Re-targeting is often touted as one of the Internet’s success stories.
The ability to track a user’s Web behaviour based on cookies means that if I have shown an interest in a product on a retailer’s Web site, the retailer can continue targeting me online with advertisements for that product long after I’ve left its site and headed for someone else’s.
Note that word “long.”
I’m not sure whether the average Web user realises what’s happening when they start seeing ads on one Web site for something they looked at a few days ago on another. And if they do realise, whether they think it’s clever, useful, or just plain spooky.
Re-targeting is less established on mobile, partly because of the fact that tracking-by-cookie isn’t possible. But ad tech companies are finding ways around the problem, so it is happening.
The problem with re-targeting, in my experience, is the lack of intelligence to put an end to the re-targeting when the ad is no longer useful to the user, and, in fact, becomes nothing short of annoying.
I’ve experienced this several times, most recently, when I was treating my family (OK, myself mainly) to an outdoor table tennis table. I’ve wanted one for years. As we’re having a new patio laid, I figured there was room for one, so I started doing the research for it while on holiday a couple of weeks ago.
I Googled “outdoor table tennis table” to find some online sellers. Out of all these, for the model I was looking at, Amazon, interestingly, was by far the most expensive, wanting almost US$330 more than I eventually paid for the table at a Decathlon store on the Lakeside complex in Essex.
Let’s put a timeline on this: I bought the table the day after I returned from holiday, so the purchase date was August 7. A search in my browser’s history reveals that the online research started five days before that, on August 2.
Since I bought the table, I’ve found myself chased round the Web by re-targeted ads from two of the sites I looked at when researching the purchase. The first is Amazon; the second is a site called table-tennis-tables.co.uk.
I was still seeing the Amazon ads, from memory, up until last Friday, August 15. The other site was still targeting me this week, 18 days after I first started looking at table tennis tables, and almost two weeks after I bought one … and lost all interest in looking at ads for them.
I’ve seen both ads several times – 10 each at a guess – and not clicked on either of them. Why would I?
I’m guessing that if I had bought the table from Amazon or table-tennis-tables.co.uk, the re-targeting would have stopped the instant I hit the “Buy Now” button.
But because I bought it elsewhere, and not even online, they kept on coming — the Amazon ad only serving to remind me what a bad place Amazon is to shop for table tennis tables.
I ran this theory — that if I had bought from one of the sites, they would have stopped retargeting me — past Russell Buckley, who, as a veteran brand marketer and the first employee of mobile ad network AdMob, which Google bought for $750 million, knows a thing or two about advertising.
Firstly, he reminded me how re-targeting works, online at least.
You see an ad and you click on it. That places a cookie on your PC (as long as you don’t clear the cookie cache) and the ad server continues to serve that ad until told to stop. That could be frequency capped to x number of servings or to a period of time.
What it doesn’t know is who you are. So, if you buy the table online and it doesn’t pick up that you’re a returning customer via the cookie, they don’t know that you’ve bought it. Same applies if you make the purchase on another PC, or tablet or mobile phone, or if you cleared your cookie cache.
So in answer to my first question: No, there’s no guarantee that either site would stop targeting me with ads for the thing I just bought from them.
How about the next scenario: If I had bought the table online from someone else, the retargeting advertisers would have no way of knowing, so the ads would keep coming. Correct, says Buckley.
How about if I bought the product from one of the advertiser’s stores?
Presumably the ads might stop then if their in-store POS was wired into their CRM system. Unlikely, says Buckley, unless the store could identify me from the cookie when browsing the Web and so recognise the fact I am the same David Murphy who just bought the product we are targeting him with from one of our stores, and, therefore, no longer in the market for the product.
Finally, I reasoned, if the advertisers had any sense, they could frequency- or time-cap the campaign so that after a few days or instances of serving me the ads, and me not clicking on them, they would stop. Correct, says Buckley.
So what can we learn from all this?
Firstly, that re-targeting is a much blunter instrument than it might at first appear.
Secondly, that there’s little real intelligence behind it, other than a basic premise of “he looked at this on our site but didn’t buy, so let’s bombard him with ads for it until he does.”
And thirdly, that if you’re a publisher who wants to let advertisers re-target people on your site, you should set strict guidelines around frequency- and time-capping, particularly with respect to users who see the ad and don’t click on it.
I assume both the Amazon and table-tennis-tables.co.uk campaigns are, or were, capped one way or the other, but speaking as the target of the re-targeting, the cap was set much too high.
I’ll be keeping a watchful eye for them over the coming months or weeks to see how long they carry on stalking me, and wasting their ad budgets.