incrementally. We tested each update against the old version to ensure it performed better on every metric. Only after it surpassed the old version completely did we make the switch permanent.

Reinventions should also be in line with your values as a business. Ideally, these values are well-known to everyone at the company, from C-suite executives down to interns. At Jotform, our primary objective is to make our users’ lives easier. It’s a simple mandate, but it’s also our North Star, and every pivot and new product has to align with that mission. As Harvard Business Review puts it, the human mind values consistency. According to research, people “are less likely to register deviations as significant if they seem to be in line with larger aims.”

We’ve all been there — the product we love that gets discontinued, or sneakily replaced without anyone asking us whether we wanted it to change. It’s frustrating, and it erodes trust. The smartest move a company can make is to slow down, listen to its users and carefully build on what already works.

Sign up for the Entrepreneur Daily newsletter to get the news and resources you need to know today to help you run your business better. Get it in your inbox.