23 January 2020
First published on decisionmarketing.co.uk
With the new decade well and truly underway, this landmark is the perfect time to think what the next few years will bring for data science and marketing. Ten years ago, data science was, for the most part, an academic term. Indeed, complex use of data in marketing was the preserve of only the most progressive organisations.
How far we’ve come. In a relatively short space of time, many businesses have moved from basic data analysis to building predictive models to anticipate customer churn, while attribution analysis is being used to discover the true impact of marketing spend, and data-driven personalisation. In short, marketing has never before had the capacity to be so effective and cost-efficient.
So, where do we go from here? It won’t have escaped your attention that artificial intelligence is having a “moment”; chatbots, assistants, self-driving cars are all now mainstream concepts.
Within marketing, AI has the capacity to make a big difference with simple changes over the next year. Data scientists have, for the past few years, deployed machine learning to enhance the customer journey. This is most apparent with recommendation engines; however, it has also profoundly changed the “back end” of marketing – customer segmentation, sentiment analysis and product categorisation. In 2020, I believe AI will step up another gear.
The first visible leap forward will be in relation to image recognition-driven personalisation. This has several applications. For example, instead of making recommendations or providing marketing content on the category of clothing item someone has looked at, we can look at deeper attributes to determine the style – the shape, pattern etc. Uncovering what people are actually looking at and liking online will radically improve product development, merchandising and online advertising.
The second will be further marketing automation. So much of the process is still labour intensive. Not only is this costly and time consuming, it makes it hard to recruit and keep good marketers who get frustrated spending their time executing instead of being creative. Perhaps most importantly, it is prone to human error and can mean some decisions are influenced by our own individual bias.
Now, before you spit out your “Veganuary” shake – I’m not talking about the creative aspects of marketing. I’m talking about using AI to collect, monitor, clean and categorise marketing data and then load it into the right pre-built campaigns. It will also give you the ability to marry the insight from your one-to-one communications campaigns to drive your performance marketing.
The cost-saving alone for regular email marketing campaigns by using data is significant – once you intelligently use your insight to inform the customers you are looking to acquire; the cost saving is exponential.
Of course, we’re still a few years away from compelling AI-driven copywriting and graphic design; you can spit out your “Veganuary” shake when that happens (if Veganuary still exists).
However, if I’m writing this article again in 2029 (what a thought), I believe AI-driven marketing campaigns will be the norm. Although, it will be an incremental process, I believe 2020 will be noted for a spike in the number of companies willing to hand over more of their marketing function to algorithms, and let their marketers spend their time being more creative.