“Customer Analytics: [when] everybody talks about it but no one really knows how to do it”

“Customer Analytics: [when] everybody talks about it but no one really knows how to do it”

Once upon a time: there was a company.

The Company dealt with tons of data coming from all over the place. Customer data, customer service data, sales & marketing data, etc. The Company always thought it knew its customers, until one day: it realized it didn’t!  It decided to take action and undertook a transformational journey into the valley & cliffs of analytics. Here is a bit of the story.

A Case Study/Story

Let’s say The Company is Financial-Service and/or Retail. The following post will try to underline how, through analytics, companies should be able to (1) Know their Clients, (2) Enhance their products and services, and (3) Outperform their Marketing.

What if you could analyze a complete year of customer service related data?
What if you could analyze a complete year of customer service related data?

Know Your Customer

How can you know who your customers are? What if you could analyse a complete year of customer service related data?

  • All transactions your customer achieved
  • All communications you had with your customer on all channels
  • All your marketing and sales efforts and their potential measured impact
  • And finally, your Customer Satisfaction [CSAT “grade”] evolution along the year

Then, you’ll be able to cross reference all these data and populate it into myriad representations such as the following:

CSAT = function (customer transactions);
Chart#1: Identifying Customer Clusters

CSAT = function (Customer Spendings in €). “How happy or unhappy your customers are based upon how much they spend on your products & services?” (CSAT: Customer Satisfaction)

From the previous chart, you will notice Your Customer Clusters, beyond traditional segmentation. Customer Clusters are groups of “personas” holding, in this case, both similar CSAT levels while having a similar purchasing behaviour. As a result of these readings you start having a clear understanding of your customer base which will allow you to focus your future efforts on delivering a Customer Service Strategy based on these identified Customer Clusters.

Know Your Customer: A Survival Ability. Possible Today Through Analytics
Know Your Customer: A Survival Ability. Possible Today Through Analytics

Enhance Your Service

Obvious benefits of giving your company the possibility to Know & Understand Its Customers via Analytics include: Defining which KPIs are the relevant ones to focus on.  The ‘extra mile’ here should include figuring out how will those KPIs help shape your customer service strategy.

Let’s go over the following chart: CSAT = function (Average Waiting Time or AWT). We’ve noticed that a short AWT does not mean higher CSAT, on the phone channel, for instance.  This is true for multiple industries.

Customer Satisfaction in terms of Average Waiting Time (AWT) on the Phone Channel
Chart#2: Customer Satisfaction in terms of Average Waiting Time (AWT) on the Phone Channel

How should this be interpreted? It helps identify the range for which your KPI makes a clear impact on your CSAT in order to later be able to act upon it. Following this example we will “read” that on a KPI=AWT, a range between 60s & 90s does NOT make a big difference in terms of CSAT.

So, if  your currently on a 90s AWT: you should NOT try to enhance this KPI by bringing it down because the [financial, operational and HR] efforts needed to do so will be far greater than what you’ll be able to gain on a CSAT scale.

Now, remember the Customer Clusters? We can extract some relevant info from them. Let’s check it out.

Identifying Customer Clusters: Further Digging
Chart#3: Extracting Customer Clusters’ Properties via Data Processing & Visualisation Tools

And here comes the tricky part: each Customer Cluster will have different “expectations” with regards to the performance of the company and how it “treats” them. What does this mean? It means that each Customer “Persona” will generate a different CSAT [“grade”] given the same conditions.

How can we work this out? First, let’s take a step back and underline why this occurs. Often, companies will define, for instance, a QoS [Quality of Service] for their whole customer base. This won’t work.

In the same conditions, or QoS, or Customer Experience: your customer clusters will give you different CSAT grades. Customer Cluster #5 would for example be more tolerant than Customer Cluster #1 when it comes down to Average Waiting Times.

Enhancing Your Service: A Fine Equilibrium and Slalom Between "Categorisation" & "Personalisation"
Enhancing Your Service: A Fine Equilibrium and between “Categorisation” & “Personalisation”

Treating all your customer base “equally” will harm your business.

Instead, and following a QoS example: defining and delivering a Quality of Service per each Customer Cluster is a good first step.

A QoS based upon each customer’s cluster “value perception”: a good first step
Chart#4: A QoS based upon each customer’s cluster “value perception”: a good first step

Outperform Your Marketing

This should be the “easy” part. Once you have done all this previous work and you’re able to understand your customer clusters: you will focus on the NBA. And that means The Next Best Actions. 🙂

NBAs need to be simple, sequential, solid actions to take for instance Cluster#1 and work on moving it to the “sweet-spot” in the upper right side of Chart#5 (see below). For Cluster#3, an intermediate step is likely reasonable: from “3” to “4” to “5” or from “3” to “2” to “5” or both in an A/B Test type of fashion. This is an agile oriented set of processes.

Outperform Your Marketing & Sales : The Cherrie On Top, The Most Fun Part Of The Journey :)
Outperform Your Marketing & Sales : The Cherrie On Top, The Most Fun Part Of The Journey 🙂

A little bit more on our Clusters examples and then ciao:

  • Customer Cluster #1: these “Personas” have to be full on “baby sat”. They spend money and are not happy with us. Cocooning, ultra flexibility, special Care programs…
  •  Customer Cluster #2: a great cluster! A targeted set of outbound campaigns as the company drills down into the cluster properties will be useful. Within the cluster, different types of channels could of course be used.
  • Customer Cluster #3: a very important cluster. Opposite to millennials or digital-savy (all ages included), senior profiles need to be taken care of first. Here is where your customer interaction routing engine will play a corner-stone role. Special customer care should increase loyalty which in turn would drive sales.
  • Customer Cluster #4:  these guys & gals are neither happy nor unhappy, they also dont spend the targeted “shopping cart” amounts to spotlight them. These include millennials and digital-savies who will be more than fine if the company holds state of the start self service, automated virtual type of experiences.
Ex: Customer Clusters’ Next Best Actions for this case
Chart#5: Customer Clusters’ Next Best Actions for this case

In a Nutshell

An analytics project taking a one year data-scope on all your customers’ Transactional and Customer Service related data will allow you to: know your client, enhance your service and outperform your marketing & sales efforts thanks to crafted NBAs. If you are interested in further digging on this topic, i’ll be happy to point you to the right direction and resources.

P.S: Acknowledgements, many thanks and “spéciale dédicace” to Sébastien Parmentier, Data Scientist, “behind the scenes” Visionary and Colleague from Capgemini Consulting France.

Leave a Reply

Your email address will not be published. Required fields are marked *