Every good product owner, whether for a physical product, Web application, or Web page, knows the importance of measuring the success of your efforts through conversion KPIs. But it isn’t just enough to know you need to be tracking your performance – you also need to select the right indicators of success to really get down to the truth. This can be a challenge for many, especially given that different metrics are optimal, depending on your position within the hierarchy of the business.

First, it helps to identify conversion KPIs that are commonly used, but don’t really tell us as much as you might think.

Bad KPI #1: Anything that isn’t relative to your own performance

Focusing on the wrong KPIs

As the headline suggests, this is less of a specific KPI, and more of a rule of thumb: Always compare your performance against your own past performance. I often hear business owners asking what they should expect from their bounce rate, or sitewide conversion rate, thinking there is some magic number they need to hit. But every site and every product has its own intricacies and quirks, so it’s difficult to make a valid comparison of metrics against another business.

Here’s an example. One of my clients does more than 50 percent of their revenue over the phone because their most popular products currently require very specific information to be gathered by customer service – they don’t even have a purchase button on the PDP (product details page), the main CTA is “Call to Order” with their phone number. Their on-site conversion rate, as a result, is significantly lower than most other eCommerce sites. Their bounce rates may also be higher – a prominently placed phone number could result in users landing, calling and leaving the site without hitting another page, even if that user has placed an order! More important than comparing their bounce rate against another site would be to compare against how they were doing one month ago, or one year ago.

As with any metric you use, remember these are just an indication of what you might need to investigate. Bounce rate going up in the above scenario could be a good thing if you increased the prominence of the phone number between your comparison periods. Taking these metrics in the context of your own site and changes you’ve made will help you continue to improve, regardless of what others in your space are doing.

Bad KPI #2: Sitewide conversion rate

This is a common one, and as a CRO advisor, it is both a blessing and a curse that this carries so much weight. On the one hand, tanking conversion rates bring clients in the door. On the other, if the slide continues during the CRO program, it can lead clients to think you’re incompetent.

The main reason this high level metric is problematic is that one of the biggest factors in conversion rate is traffic. Blog posts on this topic abound, but the long and short of it is that all traffic is not created equal. Branded search, for instance, tends to convert better than most other keyword terms. Likewise, referral traffic tends to convert higher than display traffic. If you expand your keyword footprint to target new markets, or start running a new display campaign, you’d expect to see your sitewide conversion rate drop, even as other channels might be doing better.

For an even larger contrast, consider your conversion rates for desktop vs. mobile – with the growth in mobile traffic, nearly every site on the Web should have seen their conversion rates dip. If you haven’t experienced this, either you’ve got a great mobile team or a non-existent one. I have yet to see a site where mobile conversion rate was anywhere near as high as desktop, and many sites I review now have upwards of 50 percent of their traffic coming from mobile. As traffic has shifted from the higher converting desktop to the lower converting Mobile, overall conversion rates go down, even as your conversion rate by device improves. Consider the below (oversimplified) example – note how the conversion rates for both desktop and mobile have increased, but the sitewide rate is lower.

The other problem posed by mobile is that users can get counted up to three times if they visit on their three devices (desktop, tablet and mobile). Throw in multi-computer houses, work computers, multi-device users, and you can see how this might get out of control. A user who converts after viewing your site on three devices will show as a 33 percent conversion rate instead of 100 percent!

Based on all this, sitewide conversion rate is clearly not a good indication of the health of your business.

Bad KPI #3: Revenue per visitor

Average order value is an unreliable variant | woman on a bench with shopping bags


The main reason this is a bad KPI is that it is a composite of two totally separate metrics, both of which have their own nuances. The first of the metrics, Sitewide Conversion Rate, is addressed above at great length. The second is Average Order Value.

Average Order Values have less of the problems listed above; however, they can still fluctuate greatly. Things like thresholds on free shipping, sales, bundles and item pricing factor in here, and frequently these details are not noted in analytics. When comparing against things that happened 1+ years ago, very few people in your organization will remember these things well enough to account for them in their analysis. And even if you do have everything perfectly annotated in analytics, how much of the impact can be attributed to your recent performance vs., for example, an increase in the minimum free shipping threshold?

What conversion KPIs should I be looking at?

Well, that all depends on who you are. Let’s start at the top.

Conversion KPI’s for Execs #1: Bounce rate*

These metrics are key because they help you figure out if you’re meeting users’ expectations when they land on your site.

While site speed can be a factor here, your site would have to be extremely slow to be significantly impacting bounce rate. Below are a few resources we use to analyze site speed and identify areas for improvement there. We recommend aiming for a score above 80 in the Google tool for both desktop and mobile – not blindingly fast but not painfully slow either. To put the import in perspective, I recently worked with a client to more than double their site speed score (~40 to ~80 in the Google PageSpeed Tools) and we saw almost no difference in their bounce rate after two weeks. That being said, I expect a major reason for that is user expectations are not being met; in which case having the fastest site on the planet won’t make a difference.



If you’ve determined your site speed isn’t the main issue, you likely are not handling the things that matter to your users when they land at your site. Your landing pages should do two major things – highlight your value proposition and handle any major objections they might have to continuing to explore your product. These might seem like in-the-weeds kind of items, however, in my experience these are top-down problems – if the leadership team has a clear message for the brand it is easy for the rest of the company to execute. As an executive, the bounce rates can be a clue to understand the strength of this message.

Your Message Here sign

One very important note: As you improve your message, you could actually see your bounce rate increase! This could indicate that a good portion of site visits were from users who were not actually potential customers. In this instance, the increase in bounces is actually a positive thing since you won’t be wasting resources on unqualified users. While you might take a hit initially from these bounces, ultimately it will result in more of the traffic you want coming to the site. In order to tell if this is the case, have a look at your Return Visitor Rate. If it’s going up, it means that you have gained customers, despite the higher bounce rate.

KPIs for Execs #2: YoY revenue and net profit

This one is probably a no-brainer for most executives, but then again I’m sure many considered tracking conversion rates as a no-brainer. Revenue is why most companies exist – even for non-profits, as revenue is what allows you to complete your mission. Tracking revenue on a monthly basis year over year lets you know if you’re moving the right direction. It should help remove major seasonality from your results giving a clear picture of whether your efforts are gaining a return. Be mindful though – if your expenses have gone up in order to drive revenue growth, it’s important to make sure the costs haven’t outpaced the growth!

KPIs for Execs #3: Device and channel mix and conversion rates

Harken back to the first Bad KPI. Understanding device and channel mix is important to an exec not because of sitewide conversion rates, but because it helps understand where the opportunities will lie in the future. This is most obvious for device type – most sites will have seen an explosion of mobile traffic over the past few years and should be making significant investment there.

Getting a bit deeper, understanding channel mix and conversions gives insight into both your internal teams’ performance and to your place in the competitive landscape. For instance, if you see your Organic traffic declining you know users are finding other sites more relevant for things you used to dominate, or that the number of competitors in a certain sector is increasing. Knowing there is increased competition might impact your decision to either double down in that niche or look for other opportunities that fit your core competency. Or, it could lead you to consider a totally new direction for your company if it turns out that your decreased traffic is a broader industry trend and users just aren’t interested in what you are currently offering.

KPIs for non-executives

As an executive, you’ll also want to know that your department heads and their teams are looking at the right metrics. Here are a few pointers to help you understand if they are considering the right information.

  • Do their metrics have a direct impact on your top KPIs? If not, can they explain why you should care? For example, capturing emails does not directly impact the bottom line, but if the email program is strong and converts, well then this can still impact your bottom line.
  • Are they accounting for all traffic sources/types that impact their program? Lots of companies made the mistake of neglecting mobile because conversion rates were low, rather than addressing this as a new paradigm – it’s important that you at least discuss how to handle each channel and understanding performance is a necessity.
  • Are they using composite metrics (like RPV)? If so, are the individual measurements useful? It is likely better to break them out into the components rather than look at the composite.
  • Are “micro-conversions” being tracked? These would include things like email capture, items added to cart, or completion of specific important flows, such as a custom building tool. While you as an exec don’t need to have this visibility, it’s important your department leaders do.
  • What questions are being answered by these metrics? What questions are not being answered that are important for this group?

Tying it all together

Ultimately, you want to be sure you are looking at metrics that really tell you about the health of your business and can help drive future strategy. These will be different for every business, but the above should be a good start and help you analyze other metrics for their viability.

When considering your teams and their metrics, it’s not important to get into the weeds with them. All you need is the confidence that they have chosen their metrics with the same care as you have selected yours, and that they understand how those metrics fit into the bigger picture.

*These are conversion specific KPIs. Inflow recently published a similar article on digital marketing metrics where we named overall bounce rate a bad metric. Read about the differences here.