Whitepaper: Estimated Total Conversions

Word on the grapevine is that once again, it's the year of mobile. While it would probably be better just to call it the decade of mobile and be done with it, there's no escaping that the days of a user searching, browsing and converting on one device are long gone. Mobile is playing a more important role than ever in the user journey, with 90% of users moving between devices to complete a goal.

Google recently came up with a solution to actually keep track of these fickle users: Estimated Total Conversions. This new metric is determined by looking at users logged into Chrome on all devices, and tracking when cross-device or cross-browser conversions are completed. This data is then extrapolated out to all users and an estimate of how many conversions were completed on other devices or browsers is calculated. ETCs are completed by including calls made using a Google call forwarding number viewed on a tablet or computer. With this comes a host of new columns such as estimated cross-device conversions, estimated total conversion value, and many more.

Here I'll be looking at Estimated Total Conversions (ETCs) in several accounts here at Periscopix and seeing how different our data looks when they're taken into account.

Overview

Looking at a selection of accounts across several verticals, a total of 1,250,240 conversions were completed between October 1st 2013 and September 30th 2014. During that same period, ETCs were 1,310,350; that's an estimated 4.8% more conversions completed on a different device or browser than the initial click. Taking those conversions into account, CPA drops from £11.27 to £10.75

Looking at the percentage difference between ETCs and conversions over the last 12 months, it's surprising to see that there isn't a clear pattern. You'd expect that with mobile constantly growing, a greater proportion of people would be completing cross-device conversions than this time last year, but in fact there are peaks in Oct 2013 and June 2014.

Of course this data includes several verticals, and there can be huge differences in mobile behaviour which need to be taken into account.

Retail

Looking specifically at retail, ETCs were 2.2% higher than conversions. While that may sound like a small percentage, those conversions brought in an estimated £271,790.72 in revenue.

Also if we look at how the difference between conversions and ETCs has changed over the past year, we can see very clearly that in the retail sector, there's been a huge growth in
people switching between devices. In fact in September 2014, there were 106% more estimated cross device conversions than in October 2013.

If we delve even further and look at what percentage more ETCs than conversions there were by device, we see some interesting results. By far the biggest difference is seen on mobile. What that implies is people searching on a mobile are then far more likely to convert elsewhere, than someone on a desktop. This makes perfect sense when we think about user behaviour; a typical shopper could be very likely to browse a site on their phone while they're on the go, but then actually complete a conversion later on a desktop or tablet, where they can use the full site.

Travel

The travel industry tells a very different story from retail, with only 0.05% more ETCs than conversions. In fact in the space of an entire year, it's estimated only 16 cross-device conversions were completed.

We can make guesses here about what this implies; perhaps for a purchase as big as a holiday people want to search on a desktop or tablet, where they can easily view more details and photos, and then convert. Or perhaps micro-conversions, such as getting a quote or calling an operator, can be completed easily on one device so there's no need to switch to another.

Finance

If we look at the finance industry, we see that there were 2% more ETCs than conversions, implying that users searching for finance products are more likely to switch between devices before converting than say, someone looking for a holiday, but not more so than someone doing online shopping.

If we take a look at how that behaviour has changed over the course of the last year, we can see that aside from an unusual spike in October last year, there has been a much greater difference between conversions and ETCs recently, which would indicate that multiple device use is becoming more prevalent. In fact since January 2014, estimated cross-device conversions have increased by 293%!

Recruitment

Looking at a selection of recruitment clients, we see the most dramatic difference, with 9.6% more ETCs than conversions. This is a huge difference; it means that essentially 10% of their conversions aren't being accounted for in reporting, simply because they occurred on another device.

If we look at which devices saw the greatest percentage difference in conversions and ETCs, we can see that just as with retail, it's mobile where most people are searching, not converting, and then switching to another device. In fact in this case, mobile saw 49% more ETCs than conversions. It appears that users looking for jobs do a lot of browsing on multiple devices (possibly sneaking a peak on their phone at work!) before finally converting.

Conclusion

So what can we take away from this report? There are several points to consider with this data in particular. Because it's taken from such a wide selection of clients, it's not possible for us to know what the mobile experience for the user is like in every case. Also we can't know what mobile bid adjustments are in place on all the campaigns involved. We're also looking at a variety of different conversion types, some of which may not be available on all devices. All these factors could have an influence on people's likelihood to switch between devices, and as a result, this data. Without the ability to account for all of those potential variables, we simply have to take this data at face value.

Of course the most important thing to remember is that this data is based on estimates. While Google assure us it is based on 90% accuracy, it shouldn't be used as the same concrete foundation for optimisation and strategy as finite metrics. However it's shown us what is undeniably some interesting data. It appears that all the hype is true: this really is the year of mobile. Users across several industries are becoming more likely to complete a conversion on a separate device from their initial search, most likely searching on a mobile and then converting on a desktop or tablet.

So whichever industry you or your client is in, it's essential to look at Estimated Total Conversions when thinking about your mobile strategy. If it seems mobile isn't generating the same direct response, a look at the bigger picture could tell a very different story.

To view the whitepaper as displayed on Periscopix's website, please click here.