Top 3 things you should look for in legitimate contextual technology

Many companies are claiming expertise in contextual targeting technology, but not all of these claims hold water says Hailey Denenberg, VP of strategic initiatives at GumGum

The ongoing demise of third-party cookies has brought a surge of interest in contextual technology, and it’s great to see. That’s where we started - back in 2008, and over 15 years later, contextual remains the bedrock of our offering. We believe it’s one of the most powerful tools available to an advertiser, especially in an era in which other signals of interest and intent are being dimmed to the point of invisibility.

Where there’s gold, however, there is invariably also a rush - often led by people who don’t have the knowledge or the tools to extract value from the precious commodity they’re chasing. That’s exactly what we are seeing with contextual right now.

In the context of digital advertising, all kinds of companies falsely claim to have advanced contextual advertising capabilities. We’ve all been on the wrong end of a convincing sales pitch for a service that disappoints, so to avoid getting hitched by one of these companies who can’t deliver on their promises, we recommend you ask a few key questions of anyone pitching you with contextual tech.

1. Transparency

Transparency is critical in ad tech, and the best way to ensure that a vendor’s tech does what it says it does is to review its methodology. If a potential partner does not publicly publish or willingly share the methodology behind their contextual systems, there’s probably a good (or more likely bad) reason why.

Likewise, ask them if they are using their own tech or tapping into someone else's. Licensing a good piece of tech, contextual or otherwise, is perfectly fine and normal, but the vendor should be forward about this and happy for you to take a detailed look under the hood so you fully understand its capabilities.

2. Accuracy

Ask the vendor how they validate the accuracy of their contextual approach. It’s never acceptable for an ad tech company to mark its own homework, so you should ask for evidence that the tech has been independently validated by real-world tests, both at a platform and campaign level. If they cannot provide proof points that the tech does what it claims to do, walk away.

In a similar vein, you should seek evidence that the tech has been audited or accredited by a credible third party, such as the Media Rating Council (MRC) or SOC II. If a contextual provider has third-party validation, that is a key proof point for their capabilities. If they don’t, that should give you pause for thought.

3. Fit for purpose

Finally, take a close look at the tech itself, and at what it is capable of. Is it limited to keyword and text analysis, or can it analyze all digital signals - including text, but also audio, images, and video? Does it leverage human annotators to appropriately label the semantic classification of content? Because even if the platform you are looking at satisfies all the points above, that doesn’t speak to how sophisticated - and therefore useful - it is. Contextual technologies have evolved so much over time; today, they should be significantly more sophisticated than relying solely on keyword and URL analyses.

If the platform you’re looking at is not capable of analysing audio, images, or videos, you should look elsewhere, because, in an era where so much of the content we consume online is not text-based, such tech is not fit for purpose. To tap into the power of advanced contextual, and to drive optimal brand outcomes in a cookieless world, the contextual platform needs to understand all available signals in a digital environment.

Good contextual technology, developed and refined over many years, is a powerful tool in the digital marketer’s toolbox. But not all contextual tech is built equal. The time you invest in putting vendors and their platforms through their paces will be time well spent.

Also published in: The Drum