Industry update - Intelligent Tracking Prevention

Related Articles


Delivering connected measurement in a fragmented world

Google Analytics 4

What is data thresholding in Google Analytics 4?


Industry update - Intelligent Tracking Prevention

Hello everybody. Today we’re going to have a quick industry update. We’re going to be talking about intelligent tracking prevention.

Apple implements intelligent tracking prevention

Can anyone hazard a guess as to what this might be showing? Any ideas? This is Safari’s new release of Safari 11. Here, we’re seeing all other Safari traffic and we’re seeing the growth of Safari 11. I believe, at the moment, it’s already got about 30/40% adoption, from what I’m seeing in Google Analytics. It’s growing quite quickly. Why is that significant?

Apple has caused quite a fuss recently with their latest update by implementing something called Intelligent tracking prevention.

What’s changed?

What’s changed? What Apple is now doing in Safari 11 is restricting the way that third party cookies can be used by the browser. Third party cookies are used by the likes of Google and a lot of big display platforms and big data companies, and it’s essentially a cookie that can store and identify for a user, and it can be accessed by the owner of the cookie, for example, Google, as they browse around the various sites. This is hugely useful, there’s a mutual benefit, so I can tag up my site with Google’s tags, and I share my information with them. In return, they give me lots of information about the kinds of people who are visiting my site. I can use platforms like DoubleClick and AdWords to be able to market to them on other sites. That’s hugely beneficial.

Limiting how third party cookies can be used

What the latest version of Safari is now doing is putting a limit on how these third party cookies can be used. If you visit someone’s website, for 24 hours after that, that third party cookie can be accessed by that site. After that, it can no longer be used by platforms that have been recognised as advertising platforms. That is the likes of AdWords and DoubleClick and the various other platforms that we know and love.

You then have a remaining 30 days, which can be used for things like log-ins. Cookies are very valuable for logins. When you go back, and you visit a site again, you’re still logged in. That’s all facilitated by cookies, when you have to log-in with Facebook and log-in with Google and so on. The latest version of Safari will give you another 30 days through which you can use that cookie for logging in. If you don’t revisit the site in 30 days, then they will just delete that cookie.

How it’ll impact consumers and marketers

The important bit to think about here is this first 24 hours. From a marketer’s point of view, this is critical to us. We can only see the cookie within a 24-hour window for the purposes of our marketing. After 30 days, that cookie is gone entirely. What difference would we expect to see from this? There’s a few things. There’s two ways that we can look at this. One is from the consumer’s point of view, and the other is from us as marketers.

Improved privacy

For consumers, there’s going to be improved privacy as a result of these. Apple has made the decision, on the behalf of its consumers, that they will restrict the data that is shared with all of the major platforms around the web.

This could be a better or worse experience. From an immediate point of view, you’d suspect that being able to block one of these ads and not be tracked around the web, brilliant. Users are going to love it. It’s the whole reason behind the rise of the ad blockers.

Ads could become less relevant

At the same time, if the marketers are given less information and there’s less ability to understand more, then potentially the advertising for these consumers can get less relevant. Really, it comes back to us having to do a good job as marketers. We need to be sending the right message to the right people. Basically, not to be irritating. This is one of the challenges we’re going to face potentially with users of Safari, is that without this more rich, contextual data, we can potentially struggle to reach the right people.

It could be tougher to track users over time

In addition to that, you’ve got, for example, b2b, or particularly high-ticket items, typically have a longer consideration cycle before getting in touch with the buyer on the website or buying a product or whatever it might be online. Where these conversion paths might span over several days, we’re going to see it being very difficult to track these users over that period of time. Potentially, we could be losing some of the data quality that we have for users of Safari.

Important information

What do we need to know about this? There’s a few things. There’s been quite a significant backlash from a number of advertisers. Many of them are taking steps to either mitigate against it or being quite vocal about their objections to it, saying that one of the major criticisms has been that Apple is enforcing this upon their users. They’re making the decision about what their consumers want, rather than giving them an option to enable this feature, for example.

Google introduces a first party cookie

Google have done a few things. They’re typically one of the first to act. They’ve introduced a first party cookie. First party cookies, not effected, it’s the third party ones that are effected here. First party cookies are the ones that you set on your own website, but they cannot be accessed by other third party platforms. This is very similar to the Google Analytics cookie, for example. As a first party cookie lands on your website, no other website can access your cookie. If you have AdWords auto-tagging enabled, this will be done for you. Absolutely brilliant.

There’s also a new Google tag manager, a tag that’s just been released. I haven’t had a chance to explore it yet, but keep an eye on that. That will allow you to grab the linker and essentially store things like the gclid, campaign parameters, and access them nice and easily through Google tag manager.

Google has implemented predictive modelling

To make up for the missing conversions as well, we would suspect that if we’re less able to accurately measure these users from Safari, we would see a drop in reported transactions, in platforms like AdWords and DoubleClick, which would reflect the loss of tracking that we have or the loss of visibility we have for these Safari users. To make up for that, Google are looking or have implemented predictive modelling to essentially estimate what that loss has been, and they’re now including those back in with your reports.

Whether that’s a good thing or a bad thing or how accurate their modelling is or how transparent their modelling is is yet to be seen. They are certainly trying to fill in that missing gap that you’ve got. That’s worth bearing in mind when you’re looking at your reports. If you’re not seeing a drop, that’s more than likely the reason why.

Create a segment of Safari users

My biggest recommendation would just be getting whatever analytic platform you’re using. It’s very, very easy to segment by browser and by device, particularly for mobile Safari. It is significantly more used on mobile devices. Create a segment, track it, get into the dashboard, keep watching it for the next couple of months. Keep on top of it.

Explore other methods of user identification

Finally, I think there’s a bigger trend that we’re going to see here. We’ve got rise of ad blockers. We’ve got massive, massive technology, mobile providers – Apple – making these decisions about pushing the industry in a certain direction by enforcing these choices upon consumers. I think we could potentially see more of it.

That doesn’t mean the end of us as marketers. We still have other ways to measure and to infer individual users. These are probabilistic and deterministic user identifications, using other indicators, so Geo IP, location, timestamps, time of day, devices, browsers, browser fingerprinting. There’s all sorts of methods that can be used to either exactly identify an individual, such as when they put an email address in your website, or to probabilistically identify someone. For example, if you see a user using a very similar browser from a similar IP address and a similar location multiple times, you can usually infer that that might be an individual.

These technologies are already around, but I think we’re going to see a significant rise in those over the next six to 12 months or so. Thank you very much for listening.