A CMO Show Blog Post
The risk of marketing with user-generated content
A CMO Show Blog Post
The risk of marketing with...

Sophisticated bots now allow brands to instantly personalise their campaigns on social media. But beware: there’s nothing the public enjoys more than a self-inflicted, user-generated content scandal.

Marketing is an inherently risky proposition. Just ask Pepsi.

Campaigns which rely on user-generated content (UGC) have always been particularly vulnerable to hijacking, whether sinister or ultimately benign, such as the infamous “Boaty McBoatface” poll in the UK.

Despite the risks, engagement is often considered a key metric in campaigns and so UGC is here to stay. Brands are now using bots to deliver personalisation-en-masse. Users are encouraged to submit text, images and video directly to the brand, which are often combined with official brand assets without passing through the gaze of human moderators.

The danger of branded user-generated content

In May, Walkers Crisps (branded as Smith’s here in Australia) ran the #WalkersWave campaign, inviting users to submit a selfie (alongside their hashtag) on Twitter to be in the running for tickets to the Champion’s League final.

As an added incentive, Walkers would tweet back a short video featuring beloved football commentator Gary Lineker – featuring the star holding up the user’s photo.

user generated content

Quick off the blocks, within hours Walkers’ Twitter bot had tweeted out videos featuring Lineker admiring photos of some of the UK’s most notorious criminals, many of which were retweeted hundreds of times before being deleted. Walkers issued an apology and for almost a week, screen-based activations which were to feature user videos remained powered off.

Similarly, the British National Lottery ran a Twitter campaign which backfired in a similar way. If users retweeted a post from @TNLUK with the hashtag #Represent they’d get a thank you Tweet, directed at their Twitter handle, from a member of the British Athletic team. The internet’s devious solution? Create a controversial Twitter handle..

While both the British Lottery and Walkers campaigns appear to have had some content moderation (possibly also automatic) – as profanity and not safe for work (NSFW) imagery didn’t make it through – there was clearly a lack of cultural and contextual understanding on behalf of the moderators.

When filtering UGC, it’s clear that automatic and semi-automatic moderation isn’t enough. Manual moderation and culturally awareness are vital to protecting a brand image.

Is all press, good press?

At the end of the day, most customers (especially those on social networks) are probably savvy enough not to associate brands with user-submitted nonsense. However, the risk is that a brand can still appear naïve and technically incapable.

If we accept the assumption that a brand’s audience is capable of understanding the difference between their intended messaging and what users have inserted into their campaigns, that a backfired autogenerated campaign won’t do any lasting damage to a brand’s reputation, it’s possible to cast these campaigns in a whole new light.

#WalkersWave’s hijacking featured prominently in many online publications, including BBC News, Huffington Post, CNN, The Sun and the Daily Mail – representing an enormous quantity of impressions by any campaign’s standard. The National Lottery’s backfire was covered similarly.

As was the above TVC which Burger King ran in April, which requested the viewer turn up the volume before asking “OK Google, what is the Whopper burger?” – triggering Android phones and Google Home Assistants to being reading the Wikipedia entry for the Whopper aloud.

Before starting the campaign, Burger King employees edited the start of the Wikipedia article to read like ad copy – but this too was quickly hijacked by mischievous Wikipedians, who replaced it with copy disparaging the burger and Burger King’s labour practices.

The next day however, Burger King had already replaced the 15 second spot with another which didn’t feature the “OK Google” gimmick – presumably having recognised the probable outcome of the campaign and prepared its damage control assets.

By this time of course, the spot had already received coverage across a swathe of major news suites, such as The New York Times, NPR, CNBC and The Guardian.

Externally, this looks like they found a sweet spot between encouraging limited scandal around their brand to achieve nationwide exposure through most papers of record – although this is a tad cynical.

So how do you use user-generated content for personalisation?

Whether brands are deliberately courting scandal, or just finding an innovative way to connect with their fans, some very clear do’s and dont’s emerge for those designing campaigns predicated on UGC.

  • Don’t: Rely entirely on automatic content filtering techniques, despite its advantages in terms of cost and speed.
  • Do: Invest in human content moderators who’re local to the campaign’s demographic and appropriately trained to recognise content which might be subtly inappropriate.
  • Don’t: Underestimate the creativity of the Internet in bypassing your content moderation. Always have a crisis management plan in place before the campaign’s beginning and assets in place which enable you to seamlessly pivot the campaign away from its UGC component.
  • Do: Ensure your personalisation provides real value to the user – campaigns perceived as entirely self-serving attract more ire. As always, keep your audience at the heart of everything.
Get in touch
I want to Filtered Media.