Don't Miss Garden Party - Media & Advertising's Summer Celebration, July 12th: Buy Tickets

How GumGum Plans to Stand Out in a Crowded Contextual Market

Vincent Flood 30 June, 2021 

Amid privacy changes, contextual solutions are gaining traction in the market. GumGum is one such company that provides tools to analyse and place ads based on content across desktop, mobile and CTV platforms. 

VideoWeek spoke with Peter Wallace, managing director EMEA at GumGum, to discuss what sets the company apart in a crowded market, how they intend to use recent funding from Goldmann Sachs, and whether privacy changes mean contextual will become the dominant form of targeting. 

We’re seeing a lot of contextual tools come to market. What sets GumGum apart from other contextual technologies?

Everybody has been talking about the re-emergence of contextual. So it’s important to set the tone in terms of what contextual is today, versus what it might have been in yesteryear. Previous contextual solutions would have been based on domain page level, or on bland keyword analysis. 

These days, contextual needs to be more advanced and people need to understand exactly what’s under the hood. So context for GumGum is about full content-level analysis. That includes natural language processing to understand the sentiment of the text, it includes image recognition to understand exactly on a pixel by pixel basis of what is contained within an image. Similarly, with video, we use that computer vision to understand and analyse the frames of a video, and transcribe the audio to then apply our natural language processing to the audio.  

For GumGum, that level of technology is what really sets us apart from the rest of the market. We are currently going through a Media Rating Council (MRC) accreditation for content level verification, to really rubber stamp the fact that our technology is better than everybody else’s in the market. We will very soon, hopefully, be the only business globally with that accreditation that is applying this technology to the open web. 

We feel like we have the best contextual technology in the market and that we are pushing the boundaries of what is possible within contextual. Today, it’s about looking at the text, imagery, audio and video. We’re already working within the CTV and OTT environments. But we’re always looking to see how we can push the boundaries and take new contextual signals and apply our learnings into targeting and brand safety. 

In April, you raised $75 million from Goldmann Sachs, how do you plan to use the money?

We have two ambitions with the cash. One side of things is continual development of our technology. Keeping constant that element that sets us apart from the rest of the market and making sure that we’re always at the cutting edge of contextual technology and also creativity.

And then on the other side of things it’s about expansion, with mergers and acquisition activity. So we will be exploring the market to see how we can apply that cash for interesting opportunities across the market, whether that be technology opportunities, or whether that be international growth opportunities. Watch this space. 

Brand safety tools often take the form of blocklist technology, how does GumGum ensure brand safety with its contextual technology?

For us, brand safety isn’t about blunt keyword block lists, it is about understanding the full contents of that page. And it’s about not taking immediate inferences from words. For example, from the word “shoot”, we are able to understand whether that refers to a horrific thing going on or whether it refers to a sporting event. We go much deeper into that content than blocklists. Brand safety for us is about using our technology like natural language processing, computer vision and seeing how we can be more intelligent around brand safety.

A good example is the analysis we did around COVID-related content. A while ago we were at a point where brands were blanket blocking all COVID-related content, regardless of the sentiment around that, regardless of the general content of the page. And we’ve seen this occurring in other events. For example, there were instances where brands were blocking Black Lives Matter content. I think, from an ethical standpoint, brands need to consider whether it’s in keeping with their outlook to be blanket blocking certain content. And then, they’re also making it a lot more challenging for themselves to gain scale and appear in places where people are reading content. 

We did analysis on COVID-related content a while back and found that upwards of 60 percent of all content containing COVID references was brand safe. So, in that instance, our technology would have opened up a lot more inventory than other brand safety solutions would have done. We’re really supporting the advertisers on that, and also the publishers because we can also open up an awful lot more inventory and monetise that inventory for them by having more effective brand safety solutions.

After the degradation of the cookie, do you think contextual targeting will be the dominant form of targeting or will it be used in combination with the new breed of cookieless targeting methodologies?

I would need a crystal ball to give you an exact answer on this, but I do think that contextual will make up a dominant portion of how people operate their targeting methodologies. I think, if we’ve learnt anything in the last number of months, it’s that anything that relies on a cookie or any kind of identifier, can be turned on its head very quickly. 

You’ve got GAFA at loggerheads with each other, and I think they’re turning this idea of data and privacy on its head to be a self-serving, money-spinning opportunity for them. With all of that going on, the one consistent thing has been that contextual has never been impacted. 

So do I think that there will be other targeting solutions that are important outside of contextual? Yes, right now, there needs to be a process of re-architecting the industry so that we can even just operate without a cookie. There are big questions remaining about what that user ID solution looks like. So, contextual targeting is definitely at the forefront.

Do I think it needs to be some level of hybrid model? Yes, but I think that’s more with regards to the piping rather than the targeting solutions. But I think there’s a lot that remains to be seen in terms of how a lot of those solutions will actually work and how they will operate amongst the walled gardens. 

What do you think the future holds for independent media publishers? Will privacy changes benefit these smaller publishers or will more money go to the tech giants?

I totally admire businesses like The Ozone Project, who are bringing together publishers outside of the walled gardens. That allows publishers to come up with solutions that feel sustainable and create a little bit more of a critical mass for them to be able to go out to market with significant data, a tech stack, and really be able to tell a compelling story.

I think ultimately publishers need to be smart about the way that they’re going out to market and make sure that they are adding value in the supply chain. And I think a lot of them are doing that, a lot of publishers are looking at how they can work on their side with things like brand safety solutions and contextual solutions, rather than leaving advertisers to deal with everything downstream. They’re adding value in ways that they probably haven’t done before. 

I think that the changes have pushed them to think a little bit more about their solutions. I think it has prompted them to get their ducks in a row around their tech stack. They need to keep on doing that and evolving. And I’m all up for people coming together and collaborating. The more publishers can collaborate with one another, the more value they’ll bring to the market.

Finally, do you have anything else you want to add about GumGum and the future of contextual targeting?

When it comes to the future of contextual targeting, we’re in a very cluttered market at the moment. I think that for a lot of businesses, it’s difficult to see the wood through the trees. Currently, there’s a huge boom in the number of solutions out there. But there’ll be a contraction in those numbers at some point, when the best solutions rise to the top. 

In terms of how contextual targeting evolves and how GumGum sees the world, our ambition is that we deliver ads that make sense to the user, because they’re actually contextually relevant. We also know that they drive value from an advertiser perspective, because of the results that we see. 

But fundamentally, we are not using any personally identifiable information, and we are not using anything which is reliant on a cookie. We want to continue to apply that to different media channels. We’re already working with OTT and CTV, and want to see those through to the endgame. Then, we can start to look at other media channels, such as out of home advertising.

So I think there’s a super exciting space for contextual right now. And again, it’s businesses that can take that contextual understanding of content level analysis and apply it to different scenarios who will be the ones that rise to the top and really succeed. And that’s going to be a key focus for us as a business as we keep on evolving.


About the Author:

Go to Top