It is nearly one year since life as we know it was suspended. Almost overnight humans seemed to disappear from the physical world.
Schools, offices, high streets, shops and aeroplanes emptied. It was as if the pause button was pressed on life outside our homes.
Of course, we didn’t really vanish. Instead, we emerged as almost fully digitalised beings.
As our world contracted, it seemed as if our whole life moved from the physical to the digital world. The shift took place at a scale and speed no one could have imagined. Transformations in consumer behaviour that were previously set to take months and years were compressed into days and weeks.
For consumer-facing businesses, this accelerated digitisation of human activity has shone a stark light on their data capabilities.
“The challenge facing many businesses is not a lack of data but that the consumer data they have is siloed, incomplete, inaccessible and typically aggregated around traditional segmentation”
Few were prepared to harness the opportunity afforded by the extraordinary shift in customer behaviour to being fully digitalised – i.e., observable.
Specifically, the challenge facing many businesses is not a lack of data but that the consumer data they have is siloed, incomplete, inaccessible and typically aggregated around traditional segmentation.
This is where the real power of the cloud comes in. The ability to gather, join and process vast data sets from multiple sources in real time is only possible – and economic – in the cloud.
As Thomas Kurian, chief executive of Google Cloud, recently said: “Customers historically saw cloud technology as IT infrastructure. Increasingly, they are saying, ‘Can you help me with some problems that I couldn’t solve before? Can you give me a real-time customer view?’”
IT infrastructure may have dominated the conversation in the past. But businesses are now increasingly seeing the possibility of cloud in tackling problems that just couldn’t be solved before.
To fully understand customer behaviour in real time, at an individual level, data is needed from myriad sources.
A ‘customer 360’ is a simple concept, but it is extraordinarily complicated behind the scenes to execute at an individual customer or household level in real time.
The primary customer data set – transaction and margin data – is the backbone of your customer 360.
Then you layer in intent and engagement data, such as browsing data, email engagement, customer service interaction, customer satisfaction data such as NPS, contextual data (geo), marketing exposure, socio-demographic data – the list goes on.
Once all of that data is connected at an individual customer level, what the business then needs is a customer intelligence platform.
This is so much more than a customer data platform – it is a powerful intelligence tool that sits on top of the customer data and uses analytics and machine learning to turn data into intelligence and actionable insight.
But once you have your cloud-based, real-time customer intelligence platform, what do you actually do with it?
The key is to start with a set of high-value business problems and work backwards from there. How can you turn customer data into customer insight that can help you take action to solve a business problem?
Take, for example, the common business problem where customer data is used: customer churn.
The retention of customers, particularly high-value customers, is of critical importance. Loyal customers are in effect an annuity that can be factored into growth plans.
Obviously, having an early-warning mechanism to prevent customer leakage is a high-value business problem.
Churn-prediction models are typically built using transactional data and modelling customer recency and frequency of purchase.
Retailers then end up concluding the blindingly obvious: if they haven’t seen a customer for X weeks they might be at risk of losing them and they should probably do something – usually send a panicked promo.
Now envisage tackling the churn problem using an exquisite, holistic, real-time customer 360 and a suite of sophisticated data models in the customer intelligence platform.
“Customer insight and a fixation on retaining customers is now reaching deep into the heart of the organisation and is no longer siloed in marketing”
In the first instance, retailers would want to diagnose what is going on by either letting the machine scan the data to look for patterns or testing hypotheses of why churn may be occurring.
Are there patterns or events that predict a customer is getting dissatisfied? What are the key factors that raise the risk level?
Churn is usually a symptom of something that is broken/sub-optimal deep within the organisation: availability issues, product faults, poor customer service, missed deliveries.
These are the underlying issues that need to be diagnosed and fixed. Customer insight and a fixation on retaining customers is now reaching deep into the heart of the organisation and is no longer siloed in marketing.
By building a churn model using the customer 360 and bringing in vast, detailed and contextual signal data from across the business (long angry calls with customer service, no delivery slots for weeks ahead, out-of-stocks on the website, not opening any emails, not opening the app or website), retailers have an early-warning mechanism that gives time to recover a situation and turn a lapsing customer into a loyal brand advocate for years to come.
The customer intelligence platform acts as a vast intelligence system that both signals deep into the organisation where problems lie and surfaces incredible consumer insight that can transform what and how you sell to customers.
Having that customer capability will truly be a superpower.