If your mobile campaigns keep failing, it could be time to rethink your tracking
For the first time this year, mobile usage has overtaken desktop web access in the US. This is a scary thought; if we can’t sell things to people when they are using their phones the ad industry could have a major problem on its hands.
Time to rethink your mobile web tracking?
Of course, this can’t be true. Mobile advertising does work. People spend a huge amount of time on tablets and smartphones. During that time they are exposed to a wealth of branded content and ads, then subsequently make purchases. They must be influenced by these ads, so there has to be another problem. There is. It’s called Tracking.
The majority of our clients don’t want to pay the fee Apple charges for in-app sales, nor do they want to develop separate solutions for iOS and Android, so here I’m addressing the problems associated with mobile web tracking.
Cookies
Attribution – assigning value to the specific ad exposures or clicks that are driving sales – has been a hot area in digital for a long time. To perform attribution work, we must be able to link the impressions or clicks to the conversions. Cookies have traditionally provided this link.
Cookies are small, encoded files that are saved by your desktop or mobile browser when you see an ad or click on it. Your browser’s cookie folder holds all sorts of information on the websites you’ve visited and the ads you’ve seen.
For example, you might see an ad for TVs at Amazon on the New York Times website. You don’t click on it, but your browser saves a cookie. Two days later, you then go to the Amazon website on the same device and buy a TV. When you complete the purchase Amazon will check your cookie folder to see whether you’ve seen any of their ads. It will then make a record of the ad (or ads) that you’ve seen or clicked on in a set time period and attribute a value to each ad placement.
How cookies worked
In the pre-mobile world, this worked very nicely based on four assumptions:
- People mainly accessed the internet through a single device.
- People mainly accessed the internet through the same browser.
- All major browsers allowed cookies to be saved unless a user changed the default settings.
- The average person hardly ever deleted their cookies.
Why cookies don’t work any more
Now, in the mobile age, things are different:
1. People access the internet from an ever increasing number of devices.
Result: Cookies can’t track if someone sees an ad on one device and then makes a purchase on another.
Solution: Companies like Drawbridge or Tapad analyse anonymous browsing data to make a connection between different devices used by the same person. They don’t ‘track’ in the traditional way; when a device is served an ad they use behavioural identifiers (location, time of day, past behaviour) to match multiple devices to an individual user. These solutions enable us to link exposure on one device to conversion on another.
2. People now switch between mobile apps (the Facebook app, the Twitter app, the Google app) and the mobile Web (Safari, Chrome).
Result: Single browser cookies or app-based tracking is fairly useless as people may be exposed to an ad on one browser (e.g. within the Facebook app) then convert in another (e.g. Safari).
Solution: We can combine a variety of tracking solutions. We can drop cookies on the landing pages instead of where the click happens in case people jump from one browser (e.g. an app) to another (e.g. a mobile browser).
There are also cookie-less options: Facebook and Twitter track user IDs instead of cookies, which means that they can match users across devices within their own platforms.
Medialets has developed Servo, a dedicated mobile ad-serving platform. It uses behavioural signals to identify a single device, multi-browser user in the same way that the solutions mentioned above distinguish multi-device users.
Servo identifies devices with an 86 per cent accuracy based on 120 signals such as location, WiFi network and response time. It can perform attribution across all the boundaries on a device: app-to-app, app-to-web, web-to-app, and web-to-web.
3. Some desktop and mobile browsers, including Safari, block cookies from ads (called third-party cookies, as they don’t originate from the website you are viewing, they come from third-party ad servers).
Result: View-based tracking can’t be trusted. Customers may be exposed to ads, then make a purchase using the same device and browser. The sale won’t be attributed to the ad exposure as the browser blocked the cookie when the customer saw the ad. This is usually an important part of attributing value to advertising.
Solution: Servo’s cookie-less tracking also solves this issue. A large UK retailer was able to see a 7x performance improvement on conversions by using Servo to optimise their campaign mid-run based on attribution.
Some ad servers, such as Trueeffect, now use first-party cookies. Safari doesn’t block first-party cookies as they are used to remember passwords and website preferences.
They can also be used to store ad exposure information, which can then be used in the same way as third-party cookies to attribute conversions to specific exposures.
4. The average person still doesn’t delete their cookies, so we’re safe here!
A big problem
The details of cross-platform tracking will cause most people’s eyes to glaze over, but it is probably the biggest problem in digital advertising today.
If your mobile campaigns don’t seem to be performing, it’s easy to point the finger at the medium, but it may be that your tracking solution isn’t up to scratch.
It might be a relatively small problem for you now, but as more and more of your customers ditch their laptops and start transacting on mobile, it’s essential to understand how to accurately attribute value to mobile advertising.
Simon Mansell is CEO of TBG. You can find him on Twitter @simonmansell