What do people actually mean when they talk about automating customer marketing?
By ignoring the vast ‘black holes’ of programmatic, social, and other forms of adtech used to recruit customers, we can focus in this paper on automating marketing to existing customers.
This usually happens through three main channels; website, email and direct mail. However, what we would like to consider is not the channel used for the delivery process, but how the automation that applies to that should be set up.
To us automation can normally be applied either when a; a customer event takes place and you want to respond to that trigger, or b; when a campaign is being sent out to large numbers of customers to whom you want to deliver relevant communications that will make a return. In plain English, trigger or batch campaigns.
When planning trigger campaign automation there are always two perspectives to consider; which customer group or segment are we expecting to involve, and what actual trigger should initiate the communication.
When thinking about customer segments, it makes sense to first plan for the different stages in the customer life-cycle. This would be from recruitment to attrition, but also to look at the customers’ context. Are they for instance browsing, enquiring, or having just made their purchase?
Here are some of the segments we like to use:
Having analysed the segments to use, the next step is to plan the triggers that apply to them. Here some examples:
Clearly creativity needs to run alongside automation to spot these trigger opportunities, and to provide a sufficiently interesting a response to increase customer propensity to purchase.
Automating batch campaigns is however a very different type of activity, as in effect the marketer is blind to what the customer is actually doing. There is no pressing need to communicate at that point in time, but nevertheless sales must be generated.
For batch campaigns we like to use propensities as our alternative to triggers. Propensity models can for instance tell us the product category that an individual is most likely to purchase next. Whether they are at risk of attrition, if there is likely to be a return from sending them a catalogue, or if they are sensitive to price reductions and more likely to buy from a sale offer.
Just as with triggers, there are no limits to the propensity models that could be developed to score customers, but we have over time developed a short list of some that we find most helpful for batch customer marketing:
There is great value to be obtained from automating customer marketing, and as this paper will have shown, a successful outcome from it is as much a question of creativity in terms of how to go about it, as it is one of technology or data science. In fact, creativity, technology and data science need to work in combination for success.
If you would like to discuss partnering your creativity with our technology and data science please email to arrange a call.
UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.
Our technology focus has been to develop UniFida. Our data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.
Marketing is changing at an explosive speed, and our ambition is to help our clients stay empowered and ahead in this challenging environment.