The imminent death of third-party cookies, and the increasingly likelihood of legislative crackdowns on hyper-targeted advertising in the US and Europe, are helping drive renewed interest in contextual advertising.
But contextual advertising was already seeing something of a resurgence prior to any of these trends, as new companies used emerging technologies to develop much more sophisticated forms of contextual analysis and targeting. Siham Fadili, co-founder and CEO of Bliink, one of this new breed of contextual specialists, says contextual targeting’s value stretches far beyond replacing cookie-based behavioural targeting.
Fadili spoke with VideoWeek to discuss how Bliink differentiates itself from competitors, whether progress is being made on bogus brand safety blocking, and why France is home to so many successful ad tech companies.
There are a lot of contextual advertising solutions on the market, all seeking to capitalise on the industry’s move away from cookies. How do you differentiate your technology from the competition?
If you want to be in the game these days and replace what’s been the industry’s favourite targeting tool for years, you must be early and have robust tech.
Last year, 76 companies claimed to have cookieless solutions. Bliink was founded in 2018, long before Google’s announcement, and before it was cool to talk about cookies. We chose contextual targeting, the most advanced and efficient cookieless solution, to help brands and agencies reach their audiences in this increasingly fragmented media landscape.
Our recipe for efficient targeting is based on artificial intelligence combined with sophisticated proprietary machine learning, which lets us understand what audiences are reading and watching beyond what’s shown in keywords.
The important thing is not content identification, which is already a job that anyone can do. Rather, the most interesting thing is comprehension, where we train our machine learning algorithm to understand content the same way a human would do.
For the past four years, we have been able to provide brands and agencies with very precise targeting segments and reach their audiences where it makes sense, using this robust technology. We can provide over 1700 segments that echo the 350 IAB generic categories.
The demise of the cookie is obviously helping to drive the adoption of contextual advertising solutions. Do you see contextual solutions as an alternative to more identity-focused solutions, or do they work alongside one another?
There are several cookieless solutions available, which is great for our market.
Some of these offer behavioural targeting, where you base your targeting strategy on what the user has previously done via pixel or cookie, whether anonymously or not.
Firstly in this targeting model you have individual IDs (a unique ID, often a cookie in the current system). However, it is extremely difficult for brands to design and reconcile consumer paths as these systems generate a new ID for each new website visited by the user. Additionally, the solutions are frequently reliant on IP addresses, and we are all aware of their limitations (for example several machines can have the same IP, you can have one IP for three different people in the same house etc). These solutions are effective, but you must frequently choose between reach and precision marketing.
Secondly you have cohorts, which allow you to target groups that exhibit similar behaviour. The best example here is Google’s Privacy Sandbox. Google proposes two products: ‘Topics’ (Google’s replacement for FLOC) which is based on user navigation and classifying people’s behaviour into 5,000+ people groups; and ‘Fledge’, which creates groups based on user activity (such as add to cart etc). The recent tests have not proven the efficiency of these two solutions (since they’ve used groups of less than 5,000 people, so not as anonymised). And Google’s solutions are highly contentious, because you can’t judge and be judged.
Then there’s the contextual opportunity, where your targeting strategy is based on what the user is looking for. Here, we identify and understand what consumers read or watch in real time, allowing us to provide a more relevant and engaging advertising experience. The key difference with contextual is that consumers are more receptive to your message because it is aligned with the content they are consuming. Whereas other solutions are still being tested, contextual has the advantage that it’s already proven its efficiency.
How effective is contextual targeting when it comes to video advertising?
In 2021, we spent 100 minutes per day per person watching video. That’s one month’s worth of days and nights! Usage patterns have shifted, so ad tech must adapt.
By clicking on the play button, the user expresses their desire to watch the video – their attention is at its peak, and they are receptive and ready. So it’s very effective to provide them with an ad experience that corresponds to their needs, what they like, at this specific moment of their consumption. With our Capture product, we’ve combined the power of video instream with the precision of contextual intelligence, and we’ve registered an increase of 34 percent in view-through rates compared to non-contextual video instream campaigns.
In this age of advertising fatigue, ad blockers, GDPR and privacy requirements, contextual brings more sense to advertising. It recreates a link between ad messaging and consumers, and has a genuine positive impact on their sentiment towards the brand. We know that consumers are 69 percent more likely to view an ad if it’s contextually relevant, and 44 percent have tried a new product after seeing a contextually targeted ad. As a result, the future of advertising is now, not in 2023, and it is contextual!
Publishers have struggled when brand safety tools have blocked certain keywords, often on content that is in fact brand safe. Do you think the technology is catching up to solve this problem?
We estimate that 30 percent of the ad waste is due to unsophisticated algorithms when it comes to brand safety, where 30 percent of content is identified as unsafe, when it is actually high-quality content.
That is why you must invest properly in tech in order to create a robust and advanced machine learning algorithm capable of identifying the meaning of a word in a specific context. For instance, imagine an article titled ‘The Death of Vinyl’. Unsophisticated brand safety tech would consider this article unsafe for brands. But of course in reality it is completely safe, and in fact would be a very strategic placement for a music brand like Spotify to advertise on.
When you manage millions of pages, you must be able to spot this difference. And that is only possible if you firstly have strong machine learning (a third of our team are developers), and secondly have direct relationships with publishers to solve these kinds of issues. This involves partnering with them on evolving their website code and metadata in order to requalify the content and create value for publishers, brands and audiences.
There is still much to be done, but we are very pleased that the disappearance of cookies has helped raise awareness of some of these issues, creating a more virtuous ecosystem that places content and consumers back at the heart of brand strategies.
France is arguably the most successful country in Europe when it comes to ad tech companies that truly scale. What do you think France is getting right?
In three words: talent x cash x tech.
The advertising industry had an eventful 2021 in France and poised to continue excelling in 2022, with five different start-ups raising a total of €1.7 billion in January. France is the new startup nation. And the growth of European start-ups gives birth to an increasing need to hire talent, especially within the pertinent skillsets. France has become a hothouse of startup talents.
We’ve succeeded in adapting the finance model to ad tech with an increasing number of talented engineers who are loyal and of high quality – especially mathematicians. Previously, many would have chosen a finance career. With the crisis, financial scandals and the rise of startups, today many instead pursue a career in ad tech. Especially since there are lots of similarities between the stock market and advertising.
The French government is also very supportive, launching several programs to attract and retain talents. We are benefiting from over 50 great initiatives designed to fuel the growth of nest startup accelerators across the country: French Tech Next 40/120; France’s growth-stage program which was built to provide unprecedented support for the 120 fastest growing companies in France; our R&D tax credits, dubbed the most attractive in Europe by the OECD, make launching a deeptech startup or investing in R&D even easier; French Tech seeds, a 400M EUR match fund, allowing France to co-invest and encourage angel investing in hundreds of young deeptech startups; and the French Tech Visa, a simplified process for non-European employees and investors who want to come to France… to name a few.
And of course, all of France’s successful ad tech companies have one thing in common: robust proprietary tech.
What’s your greatest bugbear with digital advertising today?
Competition! We’ve seen many companies taking the stage on cookieless, making it difficult for brands and agencies to separate fact from fiction. Standards and rules will be required in this new world.
Can we also talk about GAFA’s slice of the cake? In two years, YouTube’s revenue has doubled. When you choose YouTube, what kind of context do you offer to your brand?
