Sunday, March 2, 2008

Customer metrics: What should you measure?

By Neil Davey

"Be careful what you wish for," the saying goes... "you might get it!" And this could be particularly apt when it comes to customer data.
There's been a spike in the demand for customer metrics recently. Firstly, an increasing number of CEOs are recognising that non-financial measures such as customer satisfaction are as important to their investors as traditional financial figures, a fact emphasised in Deloitte's 2007 study 'In the Dark'. But equally as significant is the increasing accountability that is being demanded of marketers, with the sector being asked to demonstrate its value now more than ever.

What CEOs and marketers may not have realised until this point, however, is the enormity of the task that faces any firm trying to cut a swathe through the mass of customer data that is at their disposal. Put simply, companies are up to their eyeballs in customer information – and they don't necessarily know what to do with it.

As highlighted in Deloitte's study, for instance, whilst leaders have an excellent idea of what traditional financial figures to use, they are bamboozled by customer data. 87% of companies are happy that their financial measurement is good, Deloitte reports, but only 29% can say the same about non-financial indicators.

And many marketing teams are similarly unacquainted with customer metrics. A 2007 study by VisionEdge Marketing revealed that whilst 78% of respondents track leads to conversion, only a quarter track and measure the rate of customer acquisition and fewer than 10% measure customer lifetime value or customer advocacy. Furthermore, a third of the marketing professionals questioned omit metrics from their marketing plans altogether.

Without a doubt, the sheer volume of customer data that is out there presents a daunting task. It's little wonder that firms are asking themselves what customer analytics and metrics they should – and shouldn't – be focusing on.

A broader focus

If you are looking for a showcase example of a company that has put metrics at the heart of its business, then supermarket Tesco is an obvious choice. However, according to Andrew Jordan chief operating officer of beyondanalysis, it also represents a good example of where customer metric models can frequently go wrong. "There are two fundamental flaws in its model," says Jordan. "Firstly, it relies on a very heavy level of transaction data and secondly it only attaches itself to customers."
An oft-quoted problem associated with transactional data – as with focus on similar financial measurements such as profit margins – is that it encourages leaders to drive their firm using 'the rear view mirror'.

"Purchases, repeat visits, length of call time… many companies track key performance indicators (KPIs) to monitor the successes and failures within the business – including customer satisfaction and churn rates – but the data produced only tells you what has happened and nothing about the underlying drivers of these trends," says Gary Schwartz, VP of product marketing at Confirmit. "The CRM industry is based on examining historical purchase behaviour in order to unlock the secrets and predict purchase behaviour but more often than not, however, repeat purchases are simply a function of lack of other choices!"

A focus solely on customers is similarly misplaced, as it results in the metrics completely omitting anyone who has failed to consider shopping at Tesco or those who have proactively decided not to shop there. "Anyone and everyone has the potential to become a customer, whether they have been a customer in the past or not," stresses Jordan. "So to use metrics that only attach themselves to known quantities is very traditional... very CRM-based. It misses a vital dimension because all you are doing is looking at things like share of wallet and repeat purchases and traditional value. They're all very well, but you've got to start earlier in the journey and understand how these things came about in the first place. A lot of the advice that we give companies is to think in a broader context about how they go about aligning the same metric approach to things like customer acquisition as they do with their own customers."

The internet in particular has created a wealth of data on non-customers for firms to exploit according to Jordan. "I'm referencing the rather murky world of social media, but also the fact that people are now collaborating electronically a lot more and that is creating a very rich stream of data which tells you a lot more about how people lead their lives, why they make the decisions they do, which ultimately inform purchasing decisions."

With the field of 'customer experience' gaining growing prominence, it's no surprise to learn that firms are increasingly looking to apply metrics to 'experiential' aspects. But with so many firms running on ‘command and control' metrics, this doesn't necessarily mean that companies are any better at delivering the experience to their customers that their brand values demand. Indeed, with it could be argued that in many cases the result has simply been that they only deliver what they measure. And without input from – and empowerment of – those employees 'at the coal face', the firm may not even be looking at the most appropriate metrics.

"It's clear that the metrics set at the top of the organisation – usually around shareholder aspirations – dictate the behaviours of that organisation toward customers," suggests Tony Mooney, consulting and propositions director at Experian Integrated Marketing. "These are rarely customer experiential metrics, in our experience. The nearest many organisations get is the use of customer satisfaction surveys and average call answering statistics - neither of which adequately measure customer experience.

"For example, most companies with call centres use average call answering as a KPI. This is merely a hygiene factor and, anyway, is usually inaccurate as it measures call answer times from the point at which the poor customer has made it through several layers of IVR. Of far greater importance than how quickly you pick the phone up is how the call is handled, eg single contact resolution. For other organisations, customer behavioural metrics will be key – such as downgrading and requests to cancel. Understanding the key customer metrics is a process of sound causal analysis – to identify the important drivers of customer behaviour and monitor those. Too many organisations try and manage 'output' metrics and miss the indicators."

Getting the metrics mix right

Clearly the quest for a single customer metric that holds the key to success for every company is a futile one. Behavioural metrics and experiential metrics have an important role to play alongside the more traditional ones for the modern business. But different metrics will hold a different value for different firms. So is there a way to establish the most important metrics specific to your firm?

One approach is the balanced scorecard. The balanced scorecard, arguably the most widely-used management framework of the last 50 years, allows firms to take all the potential metrics available and weight them and then track them over time. The process would, for instance, involve firms drawing up a list of key customer goals - perhaps customer satisfaction, new customer acquisition, customer retention, customer loyalty, fast response, efficiency, reliability or image – and then creating a number of metrics to measure success in the fields – which could consist of a focus on customer satisfaction index, repeat purchases, market share, on-time deliveries, returned orders, new customer acquisitions or perceived value for money.

"The question is whether all customer measures are of equal importance – and if not, how do we decide what we should be focusing on?" says Dana Guthrie of the performance management group at Actuate. "No two organisations have the same strategy, so every company needs to make its own judgement about the most important customer measures for themselves. The [balanced scorecard] process, though, is always going to be the same: a top-down approach of applying your own unique strategy and objectives to decisions about what to measure."

On the face of it a common-sense approach, the balanced scorecard actually involves a rigorous process to select and define the key measures that will ensure successful strategy execution.

Some scepticism of the technique's effectiveness in this area exists, however. "It strikes me very much of management overkill and of trying to design a complex mechanic for something that shouldn't be that complex," says Jordan. "And the problem of doing that, not withstanding the actual process of putting a balanced scorecard together in the first place, is that you'll be continually trying to challenge the validity of the scorecard rather than the validity of the results. The danger is that you will over complicate something that doesn't need to be complicated."

Nevertheless, whilst the balanced scorecard approach is not a guarantee of success when it comes to incorporating customer measures into the performance management mix, it has proven popular – as its longevity attests.
Professor Robert Shaw, though, believes there is a far simpler way that companies can identify the most important metrics and jettison those that are surplus to requirements – by evaluating their value to the decision-making process.
"The first thing is that people need to focus away from the data and onto the question of decision support," he suggests. "Many firms haven't stepped back and asked themselves if the data is actually supporting their decision making. You should ask the question: what are the main applications in this in terms of decisions? You can do audits of how you are applying the data and analytics technology and expertise to answer key decision questions. What comes out of those audits is a great deal of clarity about the value of the technology and data to the decision-making process. And then you can start to prioritise them."
Certainly firms need to take some action to wrestle control of their customer data – a problem that has been especially exacerbated by the internet. "There are a few hundred new metrics available to firms that they can capture that they didn't have a few years ago," agrees Neil Morgan, VP marketing EMEA, Omniture. "It's the biggest change I've seen in consumer marketing. Most traditional business people are struggling to interpret it or action it. The big requirement is to be able to raid this, adapt these metrics and make them useful in a business."
But Shaw insists that there is a dawning realisation amongst some firms that the focus should be on quality – not quantity.

"We used to get a bucket load of data, now it is like having a fire hose pointed at us - and firms don't know what to do with it," he concludes. "A radical rethink is needed. Companies need to take an axe to research and cut the stuff that is not illuminating the decision-making process. The enlightened companies are already doing this. They are asking themselves what decisions they are taking – whether it is to do with direct marketing, or pricing, or how much they advertise, or product/service quality. And then they are asking themselves what they can actually do about something like service quality and how they can measure if that has an effect which ultimately finds its way back to the financial results of the company. And unless it throws some light on the way that service or price or whatever hits the bottom line, then these enlightened companies simply won't be interested in that research."
The rest of the market, however, may yet be rueing the day they wished for more customer data...

No comments: