Contextual Signals in Performance Campaigns: Interview with Kenneth López Triquell, Seedtag

In association with Seedtag.

In this exclusive interview, Kenneth López Triquell, global head of performance at Seedtag, discusses how contextual signals can inform marketers on performance, what tools can help to maximise contextual advertising’s efficacy, and how the industry can assure marketers on the effectiveness of contextual data over user-level signals.

How can contextual signals aid marketers in terms of performance following the deprecation of display and mobile identifiers? How can contextual aid in optimising KPIs?

Contextual offers advertising solutions that are based on anything unrelated to who is behind the screen. Over the last few years, the industry has relied on cookies and audience targeting — everything about you as a person, or how you behave online. Contextual offers a remedy to this, as it diverts focus away from the individual and onto the web content. 

We’re able to do this using AI and machine learning models, which can analyse web content being viewed in real time and classify it into different contextual categories. These signals can then all be used by marketers to deliver ads, and the whole process works independently of display and mobile identifiers. Through our technology, clients can use contextual data exclusively to serve ads that are timely and relevant to users without compromising their privacy, which is really important.

How can locational data be used to maximise the effectiveness of contextual advertising?

Location is a key component when considering whether it’s worthwhile to make a bid. For example, if a person is reading about fashion, and their location shows that they are in London, where it’s been super cold recently, we can better understand the kind of clothing they may be interested in. For that specific user, it may be most relevant and appropriate to serve an ad relating to warm clothes, such as H&M’s winter campaign. In assessing a user’s location, and considering the implications of where they are, we can much more effectively ascertain which ads to bid on and when.

How can artificial intelligence (AI) and machine learning (ML) solutions be used to bolster the performance of contextual advertising campaigns?

Kenneth López Triquell, global head of performance, Seedtag

Artificial intelligence can be effectively leveraged to understand and categorise who your audiences are. The technology we use at Seedtag analyses text and images of the page that these users are engaging with in real time, and we combine these signals to understand the content from a human perspective and classify it into categories. 

Through this, we can offer more real-time targeted segmentation in order to create a contextual profile suited to our clients. And it does this on a significant scale; our technology can analyse about 60 million articles a day, a task which would take a human 228 years! So, these data analysis capabilities combined with our machine learning helps us to bolster the performance of our advertising campaigns.

How can page-level data such as keyword/semantic analysis be leveraged alongside network-level signals to maximise the performance of contextual campaigns?

Keyword and semantic analysis allow us to target the relevant spaces in which to find in-market audiences for our clients. Our AI carries out network-level analysis which allows us to analyse a whole “universe” of content across countries and languages. This allows us to go beyond predefined categories and leverage custom machine learning models to even create bespoke contextual categories. 

As a campaign evolves, the results are fed back into the AI, and the audience models will automatically expand, while the ads will still only appear in the relevant spaces and will resonate with consumers in a similar way to the spaces the campaign appeared in the previous day.

From this, we analyse which spaces are getting the most traction and the best results. The models that we use utilise data from the previous seven days, and we can combine this network analysis with the takeaways of the campaign to understand where to find our clients’ in-market audiences.

What more can be done by the industry to reassure marketers as to the effectiveness of running performance campaigns underpinned by contextual data, as opposed to user-level signals? 

The ad tech industry is still very reliant on cookies and, at Seedtag, we feel that it's vital to be actively communicating that the market is changing and that it’s changing quickly. Google’s deprecation of cookies is on the horizon, and the reality is that 50% of internet users are already accessing the web without cookies. Right now, the industry is in a really educational phase, where we are not only explaining that a replacement for cookies is necessary, but actually advocating for contextual as a better and far less intrusive way of targeting audiences for advertising. Through context, we can use different data points to better understand internet users' activity, the content that they are consuming in real time, and determine whether they will take action or not. 

The industry is already living in the final hours of a cookie-based world; we need to make clients understand that by investing in contextual now they can transition to a more effective and privacy-centred method of targeted advertising before the cookie crumbles for good.