The Distance Report
This is a new report that recently popped up in the interface living under the dimensions tab. The report looks into how far the user was away from your store based on the distance displayed on your location extensions and you can see how conversion rate and cost per conversion are affected by distance.
It makes sense that users that are located closer to a store are more likely to go in store and not make their purchase online. You can see in the chart below how reality fits the hypothesis. Cost per conversion is higher the closer the user is to the store and we see online conversion rate increasing the further a user is from the store.
The other interesting factor to consider is device type. Mobiles typically have a worse performance than desktop. It is quite common to make the assumption that this traffic represents people browsing whilst out and about, either going in store to purchase or waiting till they are home to use a device where they have more time or are more comfortable making a purchase on their PC (we can prove this concept with the next tool). The distance report helps with this one. Let's look at the same chart as above but segmented by device. We can see mobile users have the highest cost per conversion compared to desktop within 5 miles of store, and after this point we actually see the cost per conversion drop.
For this client in particular the CPC for desktop is 3x higher than for mobile devices and 2x higher than for tablets so despite the conversion rate being lower on mobile devices we do see the CPA drop below that for desktop after the 5 mile point. However, before this point it is higher, matching our assumption that mobile traffic is likely to be people out and about who are more likely to convert in store.
The Estimated Conversion Columns
This is another fairly new feature. Google looks at users who are logged into Google Chrome; they estimate this is about 30% of all internet traffic. When a regular cross device trend is spotted for logged in users, the pattern is scaled to give estimated total conversions. Google need to be 95% sure the user trend is accurate so if an account receives fewer than 50 conversions a day you are unlikely to see data in the cross device columns.
Google estimates the total conversions and total revenue a campaign receives and based on traffic volume and spend works out the following metrics which you can add as columns in the AdWords interface.
- Est. Total Conv.
- Cost/Est. Total Conv.
- Est. cross device conversion
- Est. Total Conv. Rate
- Est. Total Conv. Value
- Est. Total Conv. Value/Cost
- Value/Est. Total Conv. Value
- Est. Total Conv. Value/Click
At first I was a little sceptical to trust anything that has
estimate in the title. However, the data in these columns is based on real user data. After some time using them the results also fit what we would expect: brand campaigns are more likely to see cross device conversions happening whilst numbers coming through generic will be lower.
Using the estimated cross device conversion column combined with the device segment is a great way to analyse how much traffic from each device is moving to another. It is worth bearing in mind that this is across screen. As well as measuring users who are going from mobile to their PC it also gives you the opportunity to analyse the users browsing from PC to PC, i.e. starting on their work computer and then moving to their home PC.
How can we use these results to decide a bidding strategy?
The main use of these tools is to help aid bidding strategies. After reviewing the distance report it makes sense to add some location targeting to your campaigns if you see any patterns emerging. Add radius targeting based on your store locations and depending on how granular you want to go you could add different layers for each location. Add 1 km, 5 km and 20 km targets from store. Initially you could start with a 0% bid adjustment to gather data and see how traffic performs. If the goal is to drive traffic to store you can increase the targeting closer to store. If the aim is to drive traffic to the website you may want to decrease your bids closer to store in order to focus your budget on the areas with the best ROI.
Remember, if you add a mobile and a location bid adjustment to your account they will act as a multiplier. You can use the bid adjustment calculator to see exactly what your adjustment will end up as (look for the calculator icon in the Active Bid Adjustments column).
The estimated columns give us an opportunity to analyse how much more we can push our bids on each device. Google suggests altering bid adjustment based on the percentage difference of the actual and estimated columns. For desktop this means working out the difference between the actual cost/conv. and the est. cost/conv. to calculate how far we can push our bids. For example, if estimated cost/conv is 10% lower than the actual cost/conv, we could look at increasing our bids by 10% to push for more volume within our target ROI.
To work out mobile bid adjustment Google suggests the following formula.
Mobile bid adjustment = 100* [value per mobile ad click/value per desktop and tablet ad click)-1]
This calculation is working on a ratio that should bring your mobile ROI more in line with your desktop.
If using these columns to inform bid strategy, it is always worth bearing in mind that you may not see the same return within AdWords as the traffic could find you again via another route. However, they are definitely a useful tool in proving that
the world is becoming more multi-screen and we should have a rough idea of how our potential customers fit in with this.
Overall both of these tools aid us in seeing the value of all devices, helping us to piece together the journey our consumers are likely taking. We are all aware that attribution is getting more and more complicated so it is worth delving into these tools to figure out what your consumers are doing and asking if there is anything more you can be doing to help their journey.