Handle your predictions with care
Having the humility to review your "next big things" can help you avoid reputational pitfalls
Back in the day, there was always one set of predictions I'd look forward to. When
and Marshall Manson were at Ogilvy together, they'd produce a set of slides with their thoughts on the year ahead. The bit I'd look forward to most was the ranking of their predictions from the previous year. They were always honest with themselves about what had panned out and what they'd got a little overexcited about.This honesty and humility set up the new set of trends perfectly - I always felt their analysis carried more weight because of this willingness to look back before looking forward. The Guardian's Sports desk carries out a similar exercise at the end of each Premier League season, putting together a side-by-side graphic of how their writers predicted the table would look vs how it ended up.
Trend forecasting is an inexact science, and making concrete predictions for the future is even more fraught with uncertainty. That's because, as Paul Ford in Wired highlighted when he revisited a set of predictions made 50 years ago, we forecast and predict based on our current view and available information. We tend to think that the trends we see will continue to grow rather than anticipating drastic changes and updates.
Few football writers could have expected that Liverpool would struggle quite so badly this season, the loss of Sadio Mane upsetting the balance of this carefully orchestrated side. Similarly, 14 managerial sackings, with three clubs sacking two managers - the tumultuous nature of the EPL means the information we have as the season starts is irrelevant by Christmas.
Luckily in our world, we have fewer transfer dealings, leadership changes and ownership wrangles to deal with (although we do have our fair share). The information we use to think ahead when looking at trends or planning campaigns is imperfect but at least relatively fixed. But still, we get very excited about many things that tend not to catch on.
If you came to one of our SXSW trend sessions around 2016/7 time, you would have heard us getting pretty frothy about autonomous vehicles. I got very into the potential second and third-order consequences of humans no longer having to be behind the wheel 100% of the time. ICE cars invented out-of-town shopping centres and drive-throughs, I opined knowledgeably. Imagine our future cities with wider pavements and more space when we don't have to allocate so much space to parking.
Looking back on this prediction, in a James and Marshall-style, parts of it are accurate. In some US cities, autonomous vehicles are a common sight on streets - in fact, this week, Waymo partnered with Uber to launch an AV taxi service in Phoenix (via Benedict Evans).
But these shuttle-type journeys are different from what we had in mind when we imagined cars that could park themselves in cities of the future.
To be kind to myself for a moment, the information I had at the time indicated that autonomous vehicles would enter the real world from the pages of science fiction. The machine learning around driving was getting exponentially better every year; there was no reason to expect it would hit the brick wall it has. What has transpired is that moving to so-called level 5 autonomy requires a massive jump in our available tech. And that jump doesn't look like happening any time soon. Hence we're stranded in an AV winter, froth very much settled back into liquid.
Part of the appeal of autonomous vehicles is their innate sci-fi nature. And no matter how UNLIKE modern life is from science fiction, we can't seem to disentangle futurist narratives from sci-fi ones. This week, The Times warned us of AI's threat to the human race. These narratives take their inspiration from Terminator/2001: A Space Odyssey (The Times directly mentioned Terminator) and conveniently ignore the massive jump the tech behind tools like ChatGPT needs to make to get to Artificial General Intelligence.
For whatever reason, these leaps from the current nascent trend (smarter cars, generative AI) into full-blown sci-fi utopia/dystopia (delete as appropriate) remain highly beguiling. We saw it last year with the hype around the metaverse. The information we had to hand indicated that younger gamers were increasingly spending their leisure time in immersive worlds such as Fortnite, Minecraft and Roblox. That same audience spent much of this time gaming, but they spent just as much time hanging out or going to special gigs/screenings those platforms facilitated. A Silicon Valley VC named Matthew Ball wrote a series of articles speculating that this could be the start of the so-called "metaverse".
Then suddenly, everyone seemed to lose their minds and, in Mark Zuckerberg's case, their money. Mark renamed his whole company. Ad agencies the world over scrambled to sell metaverses. Companies hired Chief Metaverse Officers. And now, with the rise of generative AI, all this hype is forgotten, leaving behind a company with a strange name and a whole lot of regret.
I came firmly down on the side of bearish on the metaverse hype. Following
’s lead, I felt the information we had available didn't indicate any appetite for "immersive" VR worlds where not much happened.The likes of Fortnite, Roblox and Minecraft remain hugely popular - but they're predominantly popular with those under 20. And all three are games, even if they're not games in the FIFA/Call of Duty sense. Gaming is a big deal (circa 300m devices across PC and console), but it's not properly mainstream in the way Instagram, Snapchat and TikTok are (i.e. 5bn smartphones). Thinking that user behaviour among teens and pre-teens will crossover to a broader audience is a leap. And that's before you even factor in how uncomfortable most VR headsets are.
Apple's new Vision Pro headset is the final nail in this initial age of the metaverse's coffin, firmly pitched as mixed reality (XR), not VR.
You never know; it may make a comeback. All bets are off if we suffer another global pandemic similar to or worse than COVID-19. But the likelihood of the metaverses people want, the ones from Snowfall and Ready Player One coming to life, is remote.
The good news for companies who created their own metaverses during 2022's hype bubble is that there's unlikely to be any long-term reputational damage. At least at a brand level; I can't speak for the person that signed off the budget. Going all in on autonomous vehicles too soon is altogether more risky.
And that's what those of us in comms who are looking for the "next big thing" or want to take advantage of the latest trends need to consider. What are the risks of this activity?
If I'm going to save money on photography by using Midjourney to create the imagery for my assets, what's the potential blowback? We've already seen organisations such as Amnesty get headlines for lying about using AI imagery.
Understanding those potential risks requires making the best possible use of our available information. Study user behaviour. Follow the lead of
and get out and speak to your audience. Engage in some scenario planning to anticipate what might unfold.Adding this kind of rigour to your experimentation will ensure that when you look back on how it went in a year, in a Manson/Whatley style, you're more likely to have a campaign to remember rather than a metaverse to forget.
Great piece, Mark. Thanks so much for the kind words. A few years later, I'm still surprised by how much we got right, and the stuff we got wrong. But the work remains some of the best I've done, and the collaboration with James the best professional relationship that I've had.