Predictions of AI AdTech in 2020
2020: a new decade is dawning. The last ten years in Advertising Technology (AdTech) have been nothing short of revolutionary. Digital advertising outstripped TV. Twitter ads, Instagram ads, lookalike ads – they all arrived on the scene. The slow but growing march of programmatic. The rise of Artificial Intelligence (AI) bidding algorithms. Cambridge Analytica topped the list of data-related scandals, while GDPR and other privacy protection laws came into force, and adblockers appeared left, right and center. With that as our backdrop, what does 2020 and beyond hold for the industry? This article mentions predictions of AI AdTech in 2020.
New Channels and Formats
The most noticeable changes will be new advert channels.
The Internet of Things (IoT) is gaining pace. Like the regular internet, there will be a drive for advertising-based business models to fund free or low-price smart objects. For example, a future fridge might connect to special offers at the local shops, offering up one-click shoppable recipes. Out-of-Home (OOH) is meanwhile going digital, and from digital hence to programmatic.
Augmented Reality (AR) is back with a bang — after the misadventure of Google Glass, it is returning more thoughtfully, as task-specific tools and in the form of teaching aids. AR and VR are merging into eXtended Reality (XR), a simple version of which can now be run within a browser: one quick link click versus the need to download an app. This is a massive breakthrough for mainstream use, and with it comes huge potential for new adverts. Imagine: virtual avatar ads and Minority Report-esque personalized OOH adverts.
Expect also to see growth in the nascent field of Voice adverts, primarily for Alexa and Google Home. From there we may well move to talking chatbot ads, blurring the line between Advert and Sales assistant.
Attention to Data
Data will be a hot topic. The supply of data, which is the fuel essential for AI targeting and media buying, will both grow and shrink.
New data is coming online. Facial recognition is extending to include emotional analysis. Several vendors are already selling store video cameras with person tracking capabilities. AI is being deployed across data feeds, old and new, to infer extra insights. Eg. inferred demographics derived from video data.
And those inferences can be far more revealing than users realize. Has someone voted, for example? Location data can answer that. What is that person’s religion/sexuality/politics? Browsing or social media can give correlated guesses.
Of course, against the backdrop of this increase in how much we can know about people will come the inevitable regulatory and market backlash. New laws will be made and anti-tracking measures devised and launched, to much fanfare.
All this leaves AI in an uncertain space, where technologies – unless they embrace user rights – may be built on shifting sands.
“Half the money I spend on advertising is wasted; the trouble is I don’t know which half.” John Wanamake may have famously said that 100 years ago, but it has resonated with advertisers ever since.
Digital advertising has long held the promise that precise analytics could answer that question, and allow marketers to spend with pinpoint accuracy. Actual results have varied though — often impressive, but still far short of the dream. Some of the gap is papered over with bluff and nonsense from vendors. That’s led to a market where solutions proliferate, but trust does not.
Could 2020 be the year when measuring success comes of age?
Many brands still look at things such as impressions, clicks and completed views as the absolute measures of effectiveness. And as “indicators”, they’re certainly serviceable and solid. But nobody buys advertising because they want an 80% video-completion-rate (VCR) – they buy advertising because they want to build their brand, engage customers, and sell products.
If you’ve been subjected to a noisy, unskippable ad experience – then yes, it’ll show a high complete rate and viewability – but you don’t think well of the advertiser. But because we rely on stats like video-completion to measure success, it’s entirely possible for a brand to waste money on a channel that’s ineffective, or worse, damaging. Personally, my own video-completion-rate is quite high — because I treat any unskippable (or delayed skip) video ad as a suggestion to put the kettle on and make a nice cup of tea. (As an aside: my children can use YouTube – I am not actually as into cartoons as ‘my’ video viewing suggests. Advertisers: please stop selling me toys at work.
There are increasing numbers of feedback channels that allow advertisers to close the loop and see the causal links between advertising and brand value. These include increased tracking, but also fast feedback tools, like real-time surveys. At the sci-fi end, several companies are experimenting with mood-measurement from lightweight brainwave scanning devices. Each of these is designed to solve the accuracy gap that David Ogilvy nailed: “People don’t think what they feel, don’t say what they think, and don’t do what they say.”
From this, we’ll start to see more brands, vendors and agencies adopt “end goal” measures (e.g. purchase intent) alongside or instead of instant behavioral ones (e.g. clicks). AI will increasingly be used to merge low-level metrics into human-level stories of campaign performance.
Unfortunately, every step forward in this wonderful world of AdTech is accompanied by a certain amount of hype and pretense. To really progress, we need to become more comfortable with uncertainty and the intangible – restoring the art in advertising, while taking the best of the science.
As for the Good-Loop R&D team, 2020 will see us focus on Machine Learning media buying (because people respond differently to our ads for good experience, and generic AI doesn’t work well) plus tools for users to fairly manage their data, and better, more meaningful, reporting to our clients.