Posted on March 16, 2020 by staff

Coronavirus shows up online advertising limitations


Advertising has long been the canary in the mineshaft for the economic outlook, and in the last week alone, thanks to COVID-19, it’s gone from chirping a happy tune to picking up a nasty wheeze.

However, if we think advertising has it bad, that’s nothing compared to life in the news publishing groups.

With sales targets looming, and ad-funded news remaining the most common choice – especially vital on smaller titles – the conflict between editorial and commercial starts to intensify all the more. Already combating the air of “fake news” and spin, they’ve now got to fight internally to strike a balance. That’s a lot of fronts to fight on.

While there are many factors that fuel this squeeze, the reality is that current and topical news stories are always the most read, engaged with and clicked on content – a quick look at the BBC Most Read shows 70% of the articles being about Coronavirus.


This creates a dangerous Catch-22 for the publishing groups. On one hand, they have a duty to report the news – it’s what their readers expect and demand, and ultimately, journalists didn’t get into the game to worry about ad space. But, the more publishers write content about negative topics like this, the harder that inventory becomes to sell – in the last week alone, Coronavirus jumped up keyword block lists by 80%.

To give a simple explainer, keyword blocking (contextual targeting/blocking) is a tool that essentially enables advertisers to target or exclude their ad from appearing next to certain words or phrases – all in the nanoseconds before you see the ad.

For example, if you work for Nike, you might want to target articles that say ‘sportswear’ or ‘running’. But you might also want to make sure you don’t show your ad to consumers when they are feeling concerned or negative – so you choose to exclude terms like ‘war’, ‘guns’, ‘attack’, and so on.

Before the ad shows, this tool runs through the article to check if it’s safe/desirable to show. If the article matched a negative word like ‘war’, it won’t show the ad. If it doesn’t, or finds one of your target words, then depending on your settings it will show it.

Often, these lists of targeting or exclusion words can run into the thousands – the exclusion lists often containing the most creative insults and foul language you’ll ever see. That’s important for maintaining digital hygiene – there’s certain things you just never want to be near.

But when it comes to news, things get complicated.

In recent weeks, we’ve seen instances such as the masthead slot of The New York Times showing clouds and blue sky (a placeholder used by one such brand safety company when an ad has been excluded), instead of an ad. This is a premium ad slot, on one of the most trusted and praised news titles in the world. Blocked. Because somewhere on that page, lived the word ‘Coronavirus’.

It’s worth mentioning this isn’t new – it’s been a daily occurrence for some time. When ISIS were at peak notoriety, ‘ISIL’, ‘Islamic State’ and more started to appear – as did ‘IS’. This, of course, meant that some campaigns couldn’t run on any article containing the word ‘is’. There really ‘is’ not much you can match against there.

There’s all manner of intricacies as to whether the publisher still gets paid in this scenario or not – but the fact remains that many brands have chosen to forego coveted ad slots, for fear of association with something negative.

The brand safety companies, generally, do offer safeguards against this. For example, with some vendors I can say “I want to exclude ‘war’ generally, but I’ll go anywhere on The Guardian”. This offers a flexible way to pick and choose where you’re willing to make those tradeoffs/take risk. And some are able to offer some level of risk or semantic analysis, evaluating the page and story as a whole through their tech – allowing a brand to set their own risk tolerance.

However, many campaigns will elect for a brand safety approach that resembles a sledgehammer to crack a walnut – excluding these words or categories, and thousands of others, often unknowingly.

Many lists currently circulating still contain phrases like ‘Charlie Hebdo’, and ‘Norway bombing’ – terrible events, but that ultimately aren’t in the news cycle anymore. But, the impact on the brand remains if these phrases linger – if, for example, an article were on an entertainment site about ‘Charlie’ Booker ‘bombing’ at a comedy set in ‘Norway’, there’s a fair chance the brand safety tech won’t show the ad. It’s totally safe – but because of the keyword block, it won’t show.

Over time, as these lists grow and grow, the pool of stories and news a brand can advertise alongside gets ever smaller. The content that a publisher can monetise also gets ever smaller. This drives price pressure in all directions – advertisers who want a cheap price, and publishers who, with now limited stock they can sell – want higher rates. Nobody is happy.

This stalemate risks certain topics becoming blacklisted for journalistic coverage, since writing about them directly impacts publishers’ revenue potential. This is bad, broadly for society and our need for diversity of opinion, but also for the brand and publisher relationship. It should never be the case that editorial is having to ban the use of certain words, just to beat a faceless machine and keep the lights on. But in many offices, this is the case.

People searching out news content are, arguably, the most truly engaged audience you can find – say what you will about Coronavirus, but it’s a hot topic and people want to know anything they can. News on a digital publisher isn’t force-fed to viewers – it’s sought out by them – you don’t land on the FT by accident, for example. People have opted in.

As a result, showing people your ad in these places isn’t the brand safety flag you might think – in fact, you’re arguably missing the opportunity to engage and be on the pulse. In fact, contrary to popular opinion, being around these negative topics doesn’t inherently tarnish your brand – Corona beer sales are up 5%.

If Corona can still sell more beer despite having the most popular ‘bad’ keyword on the bottle then maybe it’s time for advertisers to loosen up a little. Block lists do have a purpose and should be used where appropriate – but it’s also time for brands to consider the opportunity of casting the net that little bit wider, and being part of the conversation, than risk living on an island of irrelevance.

The opportunity is there for those who move first. To borrow from Levi/BBH – when the world zigs, zag.