With the holidays approaching, we often hear retailers say they want to increase their mobile conversion rate this season compared to 2016. On the surface, an increase in conversion rate year-over-year sounds like a great goal, however, when we look more deeply at conversion rate and the variables that influence its movement, we see that it can often be a red herring. In fact, an increase in conversion rate is not always a good thing. Likewise, a decrease in conversion rate is not always a bad thing. Moreover, many retailers may find their mobile conversion rates have dropped this year compared to last year, even though their mobile transactions, revenue and traffic have increased year over year. Here’s why.

Melissa Andrew, senior customer success manager, Mobify

Conversion rate alone is not a good metric to evaluate the performance of your mobile site. As we know, conversion rate is transactions divided by traffic (sessions or users) in a given period. Because it is a compound metric, interpreting conversion rate requires taking the extra step to understand the movement of the variables—traffic and transactions—that comprise it.

The complexity surrounding conversion rate is best illustrated through an example (see table 1). Let’s say on Monday we had 400 visitors to our site and 20 of them bought something. Our conversion rate on Monday would be 5%. On Tuesday, we see that our conversion rate has increased to 6%. There are five different scenarios that could have caused our conversion rate to increase by 1%. In scenario A, our traffic stayed constant—400 visitors came to our site—but 4 additional people converted, resulting in a 6% conversion rate. In scenario B, our transactions remained constant—20 people converted—but our traffic decreased by 20%. In scenario C, our transactions increased by 20%, but our traffic dropped by 5%. In scenario D, both our transactions and traffic increased with transactions increasing more rapidly than traffic. In scenario E, both transactions and traffic decreased with traffic falling more rapidly than transactions.

For most retailers, the ideal scenario is scenario D—the funnel is growing (more customers to convert) and more customers are transacting. The other four scenarios are less ideal, with stagnant or declining traffic and/or transactions. If we were just looking at conversion rate to measure performance though, we would miss that traffic and/or transactions were stagnant or declining. All we would see is an increase in conversion rate and we would think our site was performing well.

Table 1: 5 Scenarios Causing a 1% Increase to Conversion Rate

On the other hand, let’s walk through scenarios in which conversion rate may decline. Again let’s say we had 400 visitors to our site on Monday and of those visitors, 20 of them converted. Our conversion rate on Monday is 5%. On Tuesday, our conversion rate drops to 4%—but why (see Table 2).  In scenario A, our transactions dropped by 20% while our traffic stayed constant. In scenario B, our transactions remained constant—20 people checked out—but our traffic increased by 20%. In scenario C, our transactions dropped by 20%, while our traffic increased by 5%. In scenario D, both our transactions and our traffic increase by 5% and 20% respectively. Finally, in scenario E, both our transactions and traffic decreased with our transactions declining more rapidly (by 20%) than our traffic (at 5%). Scenario D is again an ideal retail scenario. More visitors are coming to the site (more opportunities to convert) and more of these visitors are moving through checkout. Yet, if we were only focused on conversion rate, we would see that our conversion rate dropped and worry there was a problem.

Table 2: 5 Scenarios Causing a 1% Decrease to Conversion Rate

Scenario D is the scenario retailers hope to find themselves during this holiday season. Depending on how quickly traffic is increasing relative to transactions, however, one retailer may find conversion rate is dropping, while another may find their conversion rate is rising even though both have more mobile traffic and transactions (and likely revenue) than last year.

So, as we recover from our turkey hangovers and start to jump into site analytics, how should you approach interpreting your conversion rate?

1. Define your time period. The holidays are a highly seasonal period and the best way to account for this seasonality is to compare to the same holiday period in 2016.
2. Don’t look at your conversion rate in isolation. Recognize that conversion rate on its own is not a good measure of your site’s performance.
3. Analyze the movement of the variables that comprise conversion rate. How has your traffic shifted? How have your transactions shifted? If both your transactions and your traffic is increasing, you can be confident that your site performed better this holiday season compared to last year—despite what your conversion rate may be saying.

Mobify provides mobile commerce technology to of the Top 1000 online retailers in North America.

In an ongoing effort to understand the most vital business strategies of merchants that sell online, Internet Retailer is taking a deep dive on conversion rate. Help us improve our estimates and category benchmarks by answering the two questions below.