David and Goliath: How agencies can win the AI battle
Opinion
The big platforms continue to grow and brands keep pumping the fruit machine for that ill-defined ‘win’. But agencies can win this match against the giants.
YouTube has come to dominate the streaming TV market and is now significant in podcasts, but there is little conversation about the advertising that appears on it.
I had a recent experience where a great show re-broadcast from TV was punctuated by ads that turned up unannounced and unwelcome, like drunken gatecrashers at your wedding reception.
Needless to say, I made full use of the skip button. But if my experience is typical, then viewers get irritated and advertisers waste their production and media time and money.
Google usually gets paid anyway.
They don’t have to worry about the viewer experience, the effectiveness of the ads, whether the impressions are real or if there are user or brand harms.
They are just the dumb-pipe providers and, if people want to put terrible ads that some people skip, it doesn’t matter much to them.
Maybe it’s just me, but lots of YouTube seems to comprise terrible content with even worse advertising, poorly presented.
YouTube audiences grow anyway and advertisers keep pumping the fruit machine for that ill-defined and unmeasurable “win”, while Google has no material costs other than the peanuts it pays content creators.
The YouTube model is almost admirable if making money is the sole criterion and, with a 14.6% increase in ad revenue in 2024, there is a lot to envy too.
Audiovisual equivalent of junk mail
YouTube seems to me to be the world’s leading purveyor of the audiovisual equivalent of junk mail.
The beauty for advertisers is that the ads are cheap, easy to buy and the CPMs are low, so the cost of all this pointless advertising is so negligible that you can do it over and over again.
It’s not just the “long tail” doing this. The big brands often act like small ones by adopting the same tactics and micro-metrics, with meaningless or no currencies and ad exposure thresholds that they wouldn’t accept for other media.
They don’t seem to care, as long as they get the appearance of “outcomes” (however loosely defined) and the procurement people get “savings” because everything is so cheap and some agency fees can be eliminated.
So why is this happening at such a scale and so quickly?
A Google search reveals the following from Google’s own statements: “Google periodically offers incentives to accelerate the adoption of and investment in Google’s advertising products…
“Incentives to advertiser and agency participants may include: cash payments; discounts on standard pricing; free media and impressions; access to special inventory, such as YouTube Core Select Lineup.
“In addition, advertisers and advertising agencies may receive campaign assistance, such as industry research and on-site consultation.”
This won’t be the only reason why it’s happening, but it helps explain why big advertisers find themselves in among the patio water jet guys.
Advertising frenemies
So far, so… questionable. And AI hasn’t hit video yet.
Recent famous pronouncements by Mark Zuckerberg of Meta and Brian Lesser of WPP provide the clues to where this might all go.
The big platforms and media agency groups are in an arms race to automate the end-to-end process, with super margins on offer, but they’re locked in a “frenemy” relationship of “co-opetition”.
Sometimes, a lot of hilarity ensues.
Amazon, Google and Meta are worth 68 times the big holding companies so are likely to win but, other than YouTube, what clues exist to the form that AI-led, automated advertising may look like if they do?
Never mind the dashboards, let’s look at real-life Google Performance Max ads that appear in my Times subscription every day:
These are mom-and-pop businesses. No self-respecting global pharmaceutical company would do something like this, surely?
Next time I need to culture my cells, I know where to go. Not to these guys.
And I have no idea what this is, like nearly all Times readers.
Personalisation paradox
One thing all these ads have in common is that they are uninformed by any data about me, despite the much-touted audience addressability rhetoric of the last 15 years.
Another thing in common is that the “copy” was almost certainly not approved by the advertiser nor vetted by anyone, and no-one knew in advance that the ad was going to appear where it did and when it did. So the only human involved was me.
The only other one might be someone, somewhere, peering at a dashboard and wondering why nothing has happened.
Presumably, The Times is happy to pocket a minuscule amount of money per ad and doesn’t care about the user experience, just like Google, because it doesn’t see the ads appear either.
These ads are the personalisation paradox writ large, based on the theory that ad exposure is highly relevant and so on, but the reality is that advertisers just “spray and pray”.
No doubt AI will improve rapidly and the “creative” will get better (maybe) and the placement processes will fulfil the promise of personalisation at scale (hmmm…), but all of this suggests that automation is risky and leaving it to the platforms may be unwise, to put it mildly.
How agencies win
Agencies have a chance to win this match versus the giants they’re up against by proving they can deliver incremental and profitable business uplift through:
• Demonstrating to the C-suite in companies of all kinds that marketing investment does lead to business and brand growth, short and long term, and that brand value and customer lifetime value are significant money-drivers.
• Showing how the highly creative use of human skills supported by AI produces messaging that productively engages the mind, not just the eyes and ears, whatever the channel mix.
• Proving that a great messaging approach can and does work in every channel if applied consistently.
• Creating people-led communications strategy across all relevant channels, not just paid, backed by the right research and data, and incorporating important messaging factors such as distinctive brand assets.
• Reaffirming that the business is built by a combination of collective and one-to-one media, and rarely by using only the latter.
• Proving that the right use of technology and data really can help deliver better value and impact, and not just eat into budgets unproductively.
• Demonstrating that cutting back on marketing and advertising to save money is a false economy and that cheap ads, cheap media and low fees are counterproductive to business growth.
• Proving that the right kind of measurement and analytics informed by the right data can be an essential guide to how to construct marketing across all paid, owned, earned and shared channels and proving their contribution to uplift.
• Confirming that expertise, experience and relationships really matter, and that trust and transparency are not optional.
• Reminding advertisers that they benefit from a healthy media ecosystem, so it’s sensible to support it.
The platforms can’t and won’t do any of this.
It seems that the big agency groups are set on going toe to toe with the platforms and trying to take them on in their core area of capability and with a nano-fraction of the resources.
That might be the way to massive headcount reduction, huge arbitrage margins and a significant upgrade on their multiples, but they’ve got to get there before Zuckerberg does and they have to transform legacy structures and cultures that don’t exist in the platform businesses.
Advertising is too important to leave to people who don’t care about it other than its link to their bank account. So let’s prove why agencies matter and why caring is vital if we want advertising and an ad industry that are better than what we have now.
Nick Manning is the co-founder of Manning Gottlieb Media (now MG OMD) and was chief strategy officer at Ebiquity for over a decade. He now owns a mentoring business, Encyclomedia, offering strategic advice to companies in the media and advertising industry, and is non-executive chair of Media Marketing Compliance. He writes for The Media Leader each month.