Generally speaking, we're not big fans of display advertising here. Of course, that doesn't make us unique. The average consumer hates display advertising. So do many advertisers. People don't click them. Bots do. The whole thing is, in fact, plagued by ad fraud.
That said, traditionally we have had to attach one huge caveat to the above because one kind of display advertising does seem to work fairly well: It's called "remarketing."
Remarketing, in short, is advertising targeted at people who have already interacted with your brand in some way, most commonly by visiting your website. When they visit your site, a cookie is installed on their browser that allows you to track user behavior. This cookie can also allow you to continue advertising to them after they leave your site. For branding, this can be very effective. Someone has visited your site, maybe liked it, but didn't act on it right way for one reason or another. Then while they're reading something on ESPN or the Washington Post, hey, there's an ad for your community.
Even if the user doesn't click the ad, it is something they'll recognize, so they're more likely to notice it, and it can remind them to follow up and call you about scheduling a tour or to go back to the site and look over your floorplan again.
There's just one problem here. Maybe you've noticed it: the cookie. The cookie is what allows you to track the user across the web such that you can advertise to them. But, of course, many internet users are not wild about companies being able to do such things.
It's a give-take, of course. On the one hand, if my phone has location tracking turned on, Google knows everywhere I'm going. That is... very creepy. On the other hand, if my phone has location tracking on and I misplace it, I can use that setting to find it. With many of these new technologies the line between "useful" and "creepy" is hard to define.
Because users are not wild about companies being able to track their behavior, they have taken steps to try and prevent them from doing that. The rise of ad blockers is huge in this, of course, but it's far from the only thing we're talking about. Apple and Google have both talked about building in more robust anti-tracking mechanisms into their Safari and Chrome browsers.
Apple's Intelligent Tracking Prevention
The latest example is a move Apple has taken that they are calling "Intelligent Tracking Prevention." We've already gotten questions from clients about this issue, so if you've run into it, you're not alone.
Briefly, this is what Intelligent Tracking Prevention is: If you are a Safari user, then when a website tags your browser with a cookie, Safari treats that cookie normally for the first 24 hours after you are tagged. All normal tracking and remarketing ads will work during that window. After 24 hours, it essentially sets the cookie aside and does not allow it to be used for tracking or remarketing. The cookie is still there; it just isn't being used to influence what a user sees when they are on the internet. Then, after 30 days, the cookie is removed entirely.
This graphic, provided by Apple, explains it well:
So there are several questions you might have now based on that description. Let's take each of them one at a time.
First, what is actually going to happen as a result of this change?
So cookies like the ones being affected by this change from Apple do two different things for you.
First, this change could affect your marketing data. If you use Google Analytics and have not linked Google Ads to Analytics, you're going to have a problem: The cookie that Ads tags a browser with when they visit is not going to work for tracking after 24 hours. So let's say someone gets to your site via Ads but doesn't take the next step you want them to take for 2-3 days. They will be a lead from Ads, but you won't know it because Analytics won't be able to credit the lead correctly due to Apple's handling of the cookie. If you have linked Analytics and Ads, then Google has already figured out a workaround here so that your data is still reliable. NOTE: If you are a RentVision client, we link Ads and Analytics so all of your data measuring the effectiveness of your advertising is accurate.
The second change is the larger one. We already mentioned it: This change is going to have a significant affect on how remarketing ads work in the Google Display Network. Moving forward on Safari, remarketing ads will no longer display for users who last visited your site more than 24 hours ago. For the first 24 hours after a site visit, the remarketing ads will work as usual. But after that, remarketing will no longer work.
This sounds bad, but it's worth remembering that this only applies to Safari (on which we'll say more in a moment) and it only affects remarketing conversations that happen more than 24 hours after someone visits your site. From what we have seen, even with clients who are very aggressive on Ads, remarketing is still only one slice of their traffic and often not a particularly big one.
For example, in one client case we looked at 1,406 sessions and this is what we found: 301 of the 1,406 visits came via paid search. So 21% of their overall traffic is coming via paid ads. How many of those 301 clicks came on Safari? 102. And how many of the 301 clicks came via remarketing? 24. Only about 7% of their traffic came from any kind of ad click on a Safari browser and only 1% came from remarketing ads—and even here we don't know how many of that 1% of visits came within 24 hours of visiting the site vs how many came more than 24 hours after visiting the site. So the number of visits lost here is marginal and maybe quite marginal indeed.
To be sure, it's not nothing: 24 clicks from people who have already been to your site is good. That's 24 people who are reengaging with your content thanks to a remarketing ad. In the extremely unlikely event that all 24 of those clicks came from people more than 24 hours after visiting your site and all 24 came from Safari users, then you'd miss out on 24 visits. But realistically you're probably missing out on 8-10 visits. It's not great. It's also not a crisis.
When you're driving traffic to a website it's a volume game. You have lots of different channels and the goal is that they all add up to a significant sum. The goal is also to diversify your marketing channels enough that changes in one channel don't destroy your entire strategy.
Who is affected by Intelligent Tracking Prevention?
This is important to remember: As of now, this technology is exclusively used on the Safari web browser. It does not affect Chrome users, Firefox users, or any other browser user.
It is possible, of course, that Google or Mozilla will see this and change their own browser settings to match what Apple is doing. But Google's incentives here are very different from Apple's given how different the revenue models of the two companies are, so don't look for Google to change much. If they do change, they will likely be far friendlier to advertisements or, at least, advertisements in the Google network.
Firefox is another story, of course, and is famously anti-tracking and anti-ad, but they're also a very marginal player here relative to Chrome and Safari.
How common is Safari anyway?
To answer that question, we need to talk about "browser share." Browser share refers to what percentage of all internet users use a specific browser. At time of writing, early November 2017, Safari has a 14.59% browser share worldwide. Not a big deal, right? Well, if we get more granular, those numbers look very different. In the United States, Safari has a 30.91% share.
That said, even here we need to break things down a bit more. On desktop, Safari is small-time: In the United States amongst desktop users, Safari has a 9.84% share—making it the fourth most popular browser on desktop in the USA. Mobile is where Safari wins big. Amongst mobile users in the USA, Safari has a 52.4% browser share!
This, obviously, makes it the most popular mobile browser in the United States. (NOTE: If you click those links and see different numbers, that is because the site is regularly updating the numbers to keep them as accurate as possible. These numbers are accurate as of November 9, 2017.)
OK, so Safari is huge on mobile and small-time on desktop. How can I find out how much this is likely to affect me?
So this is an interesting question. Fair warning though: Answering it will take us into the weeds of Google Analytics.
What you need to do first is list out what you know about your particular community's website traffic. I have looked at analytics data for hundreds of communities and the variance from one community to another is significant. So this could affect some communities more than it does others.
To figure out exactly how much of your traffic is going to be affected, you'll need to do a couple things in Google Analytics.
Begin by going to Audience -> Technology -> Browser & OS. You'll see this report:
So what we're looking at right now is data on what browsers people use to access your website. The big three here are Safari, Chrome, and Internet Explorer, which are supported by Google, Apple, and Microsoft respectively. Safari (in-app) refers to visits that came on apps being used on iPhone where the user hit a link and the link was opened, in the app, via the Safari browser.
So, for example, if you are on your iPhone and open a link from inside the Facebook app, that click would be tracked as "Safari (in-app)" in Google Analytics.
What we know, so far, then is that in the above case around 46% of all visits to this site came in via Safari. But that doesn't help us with our remarketing problem. So we need to make some adjustments to the report to give us the data we want.
To do that, we'll first define a new segment for the report. Then we'll add a secondary dimension to sort the data.
In order to make a new segment, you'll need to scroll up and click "+Add Segment" at the top of the report:
Once you have done that, select "Paid Traffic" and click "Apply."
After we have defined our new segment, we'll see this report:
This is a little closer to being valuable. Now we have isolated only our paid search traffic. Now we see that Chrome is actually our top driver of traffic and Safari is about eight percentage points behind. That said, we still don't have the most important piece of data: What chunk of that Safari paid search traffic is specifically from remarketing ads?
To answer that question, we need to add a secondary dimension:
Now we need to tell it what we want the secondary dimension to be. If you type "campaign" into the search field, you will be able to select the "Campaign" dimension with the Advertising category. Do that.
The resulting report will show you what types of campaigns all your paid search clicks came from sorted by browser.
The relevant result in this case is marked with the red box.
So: In this particular case, only about 5% of all paid search traffic is going to be affected by the change. Even here, we don't know what share of that 5% clicked a remarketing ad within 24 hours of visiting the site (in which case they could still see the ads now) and what share saw the ad after 24 hours. So it probably isn't even a full 5%. It's less than that.
To be sure, this isn't a "nothing" amount of traffic. It's 230 visits and 230 visits from people who have been to your site before and, with their click, are saying "I want to visit again." So these are high-value clicks. That said, they are still only 5% of the paid clicks and only 1.1% of total clicks to the website.
Is Google going to do anything about this issue?
Here's the big thing to keep in mind here: This move basically won't affect Google or Facebook. Why? Well, it's that 24-hour window. How many people who use the internet regularly are not logging into Google or Facebook at least once every 24 hours? Because of how frequently people use these two services, they will both basically always be within that 24 hour window with users. So they can continue to track user behavior basically as they always have.
Google has rolled out a cookie update to make them compliant with Apple's changes, such that their reporting data in Google Analytics is still trustworthy. But that change aside, neither of the two huge internet giants that you might expect to be concerned by this change are actually going to be hit by it in any kind of real way.
Conclusion
Admittedly, this is always what we say when a new change rolls out like this one, but we say it because it's true: You shouldn't panic. You should gather data, assess how the change affects you, and develop a plan that is sufficient to address the problem.
In most cases, you are not going to see a significant drop in traffic due to this change both because it is only on one browser and because remarketing traffic is a powerful but limited marketing channel. Seeing such a big change to that channel is not great, obviously, but it also doesn't need to become a crisis. You may not need to make any major adjustments to your marketing in many months. If you do see issues starting to bubble up, then you can take intelligent steps to counter them: Put a bit more money into defensive advertising or see if there are opportunities for growth on Craigslist, for example. The point is you always have options. So assess the problem accurately and then proceed reasonably. And if you have questions, we're here for you.
Comments