The first question in most ad plans is who we want to reach. After working in ads strategy at Listen Network and more than 14 years in B2B demand generation and performance marketing, I think it is often the second most useful one. The first is what that person is doing when the ad shows up.
Are they reading about the problem you solve? Comparing options? Scrolling through something that has nothing to do with you? That moment tells you more about the chance of a useful response than a profile built from past behaviour.
I spoke with Isaac Rudansky, CEO of AdVenture Media Group, about why context still matters in advertising. He sharpened a question I keep coming back to: are we matching the message to the moment, or asking targeting technology to guess what someone wants?
Precision can feel like certainty
Behavioural targeting sounds like the obvious route to relevance. Find people who have shown interest in a category, then serve them an ad wherever they go.
But a signal is not intent. Someone may have visited a page for a reason that has nothing to do with buying. Their circumstances may have changed. The data may be incomplete, or the platform may read an old action as a current need. A campaign can look precise in the dashboard and still be a poor match for the person seeing it.
That gap makes teams overconfident. When an audience is defined by a long list of behaviours, it is tempting to assume the targeting is doing the hard thinking. The ad still has to make sense where it appears. If it interrupts content with no connection to the offer, the audience definition will not save it.
Isaac also raised the cost. When advertisers compete for people based on signals many of them can buy, the price goes up without a matching rise in relevance. More targeting detail does not mean more useful attention.
For B2B marketers, that is a reason to question the story behind a segment. What does the signal tell you, and what does it leave out? A person who engaged with a topic may be researching for work, helping a colleague or just curious. Not every signal is a buying decision.
Context gives the ad a reason to be there
Contextual advertising starts somewhere else. Instead of following a person based on what they did last week, it looks at the content around the ad. The message sits beside a subject related to the product, service or problem.
A reader looking at content about a relevant challenge is already spending attention on it. The ad does not need to guess as much about the person. It meets the topic in front of them with a message that fits.
Not every page in a category is a good placement. The content still has to suit the audience and the message, and a broad topic can pull in people with different needs. But it gives you a better starting point: what is this person engaging with right now, and does our message belong next to it?
That changes how I think about relevance. We talk about personalisation as if the job is to know more about the individual. Sometimes the better move is to understand the situation. A message feels relevant when it responds to what someone is thinking about, even if you know little about their history.
Here is a test I like. Open the page or the video where the ad will run, then look at the creative. If you struggle to explain the connection, the placement is leaning on audience data to do the work. If the message follows naturally from the content, context is earning its keep.
Match the message to the moment
Placement is only half of it. An ad beside relevant content can still miss if it asks for too much or speaks to the wrong stage.
Isaac's point about matching the message to the moment is a reminder that relevance is not a setting you switch on in the platform. It comes from the fit between the content, the reader's likely need and the creative. A sensible placement still fails when the ad is generic.
For a B2B team, that means thinking past the audience label. What is the reader trying to understand? What would help them right now? It might be a clear explanation of the problem, a way to compare approaches or a next step that asks for less commitment. The creative should reflect the moment, not repeat the same pitch everywhere.
Test the message in context too, not only in a preview window. A headline that looks strong on its own can feel awkward beside the content it is meant to sit with. Reviewing placement and creative together shows you mismatches a platform report never will.
Changes to cookies give you another reason to build relevance this way. If you cannot rely on the signals you used before, you need other ways to connect an offer with what someone cares about now. Context does not answer every targeting question, but it gives you a route that does not depend entirely on tracking people across sites.
Measure past the click
Even a well-matched campaign is hard to measure. Omnichannel activity spreads across touchpoints, and attribution tools hand out credit to single interactions. You get a tidy report that can paint the wrong picture.
Isaac talked about brand search lift as a proxy for campaign impact. If more people search for your brand after a campaign runs, that is a sign the activity is creating interest, even when the path from exposure to action is hard to trace.
Treat a proxy as a signal, not proof. Brand search can move for reasons that have nothing to do with one campaign, so check what else was happening and whether the timing holds up. It works best next to other evidence.
Isaac also pushed for transparency and the marketing efficiency ratio. I agree with him here. Teams need a shared way to talk about what marketing produces against what it costs. When attribution is murky, clear assumptions and one efficiency measure keep the conversation grounded in math. Nobody can credit every channel with certainty. You make the call with the evidence you have and say plainly where it stops.
So these are the questions I would bring to the next campaign review. Did the message fit the content? Did the campaign create signs of new interest? What did we spend to get the result, and what are we assuming when we hand out credit? Those beat treating a platform's reported conversions as the whole story.
Good targeting is less about collecting the most signals and more about making the message belong where it appears. Start with the moment, then measure the wider effect with clear eyes.
Want the full conversation? Listen to my episode with Isaac Rudansky, CEO of AdVenture Media Group, where we discuss contextual and behavioural targeting, omnichannel measurement and practical ways to assess marketing efficiency.