Third-party cookies may no longer be set to disappear completely (unless the W3C has its way), but signal loss is still a major issue for the ad industry. Outside of Chrome, several major browsers including Safari and Firefox block third-party cookies by default. And even within Chrome, there are plenty of cases where users browse via incognito mode, choose not to give publishers their consent to process their data, or use third-party tools which block data collection.
So what’s a marketer to do when they’re given a chance to show an ad to an individual they don’t have any data on? Well, they could just ignore the opportunity completely, as often seems to be the case given the frequency with which we hear stories of advertisers avoiding Safari.
Alternatively they could, you know, just ask people what they’re interested in?
That’s essentially the logic behind Welect, a German ad tech business founded in Düsseldorf back in 2016. Welect’s solution is fairly straightforward: audiences are shown a selection of possible ads within Welect’s format, and asked to choose which ad they want to watch. Olaf Peters-Kim, Welect’s CEO and co-founder, says that the format not only provides a neat solution for targeting ads when no signal is available, but also creates its own signal which can prove very valuable for marketers.
Putting audiences in control
Peters-Kim says the format reflects the fact that internet users are accustomed to being in control of the content they see. “Today if you’re on the internet, you select what you read, what you watch, what you click, what you forward to others, how you consume videos, how you consume games, and so on,” he said. “Users are already used to this world where they choose what’s interesting to them, but advertising has lacked this opportunity. So we’re just adapting this user behaviour, which they’re already living, to advertising as well.”
So instead of choosing which ad to show someone, Welect presents audiences with a choice mechanism which sits within the video player. When the ad slot loads, users are shown previews for a small number of ads (usually four, depending on the size of the screen), and click which one they want to watch. The ads can appear in standard in-stream ad slots, where they can be implemented through the SSP or directly through an ad server. Welect also offers a format which can appear in a publisher’s ad block dialogue screen, in the screen it shows users when they don’t consent to data processing, or in a paywall dialogue box where watching an ad can be offered as an alternative to paying for an article.
To ensure that the selection of four ads is likely to be relevant for the viewer without using any personal data, Welect uses geographic and contextual data to narrow down the selection. It also has an AI engine trained on Welect’s seven years of running campaigns, which looks at any data points around the viewer’s personal context (for example, day of the week and time of day) to help pick suitable ads. Welect also ensures competitive separation within the selection, meaning there will be only one ad option for any given product category.
By asking individuals to self-select which ad they would like to watch, Peters-Kim says Welect gives advertisers a way to ensure their ads end up in front of relevant audiences even in cases where no other signals are present.
“Signal loss traffic is quite a significant volume of the daily ad impressions on the internet, especially in the open web,” he said. “You have these situations where people sitting in front of a screen can’t be addressed or identified with your typical ad tech stack. For publishers, that creates an issue where they cannot fully monetise their total volume of traffic, and for advertisers, it’s a competitive disadvantage if they’re only reaching 50 percent of their potential consumers”. And given, as already mentioned, plenty of advertisers choose to simply avoid Safari traffic, there’s a big opportunity for brands to target environments which their competitors might be actively steering clear of.
But Peters-Kim says that the choice mechanism isn’t just a solution to signal loss. It provides, he says “what we believe is the most valuable and the most honest signal you can use in online advertising, which is real-time intent”.
Real-time intent
A lot of ad targeting is based on audiences which have been built on data collected, synchronised and matched over a fairly long period of time, and not all of the data sources involved are necessarily geared towards advertising. “As an example, I could build audiences based on your Netflix data,” he said. “The data quality would be fantastic, and would be very helpful for creating a playlist for an evening in front of the TV screen. But will it help me predict which ads will have the highest relevance for you in the next ten minutes? I don’t think so.”
That’s not to say that Peters-Kim is dismissive at all of the various data sources used to build audiences. “We’re not saying that the data is wrong, or that the targeting or audience building companies are doing a bad job,” he said. “We’re saying that from a scientific point of view, it’s really difficult to take all of this data which comes from secondary sources, match and synchronise it many times, and then expect it to enable perfect targeting.”
Peters-Kim says the ‘real-time intent’ signal provided by Welect’s choice mechanism meanwhile gives an accurate view of which products an individual is actively interested in hearing more about at any given moment. The self-selection mechanism also lets brands reach relevant individuals who might not be included in the types of audiences they’d be likely to target otherwise.
He gave an example of a recruitment company which was trying to reach young people looking for their first job after leaving school. Using Welect’s format, the company found that a lot of older people outside of that target demographic were choosing to watch those ads. But those viewers were relevant for the brand: they had children, grandchildren, nieces and nephews who they knew were looking for their first jobs, and were forwarding them relevant opportunities.
“If I was a planner for that recruitment campaign, I would probably just want to target that ad to the 18-20 age group, and I would miss a whole other relevant audience,” he said.
Adaptation required
The opportunity, according to Peters-Kim, stretches beyond impressions where no other signal is available. But for pitching to publishers, signal loss helps get Welect’s foot in the door.
“Usually the way we start working with publishers is we say to them, ‘we know you have a share of your traffic which you’re either not monetising, or you’re monetising it with really low CPMs, so give us that traffic and we’ll generate more revenue on top of it’, so it’s a really easy pitch for us,” he said. Once publishers get used to the product, they might then start looking to work with Welect across a broader set of their inventory.
On the buy-side, the pitch is more difficult. While there’s a big competitive advantage to be had in reaching audiences in traffic which other advertisers are avoiding, marketers don’t necessarily feel the same urgency which publishers do.
“An advertiser or agency might be booking 100 million impressions per month, they have their budgets and they look at their dashboards and see that those impressions are being delivered, so everything seems fine,” said Peters-Kim. “So we have to explain to them that those figures look fine, but they’re only missing out on a lot of traffic which their targeting tools and tracking tools don’t see, they’re only fishing in half the pond.”
The nature of Welect’s format also requires a bit of adaptation on the buy-side. Measurement is obviously crucial to advertisers, but third-party measurement tools have also underserved signal-free traffic (though Peters-Kim says progress is being made). And ultimately, Welect has to persuade planners and buyers to invest in a format that doesn’t necessarily fit their standard frameworks.
“It requires planners and people responsible for media to change their mindset,” said Peters-Kim. “Currently everyone has their lists of audiences, or a list within their DMP which uses cookies and other identifiers. And they decide which people should or shouldn’t see an ad, based on those audiences or those identifiers. That’s totally different from our approach, where potential customers are the ones who decide if they want to see your ad.”
But again, Peters-Kim says this friction creates an opportunity for advertisers. For those who are willing to take a different approach, the competitive advantage is all the greater so long as their competitors remain unwilling. “If you’re one of the few car companies, for example, displaying your ads to 100 percent of potential buyers, you’re immediately spending your money more wisely than those who are still fishing in only half the pond.”
Follow VideoWeek on LinkedIn.
