-2.396 + (0.01554 x ATU) - Otherwise Known as My Best Customers

You are a mobile phone company. You want to sell premium air time packages. You need to appeal to a precisely targeted sector of a market on which you have a mass of multi-layered data. Who precisely should you target? Where should resources be focused to become truly customer-centric? In fact, if you are selling anything to anyone, how can you transform records of their past behaviour into accurate predictions of their future preferences?

That’s the question a whole range of professions, from sales people and marketers to new product developers and analysts have been asking themselves for a long time. A company’s customer data may be its treasure house. Yet in Celerant’s experience, it is not always easy to find the key, unlock the treasure and realise its true value.

There is always the expectation that data can be squeezed harder to make the voice of the customer louder – and some marketers are still waiting for it to become audible at all. As one Celerant client succinctly puts it: “customer data is our most valuable raw material. Our inability to refine it effectively grows more damaging every day”.

Given the potential value of customer data and the cost of resources needed to gather it, wastage of valuable insights should not be tolerated by any organisation. Yet it is often the case in our direct experience that over 80% of the customer data within a supplier organisation remains un-worked and therefore unused.

Celerant believe that more of same is not the answer. Organisations should be thinking not only in terms of gathering fresh data, commissioning surveys and running focus groups. They should be working more effectively with data they already have. We are leading our clients in some very exciting and productive new directions. And we are doing this through the application of data analysis techniques that were at one time considered more at home in heavy industry than playing a full role in enabling consumer insight.

Celerant have deep experience in using multiple regression techniques and in applying Six Sigma, Lean, Kaizen and similar methodologies to data gathering and analysis. Our approach has given customers such as BT and Reuters the ability to predict customer behaviour accurately and to impact sales. Many others have found that, with the right analytical expertise, the customer voice is in there.

Accurate customer data analysis makes the pivotal difference in a world where being truly customer-centric is a key business driver. It also affects every operational group within even the most complex organisations, as the following brief examples – taken from a deliberately disparate range of industries – clearly illustrate.

Computer games – loved by kids but their parents weren’t playing...

If you are tasked with the successful development of the next generation of ground breaking computer games, surely all logic suggests you focus every effort on surprising and delighting the kids who play the games?

Not exactly. Data analysis in the computer gaming marketplace highlighted an important group of ‘non-customers’ – AKA parents - who actually picked up the tab and who preferred their offspring to have a more active, physical expression of play.

From this insight, the Nintendo Wii was created. The ground breaking approach taps into a previously unmet need by making the playing of a computer game an energetic experience. Emphasis on broader based physical exercise, beyond thumb skills, expanded appeal to non-customer parents and restored Nintendo’s market share in computer games.

Parents don’t shop in child-unfriendly stores...

Data on the people who weren’t shopping in a large furniture outlet showed they were parents of young children who found the store too difficult to use with their children accompanying them. As a direct result of this data analysis, the store now provides crèche facilities to look after the youngsters, while their parents – relaxed and comfortable that their kids are settled - enjoy the retail experience.

Spotting potential bad debts early

A high street bank was suffering a high level of defaults on its mortgage payments. Repossessions were rising sharply. The bank re-examined its data and found that a customer who missed a mortgage payment once was 5% more likely than average to default on their mortgage.

Further analysis showed that someone who missed two or more payments was not 10% or 15% but a hugely significant 50% more likely to default on their mortgage. This insight allowed the bank to pinpoint potential defaulters earlier. And it enabled them to take preventive and mitigating action at the earliest stage.

From retail banking to furniture to computer games once the voice of the customer is clearly audible in the data forest, everyone from product developers to instore staff can benefit from the unique insights it provides. And this holds just as true of a wide range of Celerant clients such as Orange, Centrica, BT, Reuters and many more.

Our departure point is clear. Irrespective of industry or market, every operation has customers. And every operation that wants to grow profitably needs to configure every aspect of its operation around the preferences of those customers. Preference is expressed through behaviour. Every customer action is also the expression of a customer’s voice. And if you can hear the voice...

The mobile telecoms sector is literally about the voice of the customer. It is also one of the world’s most dynamic industries, where pressures to create the right product and service bundles to generate high margin revenue streams are acute. And it is a vibrant example of progression to real customer insight, through application of deep dive data analysis techniques.

Mobile phone upgrades – who buys them and why?

A major global communications network provider wanted to sell mobile phone upgrades to existing customers. They worked to uncover the critical customer preference factors that determine the most appropriate service offers. Capturing the customer voice was achieved by detailed working through a combination of historical data and consumer research, using exactly the techniques of data collection and analysis we have been discussing.

The company had 25 possible service packages to offer to customers, depending on the mobile phone selected. The range was potentially bewildering both for consumers and service provider, extending from navigation software, MP3 players and video gaming, through to voice and text only capability. For the sake of simplicity, we have only illustrated three of the service packages (A,B,C).

Analysis of research found that a total individual customer score against critical profile data of 15 for any package or more led to a 50-70% greater propensity to purchase that package than would be demonstrated by the average buyer.

Customer data capture - Mobile Phone

Propensity to purchase score over 15 will be 50 - 70% more likely to purchase than average

In this case the use of Six Sigma enabled the company to “industrialise” the collection of relevant data while identifying any missing elements. This was achieved through building a repeatable and sustainable process for customer intelligence gathering.

This process standardises and refines the marketing inputs, thus allowing translation of those inputs into more meaningful customer information. In effect, the authentic customer voice starts to be heard.

By utilising Six Sigma to refine the process of customer data collection and then applying so-called regression theory to establish the impact of key variables on the behavioural outcome, companies can rapidly build a picture that enables them to clearly target differentiated products and services to specific – and more receptive - customer types.

Using regression analysis to predict customer behaviour (and to answer tough questions from Sales and Marketing)

The data set below refers to that same study on how consumers’ choices of mobile phone plans are related to a number of explanatory variables. The charts show data on propensity to purchase a particular mobile phone plan.

The data displayed in the chosen graphical format clearly shows that propensity to purchase (P2P) is correlated positively with airtime usage (ATU). Moreover, there is now an equation to describe their purchase tendencies: P2P = -2.396 + (0.01554 x ATU). So we have a predictive equation that we can use with confidence to define future behaviours based on airtime usage. And we have answers to practical questions such as “what should we sell?” and “who should we sell it to?”

Conversely, the same data shows no visual (or statistical) difference when propensity to purchase is segmented by gender. Again, this is vital information for sales and marketing teams. Knowing how not to target is just as important.

Both these examples rely on investigation of the response variable (in this case P2P) versus a single potential influencer (airtime usage or gender). The graphical analysis has allowed us to visualise the correlation, or lack of correlation, much more clearly.

And because customer behaviour is rarely simple, there are other tools

What happens if we want to know more about customer behaviour in the face of many potential influencers? In other words, how do we continue to hear the voice of the customer in ever more complex situations?

One of the most effective data analysis techniques when life is complicated is the General Linear Model, or GLM. The GLM is used for the investigation of continuous or discrete responses versus several discrete factors, or a combination of discrete and continuous factors at any number of numeric or text levels.

The model will describe, and therefore enable prediction of, the behaviours of customers, based on the multiple and sometimes contradictory characteristics they embody.

The benefit of utilising Six Sigma data collection processes, graphical analysis and regression analysis together is that it allows businesses to understand the relative importance of data in establishing the customer’s propensity to purchase. Moving the pendulum of customer understanding away from ‘dark art’ and into the real science of robust data collection, translation, and analysis is a welcome re-balancing in many organisations. When it comes to the application of valuable development and marketing resource, most companies would rather go with validated knowledge than gut instinct or even a good hunch.

The additional benefit of a ‘joined-up’ approach to data collection and analysis is that externally facing functions – such as market research, sales activity and marketing campaigns themselves - become fully aligned with your operational process. The right products and services are developed and sold in the most impactful ways, based on real and current customer insight that also has an accurate predictive element. The consequent benefits of this end-to-end process based connection will be very significant over time.

Nor are the benefits restricted to a single project or a one-off intervention. A Six Sigma approach provides a process to ensure relevant data is collected in an appropriate and seamless manner, as part of the ongoing customer engagement processes within your business. So, once we have determined what kind of data is going to yield the best insights, it can continue to be collected and its salience remains high.

The bottom line is real customer insight. Six Sigma contains all the tools that can help you understand which elements of your customer data are significant. It can also ensure that your business processes capture meaning in a consistent and reliable manner.

The end results are impressive. Once the drivers of your customers’ behaviour are identified and their impact quantified, you know with great precision where to apply your precious resources.

Celerant clients who have worked with us to apply and develop Six Sigma-based approaches have seen significant improvements in the areas they now know are most appreciated by their customers and most likely to drive loyalty and sales. BT for example witnessed order cycle times cut by 55% and business benefits of over £80 million in less that 12 months. Reuters achieved over a 90% speed increase and an 80% quality improvement in specific key processes they know are most valued by their customers.

Customer Data and Customer Voice – the Truth is in There

Our departure point was the frustrating, and commercially restricting, situation of up to 80% of expensively acquired customer data remaining unexploited to its full potential.

This is worse than wastage. It is a barrier to effective response to the global driver of true customer-centricity. And it does more than undermine the credibility of marketers. It reduces the effectiveness of the sales force, has profound implications for product and service development and even clouds the issue of where an organisation’s next generation of innovation should be focused. There is no point working harder to get to market faster if it turns out to be the wrong market.

The direction in which our Celerant thinking has led in the last while has been towards the ‘industrialisation’ of the data collection and analysis process. This is about more than scale – although being able to interrogate large quantities of data is a necessity.

It is about gathering the right data in the right ways and then using proven techniques such as Six Sigma, regression analysis and the General Linear Model to achieve a real progression: from a mass of half-understood, misunderstood or simply ignored data to a clear picture both of past customer behaviour and its very probable implications for future preference.

The techniques are sophisticated. Working with the right expertise, our Celerant team has both deep and wide ranging experience, these techniques are also readily applicable both to the analysis of existing data and, even more productively, to the development of consistent and appropriate data gathering models for the future.

The destination we can take organisations to is one where an informative customer voice is heard directly from their customer data. This is, quite simply, a better place to be. Better today because the organisation’s existing product and service portfolio can be configured around, and pitched with more precision to, customers’ actual needs and preferences. Better tomorrow because true alignment can be achieved between external data gathering process, internal product and service development and marketing and sales functions back out to the customer.

From radically reduced order cycle times to impressive increase in the value of those orders, Celerant has validated our initial thought on behalf of our clients: customer data and a clear customer voice – the truth really is in there.

__________________________

About the Author

Dr. Paul Donnellan, Celerant Consulting
Process Excellence Commercial Leader

During his 18 years in industry and operational consulting, Paul has led a variety of process transformations and new product introductions. He has been a key player in Celerant’s leadership of transactional Lean Six Sigma for the service industry and has considerable practical experience of delivering results in the telecom, utilities, and financial services sectors. He is a regarded voice of authority for Design for Six Sigma and is a frequent speaker, chairman and author.

Paul is a certified Six Sigma Master Black Belt and a Design for Six Sigma Champion with a BSc (Hons) and PhD in Chemistry from King's College, University of London.