It’s not hard to imagine a near future in which many of life’s more time-consuming tasks are conducted by AI on our behalf. With increasing numbers of us enjoying the convenience of online grocery shopping, this particular weekly chore is crying out for automation.
Research giant Gartner predicts that by 2018, 30% of our interactions with tech will involve conversations with smart machines. AI works best when given control over repetitive, often mundane processes, so it’s safe to assume we won’t be waiting long for assistants like Siri or Alexa to be making decisions on regular household purchases on our behalf.
But contrary to movies, AIs aren’t created as super-smart sentient beings. Like any intelligence, it begins fairly limited and gets stronger by learning, analysing data and spotting patterns.
Patterns in Purchasing
Many believe the vastness of online data will turn fledgeling AIs into unstoppable digital behemoths in no time. But while 2.5 quintillion bytes of information is created every day, the vast majority is neither labelled nor categorised for meaningful reference by retail or supermarket software developers. These developers will instead find themselves relying on more structured data fields to train their early AIs.
Pre-existing online shopping histories present the ideally structured dataset for gaining insight on what your customers need, how often and when. Early trials in voice-assisted AIs prove this – normal people don’t shout Andrex at Alexa, they shout for loo roll. Alexa won’t ask them to specify but instead will base next steps on what they already know about the consumer.
Favourite product history will be mined and extrapolated to train AIs; not just in your household’s favourite brand of loo roll but how often you buy it, in what quantities and at which price. Even those trying online shopping for the first time will find that Sainsbury’s answer to Skynet will calculate a spookily accurate approximation of their habits, based on how they stack up to customers with similar characteristics.
The consequences of this for brands are enormous. When consumer choices are dictated by a historical data set, it becomes incredibly difficult for brands to make their way into a new shopper’s baskets. The odds of being selected by the AI are minimal and with such a variety of competitors, the grocery war for attention will become even harder fought.
This could be frustrating for shoppers looking for variety, but for many FMCG brands it could be a deathblow.
The good news is that AI isn’t quite ubiquitous yet, so there is time to prepare. But survival depends on carving out virtual real estate in favourite item lists and shopping histories now.
This is only possible if FMCGs begin to play a far more active role in the purchasing journey. In particular, brands must get smarter about measurability. Too often, digital’s results don’t live up to its sway over marketing budgets. Procter & Gamble's recent assurances that multimillion-dollar digital spend cutbacks had little impact on growth is very telling.
Rather than just aiming to boost brand awareness and hope that a customer remembers that fantastic piece of marketing content when they’re next in-store or online, brands need to be providing customers with more timely, direct and easy ways to purchase their weekly essentials online.
Shoppable marketing technology provides a way to turn wasteful online marketing into a direct checkout, as well as satisfy purchase-history-happy AI bots. Enabling customers to purchase directly from online content or add items to retailer baskets without leaving the page, FMCG brands can create seamless shopping experiences for consumers while staking their claim in purchase order history. This tech can also be deployed across an online advertising campaign rapidly and at scale – all that’s needed (before our unfeeling binary overlords take over) is great creative to support it.
In an age where AI is poised to wreak havoc on the marketing sector, it is perhaps FMCG companies that will feel most of the pressure. In this race against time, it is imperative that brands take steps to secure their future and act now to get as many products into as many order histories as possible -.or leave their future fortunes to be dictated by algorithms, rather than consumer choice.
Richie Kelly is chief executive at Adimo