Product Health Metrics: Why Success Metrics aren’t enough

Your latest product launch is crushing it!

Sign-up numbers are up, conversion rates have doubled, and revenue is soaring.

Sounds like a resounding success, but it might actually be a biased version of the full story. 

What if your latest changes compromised the underlying stability of your product, and cleared the way for the business to start falling apart? You might have accidentally slowed down your systems or created the conditions for your customers to start to churn.

The launch looks good for now, but it might come crashing down soon.

The solution is to build upon a strong foundation. To achieve this, set up and monitor health metrics alongside success metrics. 

Want to see how that applies in practical terms? In this article, I’ll describe two concrete examples of how success and health metrics can be combined to set the foundation for real product impact.

Why success metrics alone can be misleading

Success metrics track and analyse results after launching a new product or feature. 

They can relate to the product or the business as a whole and show you if the impact you expected is coming true. 

Success metrics are great at telling you what’s working, but they’re terrible at telling you what’s breaking. You don’t want to be the PM celebrating a ‘win’ only to realise your core systems are quietly imploding. 

That’s why I always recommend setting health metrics. 

They allow you to keep an eye on your product’s and business’ long-term well-being. This way, you ensure your short-term wins aren’t masking deeper, systemic issues that could backfire in the future. Think of this as taking a step back to gain perspective and look at the bigger picture.

I like to work with the three most important and often-used types of health metrics: 

• Technical health metrics

• Business health metrics

• User behaviour health metrics

So, what does this mean in the real world of product management? 

Below, I’ll explore two scenarios of how success and health metrics can work together. I’ll identify the health metrics that should be set up in each situation, and why. Both examples are fictional scenarios of companies I worked at: Doodle (a SaaS scheduling tool) and home24 (an eCommerce platform). 

Doodle: Balancing conversions with product stability

In this scenario, you’re a PM at Doodle and just launched an AI-powered scheduling feature to boost the Trial-to-Paid conversion rate.

Right after the launch, you see a rise in the percentage of users upgrading to paid plans. Seems positive, right? But how do you know this success won’t come at the expense of your product’s long-term health? By setting up health metrics that you can monitor. 

Here’s what I’d watch closely in this scenario:

1. System Uptime (Technical health metric)

The complexity of the new feature risks overloading the system, leading to downtime. 

Suddenly, users who rely on Doodle for their daily workflow wouldn’t be able to access it reliably. In the worst scenario, frustration would lead to churn. 

Monitoring system uptime would ensure the infrastructure can handle the new demands without interruptions.

2. Customer Churn Rate (Business health metric)

Conversions are exciting! But churn? Quite the opposite…

You don’t want to celebrate new users just to watch your loyal ones leave through the back door. So keep an eye on churn rates to ensure your new feature isn’t a quick fix or something that creates dissatisfaction among your existing customers.

3. Time to First Interaction (TTFI) (User behaviour health metric)

This usability metric tells you how long it takes for users to start interacting with your product after loading it. 

In some cases, this is a vanity metric, but in others, it’s strongly correlated with activation (for example through Time to Value), so it’s important to watch it. For example, if the new feature slows down the overall responsiveness of the platform, users might get frustrated before they even have a chance to use it.

home24: Driving sales without sacrificing user experience

In this example, you work at home24, a furniture eCommerce platform, and you’ve just implemented AI-powered recommendations to increase sales and average order value (AOV).

Users are adding more products to their carts, sales numbers are increasing, and AOV is the highest it has ever been.

Is this all sunshine and rainbows? Maybe. But to be sure of that, I’d look at some health metrics, such as:

1. Page Load Time (Technical health metric)

AI-driven recommendations are great, but they require more server-side processing. 

Nobody likes to wait, and when it comes to your website, users are even less patient. Shopify’s 2026 analysis found that for every 100 milliseconds slower a store loads, conversion tends to be about 3.5% lower. To ensure the overall health of your product, you’d need to keep page load times in check so that, even with the new features, users wouldn’t be waiting too long for pages to load.

2. Customer Lifetime Value (CLV or LTV) (Business health metric)

CLV is the average revenue you can expect from a customer over their entire relation with your company. CLV is calculated based on the AOV, the number of purchases, and the average time a customer sticks with you. 

In our scenario, increasing AOV and sales would likely increase the CLV. But what if your implementation is bad, and your customers churn earlier? Then your CLV will go down in the long run. 

So, as an indicator of long-term success, you could track 30-60-90-day retention.

3. Bounce Rate (User behaviour health metric)

The bounce rate indicates the percentage of users who leave your site after viewing only one page. 

In our example, sales are increasing, but if your bounce rate is also rising, it could mean that users are being turned off by something, like slow page load times or irrelevant recommendations. 

You might be driving some users to conversion but losing others who find the experience frustrating. In that case, it’s worth segmenting the numbers and understanding who is bouncing more and why.

The difference between guardrail and health metrics

There's another term that I want to mention here: guardrail metrics. I use this term more often than the term "health metrics". Technically, they have the same job. They serve as a guide for the important metrics that should not break, aka stay stable or improve.

In the day to day work, you will notice that health metric is often used for business related measures like revenue, churn, retention, etc. and guardrail metric is rather used in the operational work for technical and user behaviour related health metrics.

If you know my work, you will anticipate what I'm going to say: don't get into ideological discussions about terminology. Yes make sure that you're aligned on what you mean when you use the one or the other word, but don't turn it into a debate about which term is the "right" term or - even worse - who is using it the correct way.

Guardrail metrics are probably more of a product management term than a business term, so get ready to have a conversation (not a debate!) about what it means when you talk to business functions. And remember: they serve the same job.

To be very explicit about terminology debates, let me add another well meant recommendation. Don't even debate about the distinction between business, technical and user behavioural health metrics. I'm splitting them in this article for you - so that you have a structured way of thinking about health metrics. Nobody in your company thinks in this level of detail about them, and they don't care about the categorisation. You shouldn't either. What you should care about is to make sure you set some guardrails (there you go) for your product, some health metrics that explain what should remain or ideally even improve while your success metric moves the specified way.

Conclusion

Don’t get me wrong - success metrics are key in any product launch. 

But they are not the full picture. To have a 360-degree view of where your product stands, you need to complement them with health metrics. Otherwise, you might be celebrating wins while potentially ignoring cracks in your foundation.

Throughout my career, I've seen many products stumble not because they weren't successful, but because that success came at the cost of their foundation. By the time these issues surfaced in the success metrics, the damage was already done.

If you want to avoid that, get started with health metrics today. Identify which technical, business, and user behaviour metrics matter the most to your product and set them up. The earlier you do it, the sooner you’ll have data on which to base your decisions.

If you’re struggling to identify which health metrics to set up, reach out, and we can work together on that!

—
This article was edited by Diana Bernardo.