We've all experienced it. You research a product online and for the next few weeks, you are stalked mercilessly by ads. It's a kind of personalised advertising known as "retargeting" and it can result in ads that continue to follow you long after you have made your purchase.
The best parts of personalisation – getting relevant content, having forms prefilled, one-click ordering and delivery – often go unnoticed because they are almost invisible. They are so smooth and seamless that people take them for granted. But no one forgets ads that follow them around the internet like a lost puppy.
So if retargeting is so annoying, why do brands do it? The short answer is that it works. Marketing loves a good mantra, and you've probably heard the so-called "Three-hit Theory" or "Rule of Seven", which describe how many touchpoints are necessary to successfully reach a customer.
If only it were that simple. As marketers use digital technology to go increasingly granular with campaigns, it becomes apparent that neither mantra is true. The amount of "hits" or "touches" will depend on the product, message and specific journey point that the consumer is in. There are no silver bullets.
But retargeting must work because research shows that around half of the top 2000 websites in both the U.S. and U.K. do it and every marketing blog under the sun will tell you that retargeting improves ROI and increases conversions.
The mechanics behind retargeting are simple. When you visit a website, the company attaches a tiny piece of identifying code called a "cookie" which tracks you and reports your behaviour. The whole process isn't exactly new – ad tech called DoubleClick was founded in 1996 and used cookies – but there have been some innovations in retargeting. For example, retargeted ads now make use of “dynamic creative”, which will adjust the content according to the device you’re on, your location and viewing context.
Retargeting may work, but it has the side-effect of being infuriating and creating a "creepy factor". "Just how much do these companies know about me?" (A lot) "And what are they doing with this information?" (Depends) are typical questions that come to mind.
Lead generating software company Leadpages surveyed 2,000 regular internet users to find out if they were "creeped out" by retargeted ads. They found that the results largely depended on how they phrased the question.
In the first part of the survey, Leadpages asked 1,000 people how they would feel if they saw an ad based on their personal data or search history, and 41.4% responded they would be "annoyed" and never click, no matter how good the deal was. It's a large chunk but still not quite a majority, which points to that fact that for all its shortcomings, retargeting still does work in the long run.
In the second part of the survey, 1,000 people were asked they would feel if they saw an ad simply based on their interests. In this case, only 27.9% said they would be annoyed and never click. The more subtle wording of the second question and significantly different answers suggest that while people might be fine with personalisation when it becomes too personal it can push them away. The survey also found that older millennials aged 25-34 years old were the least likely to be against retargeting, indicating that younger people, in general, are not as prone to being creeped out by highly personalised content.
So how to solve the problem of retargeting gone wild? There are different approaches to this and most of the big players offer variations on similar themes.
One approach is to stop retargeting to customers after they have made a purchase. For example, marketers using Facebook pixel – Facebook's version of a cookie – can capture conversions in a Website Customer Audience and use exclusion targeting to stop them from receiving more ads. To identify converted customers, Facebook suggests that you have them complete a "registrations confirmation page", which essentially says "Hey, I bought this". But the problem with this approach is that it relies on customers performing a specific task and may miss out on the opportunity to continue selling to them.
Thus another approach is to modify the message. In its "19 strategies for better retargeting", Google Adwords suggests offering converted customers who have already purchased items complementary products or services to go with their purchase. For example, if Google can identify that I have already bought a car, it can start serving me ads about peripherals such as tyres or services like good mechanics in my area.
A third solution is for companies to use analytics to guide them. Ooyala is an Integrated Video Platform (IVP) company that is at the forefront of using a data driven approach to video advertising. In a press lunch in Sydney that Brad Arsenault, Fifth Quadrant’s Head of Marketing, attended last month, Steve Davis, Ooyala's GM for Asia-Pacific and Japan, described IVP as "an online video platform plus advertising technology, plus media logistics – all powered by a common data set underneath".
In its State of the Media Industry 2017 report, Ooyala says that data analytics is revolutionising online video delivery by allowing advertisers to amass customer viewing data and translate that into personalised and engaging content.
For Ooyala, a large portion of that data comes from Telstra, which acquired Ooyala in 2014. "Showing a video is not great," Davis said. "Showing a video and learning about your consumer is a lot more valuable. Showing a video for live or VOD and controlling the ad stack behind it? Much more valuable than just showing up random ads that have nothing to do with my experience." When asked how Ooyala knows when someone has bought something, Davis said that the system needs to be much smarter than simply throwing up ads for six months because you visited a page or clicked on something. "The way around it is to put automation and rules behind your advertising and not just blanket advertising, and that's what's happening now."
Engagement analytics is one way in which Ooyala can ascertain the effectiveness of an ad. Giving the example of auto-play ads, Davis said that by examining viewability metrics, Ooyala can tell publishers when their ads are annoying their customers. For example, if everyone pauses an auto-play ad immediately, there is clearly no engagement. To understand these habits, Ooyala provides quartile viewability metrics (0%, 25%, 50%, 75%, 100%) across the adtech stack and Online Video Platform (OVP).
The upside of personalisation is that it can deliver more relevant content and a better overall CX. But advertisers need to constantly be aware of the "creepy factor" and make sure they don't cross the line. Excluding converted customers, modifying messages and using data analytics are some of the ways that companies can deal with the problem of retargeting gone wild.