Advertising Week Q&A with Terry Hornsby, Founder and Executive Vice-President, Mantis

What has had the biggest impact on brand safety in the last five years?

Covid and the pandemic have had the single largest impact on brand safety and how the advertising ecosystem views it. Firstly, because it was unknown. No industry or person had experienced anything like it before. It touched every part of our lives – sporting events, work life, our freedom to go out. Secondly, because it changed how people consumed content and shopped. Streaming services saw rapid growth and every single business, whether they were a national retailer or local bakery, had to pivot their offering to be online.

During this time, brands unsurprisingly worried about their ads being placed near news stories covering the negative impacts of the pandemic – but nearly every piece of content produced at that time made reference to covid. Putting keyphrases like ‘covid’ or ‘coronavirus’ on a blocklist would have cut brands off from a majority of news stories, many of which were completely brand suitable. This is when we began to see the conversation around brand suitability pick up steam and it was also the point I knew that we needed to rapidly develop our brand safety tech to tackle this issue.

How far do you think the advertising industry has come in shifting its attitudes to brand safety in the last five years?

With advertisers, not very far unfortunately. There’s not been a large shift away from the norm of very blunt brand safety keyword blocking and a large amount of caution remains. However, I think the industry is starting to understand the difference between safety and suitability, and this will be key.

Safety is no longer up for discussion – that’s table stakes. Slowly as an industry, we’re understanding that safety is a fixed thing – terrorism, weapons, and violence, for example, are always going to be a universal safety issue for brands. But the challenge is that this binary thinking has been mirrored onto the suitability piece as well. With suitability, there is some content that a brand would consider safe and suitable, but these articles have been blocked from advertisers due to blunt tools. A different brand however may consider that content unsafe and be happy to have it blocked. This lack of nuance or personalisation in these tools is severely impacting performance.

The Olympics highlighted this perfectly. If an athlete ‘bombed’ out, or a starter ‘pistol’ misfired, blunt safety tools would immediately block brands from appearing next to a perfectly brand safe story. One of the breakout stars of this year’s Games was Türkiye’s sharpshooter Yusuf Dikeç – but brands could have missed out on the viral moment because of the very sport he competed in being on a keyword blocklist. This has been exacerbated during the Paralympics, where over-cautious blocklists can isolate any articles that mention disabilities.

Publishers have been far more forward thinking however. More of them have adopted our technology, which is in turn generating the data to show advertisers that it's effective. But we have to move as a unit if these tools are to be effective and there needs to be more opt-in from the buy side.

What is your proudest achievement at Mantis so far?

Creating an environment during Covid where publishers could monetise their inventory in a safe way. A significant portion of news stories had, unsurprisingly, a word relating to covid in them – everything from hard-hitting reporting to news of local restaurants reopening and restrictions lifting. Being able to create a solution that stopped publishers from missing out on revenue opportunities and allowed brands to continue to effectively reach audiences was really powerful and I’m proud to have been part of that.

What has been the biggest challenge during your time at Mantis?

I think we’ve encountered the same problem most start-ups do – changing people’s minds. No one gets fired for hiring a well known business or a tool that has been in the market for a long time. So trying to shift perceptions and get people to test out our solution to see its positive impact has been a struggle at times. We’ve approached this by being as clear and transparent as possible, explaining how our solution works and why it is more effective than a simple keyword block list solution that you set up and then forget about.

As brand safety risks have evolved, how have you ensured that your solution is able to keep up?

Our solution is highly adaptable by design. It needs to be able to shift to the exact brand safety and suitability needs of each of our clients, so we are constantly updating it to keep up with the latest digital trends and potential safety issues. AI and machine learning have been key to this, constantly learning and constantly tweaking to ensure that any new nuances are taken into account.

Take for example Taylor Swift. She’s the biggest recording artist in the world, undertaking a record-breaking world tour, so advertising near related content has long been a no brainer. But when she had to cancel three shows earlier this year due to terror threats, her name may have found its way onto keyword blocklists as brands tried to avoid being next to articles around these events.

Brands cannot risk cutting themselves off from that level of reach – when many Taylor Swift articles remain brand suitable – and so our solutions are updated to ensure that advertisers could successfully navigate these events and continue to reach Taylor Swift fans in a safe and suitable way.

What role does AI play in both brand safety risks and solutions?

There is no doubt that the ever-spreading use of AI, especially generative AI, could cause brand safety issues. Already we’re seeing incredibly life-like deep fakes videos, images, and audio perforating online.

But AI is also the answer to this problem. By using tools that can better understand the nuance of content and rapidly evolve to emergent threats, marketers can tackle these new problems as they arise.

Where would you like the conversation around brand safety to be five years from now?

I’d like brands to start approaching their ad buying with the mindset of suitability, not what is or isn’t safe. As I said before, safety is not up for question at this point, but the suitability piece is what we’re really after and what I believe all sides of the ecosystem want. Unfortunately, until now the tools haven’t been available for this shift. But now we have them I want to start having suitability conversations rather than safety ones.

Making this shift happen is a different thing altogether but I believe it will happen if advertisers just start testing suitability tools and having more conversations within the industry around safety. We need to make this change swiftly and ideally within the next two years.