Alex White, COO, Peer39
Everyone is talking about the end of the cookie and what kinds of targeting will survive in the aftermath. Many continually debate that context is or isn’t the answer to this loss of signal. To be honest, I don’t think that context is the answer either, but that’s because the definition of context that the market clings to is incorrect.
As many understand it, context is about the page or the content in front of a user. But that’s a very limited definition. Context is understanding the environment where an ad can be put in front of the user. Thinking about the page is limited, and thinking about the user is much more open and full of possibilities. If the ad industry confines itself to a limited perspective, then it’s no wonder no one sees context as the answer to the loss of audience data.
Users, not pages
If you broaden the definition of context to put the user first, you can start to think about contextual targeting in a different way.
Like many of you reading this, I am currently working from home. I am on a laptop with a big screen in front of me. I’m answering emails, jumping on calls. If I am consuming content, it’s industry-specific. That’s all context, from the content on the page to the physical environment around me.
But at least once per day I switch contexts and look at the stock market. My father and I like to invest in gold stocks, so when gold is up, I am happy. As I write this, gold and silver are both up 1.7%. Another few times per day I check for major news. Today it’s raining outside, and that reminds me that I need to pick up a rain cover for my patio heater. It’s Thursday so I am starting to think about plans for the weekend.
Of course, each day is different. I have meeting-heavy days where I’m on the phone a lot, gold could be down, the weather could be different, and I’m doing and thinking about different things. That’s all still context.
Advertisers can target based on these signals. Weather, stock market indexes, time, device, and location are all contextually relevant signals that extend beyond content.
Let’s expand on location for a minute. So much of the conversation around location has been tied to mobile and hyper-targeting users based on where they are or where they go. But location matters to the desktop user as well, because it’s a contextual signal. Where does the user live and work, and what current events are happening around them? What kind of neighborhood does someone live in? How good are the schools, what’s going on with the local sports teams, are gas prices high or low? Are they in an affluent area, and what types of vehicles are prevalent in that area?
These signals will have an impact on how well an ad message is received, and each signal provides advertisers with additional ways to reach audiences in those contexts.
Keywords are not context
Now, the content that consumers interact with online certainly deserves a lot of focus in conversations around context. That’s what they are engaging within the moment, and that’s the medium for ad delivery. The content will be important to them, and anything the consumer sees associated with the content will likely influence what they do next.
Therefore, it’s important for advertisers to know what the content is about. Again, the common concept of “understanding” what’s on a page is very limited. These conversations tend to focus on keywords, and keywords are not context.
Keywords are a fantastic tool that can inform context, but on their own, they are just words. To get an understanding of the content, keywords need to be in context. For example, a bunch of location-based words for travel brands don’t indicate travel content. They are simply locations. People live there, and things happen there. What the travel advertiser wants is to reach a user when they are researching that location as a destination in the context of travel. The destination by itself doesn’t indicate travel interest or intent.
Put everything in context
Keywords are a great tool because they are easy to produce and move context-based decisions in the right direction. But keywords become more relevant when used in context. Advertisers need to evolve one step beyond simply keyword targeting and overlay keywords in contextually relevant environments. That’s where “contextual signals” actually make a difference.
To better understand, look at brand safety. Brand safety is the identification of content that an advertiser wants to avoid. Like keywords though, brand safety is really only effective when context is understood. The conversation inside brands has moved from brand safety to brand suitability. This still identifies negative topics and phrases, but puts them into context and enables advertisers to better define when a word or phrase might be OK, and when it’s unsuitable.
Descriptions of crime are not desirable in a news story, but what about a movie review? Depending on a brand’s standards, there may not be a need to avoid that kind of “crime” content.
The context of what people consume on the web is critical for understanding their likes and interests, that much is clear. But advertisers who stop in understanding content alone, based on keywords, aren’t going to find any kind of cookie alternative. The old idea of context isn’t going to be enough. But if advertisers are willing to use unconventional data sets to build a deeper understanding of the person on the other end of that content, without tapping into behavioral data, they can develop a winning contextual formula that will stand up to the loss of cookies, and remain a usable asset well into the future.