The real reason Publicis aren't entering any awards in 2018
(hint: it’s not about awards)
Many people were surprised last week to hear that Publicis are pulling out of Cannes next year. A lot of the focus has been on the impact on the festival, and what that means for our fixation on awards, and whether that is positive for brands, business or indeed agencies. This is about none of that, it’s about something else: the why behind the decision.
A cynic might argue, as the Finance Director of the Publicis Re:Sources internal ops group’s email indicated, that maybe Publicis is simply trying to save 2.5% of operating costs (about €200m, based on €8.2bn operating expenses in 2016), and this is a great way of dressing up the saving of several million dollars in the short term.
However, Publicis is a smart company, probably the most forward thinking of the holding companies, as attested by the Power of One strategy, and the integration of Sapient through the business (though I still cannot work out all the dots and what it means, perhaps that’s next on Arthur’s list). So I can’t imagine that it is purely a cost saving measure, despite the fact that the AI personal assistant is unlikely to be costing the same as all awards entries for Publicis globally (I know Sapient is expensive, but this could be pushing it).
So why? For an ‘AI personal assistant’, it makes no sense.
Let’s get a rewind.
It is telling that this announcement was made at Cannes. For anyone who has been in the past few years the change has been striking with the advent of the tech giants to a pre-eminent position at the festival. Last year the Palais was covered by a giant Snapchat ghost, this year a Ferris wheel literally and symbolically loomed over the festival itself. Sadoun even referenced this in an Ad Age article on Friday and much of the commentary about Cannes last week came from the changing relationship with creative and media agencies, and the perceived loss of focus on what is important to them.
The big platforms, now essentially a duopoly, are increasingly dealing directly with advertisers. For the channel representing 50% of your revenue as a holding company, that needs to be concerning, with the resources these ‘frenemies’ command. The second point is the huge trend — the increasingly data led push towards marketing automation is coming — removal of the human factor, not entirely, but increasingly
When we look ahead it is worth considering what does the power of one look like in five years time, in ten? Publicis’ share price is currently trading at a 15x multiple of earnings (P/E), whereas Google is nearly 33x, Facebook is 39x (as of 24th June). Why? Because Google and Facebook ‘own the future’ — the market assesses their ability to generate profits in the long run to be much higher than most other businesses.
Adobe’s share price in the past few years has been helped by a move into marketing content and automation, but someone must surely be thinking, wouldn’t it be great if an ad agency was able to capture some of that value (and valuation) as well.
Automation over the past fifty years has tended to focus on blue collar work (in manufacturing, factories, etc). However, slowly we are coming to the realisation that as an industry this transformation is already upon us, but locked away in the less glamourous corners of the industry — paid search, triggered email programs, automated social and promotional Facebook posts, it is easy to look past it to the inspiring and shiny film Grands Prix.
But now, increasingly the robots are coming for us, the knowledge workers, too.
Sorry everyone, AI is not about that chatbot case study your innovation team recently showed you. Look at the news industry, we now have AI writing news articles that are somewhat more complex than the financial news and sports statistics feeds that have been running for the last 7 years. In fact Jeff Bezos’ Washington Post has been trialling AI powered creativity with its Heliograf software over the past year or so in what is hoped to be “a seamless interaction” between human and machine. Now the systems are still relatively formulaic, and require human input, but it is clear we’re now talking about a question of a ladder of capability, rather than whether or not it is possible.
Quoting the Wired article profiling the advent of Heliograf and similar tools there is this telling paragraph:
The Post’s main goal with the project at this point is twofold. First: Grow its audience. Instead of targeting a big audience with a small number of labor-intensive human-written stories, Heliograf can target many small audiences with a huge number of automated stories about niche or local topics.
Can you see the parallel here with the emerging needs for creativity in marketing during this digital age? Look at the pressures we have as an industry:
Squeezing margins from the existing ‘way of doing things’
The opportunity to take advantage of new data streams to reach audiences in new ways
Ever increasing numbers of deliverables
Now many of you will be thinking at this point that this type of automation is incredibly basic, and is never going to come close to the ineffable emotional work that represents the best of what we can do, the great work that has business changing impact, the pinnacle of our craft.
Never?
Consider this:
AI is already the cornerstone of programmatic media buying, representing around 70% of all media bought online, and increasingly is being used as part of the creative development
Landing pages on websites now regularly have content automatically generated based on what your preferences are likely to be
Search engine marketing copy is created on the fly by algorithms that recognise what you are likely to interact with
Films are being created in real time from a series of pre-set assets and then fed into Facebook video ads based on your stated preferences
This is inevitable
In what is fast becoming the seminal book on the subject, Nick Bostrom’s Superintelligence compiles the academic consensus on the ‘singularity’ — the moment when computer intelligence outstrips human intelligence, which will inevitably create a runaway advance …. Sounds mental? This study from 2014 (PDF), suggests that the median estimate from the consensus of AI researchers is that there is a 1 in 2 chance we will have high level machine intelligence by 2040, and a 9 in 10 chance by 2075.
Now artistic pursuits (and certain elements of the marketing industry will meet that definition) are likely to be close to a ‘complete AI’ problem, but much of it is not.
So, how will this go down? Rather than a quick transition, we will likely see a creeping automation of the advertising process. Initially we will be looking at humans with technological augmentation — starting with data led and data inspired services, and augmenting the process through connections and information tools.
But the march of progress could be fast, and even if you do not buy into the likelihood of a largely or even fully automated future, a forward thinking business would be considering how to adapt to this situation, and remain relevant, efficient and effective in the coming years.
So in this context, what is the value of Marcel the AI personal assistant, particularly if it is going to be worth the potential external and internal reputational risk of signalling a lack of creative intent over the coming year?
First, and perhaps most importantly, it is a stake in the ground, and indicator that the new world requires new approaches, new avenues of thought.
Secondly, it is a strategic step to look at how AI can augment the creative process — both in projects, connecting people, and I presume building towards human augmentation — think again about Jeff Bezos’ Heliograf software, the critical question is what’s the marketing use case for how this could come together.
What you should do about it
At a macro level automation in the long run does not affect employment rate, the white heat of technology reached England in the 1760s when we invented the spinning jenny for making thread, but we don’t still have unemployed hand spinners sitting on street corners. Spare resources tend to get reallocated within an economy over time.
Unfortunately, in the case of the first industrial revolution ‘the long run’ was over the span of around 200 years, and created significant economic problems for working people for several generations. So from a personal perspective, recognise the trend and go with it, there may be fewer jobs in the future, the trick will be to have one of the jobs that haven’t been invented yet.
As a business the imperative also has to be to go with it, large technological changes creates corpses of huge corporations that were not able to adapt (Kodak, Blockbuster, Borders) and this is the risk for the big holding companies over the coming 20 years. There is no reason really to assume the value chain will go advertiser > agency > media owner > customer forever, and in many cases now you can see that this is not the case any more.
So in whatever form it ultimately takes this is likely what Publicis is doing, and what you should be doing too — figure out what next, try to imagine how technology can augment the creative process to help the talented humans you have to be more effective, and look at what insight you can uncover from your data, people, process, product. Ultimately, as it ever was, if you’re not looking to change your business with the opportunities of new technology you’re doing it wrong.