The web page content as a medium to target users in the post-cookies era
As cookies are doomed to extinction, marketers are rolling their sleeves up in a quest for a viable alternative. Meanwhile, there is an actual solution. Cases show that tracking users’ real-time content environment can yield similar results to using audience segments.
Over the past 26 years, programmatic buying revolutionized the advertising industry reaching the volume of $100 billion by 2019. All this time the market was largely driven by cookies, which established themselves as marketers’ best friends when it comes to finding target audiences on the web. Data collection methods and attribution models have been improved to such a degree that one can target a relevant user with high precision.
However, recent developments suggest that marketers are soon going to lose their handy weapon. In August 2019 Google Chrome, the world’s largest web browser, announced that it would restrict the use of 3-rd party cookies beginning from 2022. The same decision had been earlier made by Safari and Firefox. As a result, market players are now searching for new ways of getting data about users’ preferences. In the long run, cookies can be replaced by a technology called unique user identifier. Teams at IAB Tech Lab and The Trade Desk are already working on it. Still, there is an already existing alternative which can be quite efficient.
The page content is a key
Efficient user targeting as long as you know what your audience is interested in. Therefore it is crucial to get access to the content they consume. Say, if someone is reading an article about allergy, it probably means that he or she has a seasonal nose itch. Which means that advertisers can discreetly offer them a nasal spray. The ad appears timely and looks like native content. An ad placed instantly on a relevant website before a relevant user may have huge marketing potential. Many market players are already benefiting from this technology. You can also try it by using “Semantica” from hybrid.ai.
How it works
At the first stage, we create a context channel and specify in the DSP relevant keywords and phrases that the page title or its main content should match. Then the algorithm parses all pages from which bid requests are received. If the page is familiar to the system, then it already has data about its contents in the database. Otherwise, it generates a list of keywords, the so-called stemming. In case the page contents match the product category, the system decides to display an ad – and sends a bid response. The whole procedure is simple, reliable and cost-effective.
Is it efficient?
Of course, we understand that the description above naturally raises some logical questions. How much traffic can the algorithm generate? Will it be efficient? Does it fully replace audience data? Let’s work this out.
Semantica in action
Let’s compare two campaigns run by hybrid.ai for a large automotive brand in April-May period 2020.
For the first one, we chose such relevant audience segments as car enthusiasts, people interested in cars and a number of affinitive segments (chosen based on a behavioural portrait of target audience) like outdoor activities, fishing, hunting, etc.
For the other one, we collected keywords and phrases which the system was supposed to look for on the web pages. For the test, we applied the same principle as while choosing segments. Namely, we used identical relevant and affinitive keywords and phrases.
Below is short the list of main phrases per categories. The complete list includes 800+ phrases:
Both campaigns had the same budget and similar settings. We did for the purpose of attaining an objective assessment.
In the end, we have got almost the same results. It means that Semantica managed to bring up a real audience interested in the product. Below are the stats for both campaigns.
It’s also important to look at the efficiency of spending. The average CPM1 for Semantica was 13,2% lower than the average CPM2 for audience segments. The absolute values of CPM are omitted due to commercial confidentiality.
This short analysis shows that Semantica, although not a new tool, but is a proven and reliable solution – both in theory and in practice. Its focus on users’ real-time content environment helps find engaged target audiences and spend advertising budgets with greater efficacy.
Today’s programmatic ad market is geared to collect users’ data simply because they make the ads that hit the target. With the data at hand, marketers can also check various hypotheses about the patterns of behaviour which are typical to the target audience of a given brand. Therefore the downfall of 3rd party cookies is obviously a big blow to the industry. It pushes industry players to search for new ways of collecting users’ data. But the market should bear in mind that there is an already existing technique which can take up this task immediately. Displaying ads based on users’ content environment is reliable and efficient.
