It feels like every booth at every show has AI written above it these days, and it can be hard to wade through the noise to find the signal, but it remains a self-evident truth that some level of automation is an integral cog in the modern retail machine.
Whether that be targeting, pricing, demand forecasting or optimising product placement, we all recognise that scale comes with costs, as well as efficiencies, and that there is a limit to how many people we can keep throwing at the problem before we hit a wall of diminishing returns.
The dilemma is how we can introduce machine learning to the balance, whilst still maintaining our hard-earned brands. Amazon product pages have widgets numbering into double figures, and they all provide ROI, but for any brand not competing purely on price, that is often not a viable option.
For the best part of a decade in personalisation and search I had been banging on to anyone who would listen to me about the importance of ’man and machine’, and that is still relevant today (although perhaps ’experts and equipment’ might be more appropriate in 2021?) We need to ensure that, as we progress through the fourth industrial revolution, we use AI intelligently (pun unintended) to do the heavy lifting and donkey work, without removing the personality that you, as an expert on your brand, bring to the table. The great thing about that, of course, is that taking away the work that can be automated and freeing you up for the work that needs a human touch is not only good for business, it is good for you too. You can do the things you wanted to do when you started your career, the things you are good at, the things you enjoy. The things that make a difference.
The analogy I used to invoke was always one of constructing an art gallery – making bricks, carrying them to the site, laying them, rendering them, roofing the building, painting the walls and then hanging the paintings, when really, what you want to be doing, what you are good at, what you love, where you differentiate from other galleries, is the sourcing, selection and curation of the art you display. Nobody cares about the cement you chose or the glazing job – those just need to be right, they do not need to be branded.
The analogy holds water too – nobody really cares what your stack is, so long as your site is stable, reliable, and secure. What they do care about is being able to find your products and getting the best journey they can. They want an experience which feels tailored to them (but without being creepy – the line between personalised and stalker can be narrow) and they want to feel an association with your brand. That you really get them.
That is the human touch, and it is where you want to focus, and you know that, but what you do not necessarily know is how to manage that. This is, perhaps surprisingly, going to be less about automation of jobs and more about automation of workflow and procedure.
‘But Manley, Why?’, I pretend to hear you ask, as it fits with my intended segue, “Why would we care about workflow over actual delivery?” Well, gentle reader, let me elucidate.
AI for tasks is important, you need it and you want it, but that is what everyone at the trade show is going to be able to sell you. It can be hard to identify the right tools, but the tools exist and are being built, but for most of those you still spent almost all of your time deploying those tools where you need them. That is not a bad thing – if a machine can get your category in the best order from the data, then all you have to do is titivate then it is working for you, but you probably have hundreds of categories, and you are still doing the donkey work.
Once you have the basics of your automation covered then the remaining task is to
add your brand flavour on top of that, and unless you know which tasks need doing next you are back to donkey work – ploughing through content and selections t identify areas which need work.
Intelligence in automation is about being able to prioritise your opportunities, identify the tasks which will have the greatest ROI in terms of you KPIs and then guide you (and other relevant stakeholders) through that process. You need a workflow which recognises the areas of your site which have issues, measures them against internal benchmarks, identifies and, more importantly, prioritises opportunities and then guides you and your team through execution and delivery.
Only then can you truly trade your site intelligently, focussing on the brand experience and trusting the AI to pick up the slack and identify your next area of focus for you, so that you do not have to carry those bricks, you can get on with hanging your Rembrandts.
Manley CEO of Corigan.io