How do you work out what customers want, when it is becoming harder to do so? It appears that customer segmentation or predictive models are no longer enough, but why?

Modern products are complex - Banks offer current accounts with a bewildering array of extras; telecoms companies offer tariffs with accompanying handsets, associated accessories and options – for our customers it has never been harder to work out what they want; for us, the indiscriminate “throwing” of offers no longer works if it ever really did. According to the Direct Mail Information Service, 55% of consumer direct mail is discarded unread and the rise of the Telephone and Mail Preference Services is putting increasing numbers of customers beyond reach.

While careful targeting of offers based on customer segmentation or predictive models can help it is not enough to guarantee either commercial success or the creation of a positive customer experience. To be successful in this more complex and discriminating world, interactions with customers need to be enhanced as an experience, offering propositions that are relevant now, not just in principle for those of the customer’s “type”.

Interactions with customers need a new paradigm

Our interactions with customers need a new paradigm. One that surveys their circumstances and interests, sensing their mood and converges on an understanding of what is relevant; one that informs and even educates as it unfolds; one that creates in the customer the confidence to act and gains credibility in ourselves as the provider of the “right” solution. While it might appear novel, in fact it’s an approach modelled on the way that humans interact – a natural two way exchange of information:

• The customer says something;
  

• The agent or system captures it, considers it and everything else that is known or can be inferred, and either works out roughly what should be suggested and asks any questions needed to confirm that, or recommends something – the resolution to a problem or the purchase of a product or service either in the attempt to get the customer agree or to prompt the customer to reveal more information;
  

• The customer replies or responds.

In short, it’s a normal conversation – but in this case, we are supported by technology that guides the conversation. As shown below in ‘Guiding a conversation: an illustration’, Chordiant has developed this approach to surveying customer circumstances and objectives that successfully converges on the customer’s agreement to act.

The approach is circular and iterative, holistic and convergent - each exchange with the customer considering everything, progressively homing in on the right offer at this time in this channel. This type of approach can be applied to resolve service issues, to retain defecting customers, to acquire, up and cross-sell or to an actual or potential circumstance or issue that warrants interaction with the customer. It can be structured as a needs analysis, as a market research survey or a data capture exercise. It continually seeks the next best questions to ask, the next best recommendation or action to take.

This is based on dynamic criteria and prioritisations varying, even during one conversation, from favouring relevant and simple to understand propositions that create confidence, to more valuable but harder to appreciate propositions that require trust, or from propositions that take time to present to crisper propositions if call centre queue lengths grow.

Optimise your propositions

The conversation can vary in style from the serial offering of one recommendation after another; to the dynamic creation and offering of a bundle of propositions with a dynamically selected incentive or pricing to reflect the value of the transaction or relationship; to a negotiated deal selecting propositions from a list optimised for this customer now. The style can be optimised for each customer as it’s how you do something as much as what you do that creates a positive customer experience. 

The approach can reflect the customer’s answers back to them as it makes suggestions, demonstrating that we are listening. It can illustrate the implications of any suggestion and even undertake what-if explorations of different scenarios, for example, that the customer doubles their use of text in the coming months.

This surveying of circumstances and objectives can encompass all the reasons why the customer makes contact such as service requests and product, need or event related enquiries. It can pose both questions to clarify a circumstance, need, event or interest and to qualify the customer for a proposition. It can capture the customer’s response to suggestions and recommendations.

Determine the customer’s mood

Sensing the customer’s mood can encompass sophisticated linguistic analysis to determine the customer’s mood. However, just an agent’s straightforward use of various smiley buttons and/or the analysis of immediately prior events and trends for problems can indicate the customer’s likely mood. This mood can change the minimum levels of relevancy at which various propositions are offered - increasing the minimum level of interest, for example, when the customer’s mood is ambivalent, decreasing that minimum when some goodwill has been created. Similarly mood can influence the relevance of individual questions or propositions and the choice of words, tone of voice or imagery used to communicate each proposition.

Critically, all of this, every reason, answer and response is captured to feed the conversation itself, inform future interactions with this customer and provide the basis for the development of generalised predictive models that will guide those future interactions.

Conversations in a shade of Orange

This conversational approach has been adopted by such companies such as O2 and AOL and is being taken up by Orange and by major banks. In each case, the primary objective is growth through making the most of the precious time in contact with customers. Alongside, costs are reduced through the more effective use of time and money – time on customers who warrant it; money on incentives for those who need persuasion, and rewards for those who deserve it. The results are impressive with customer satisfaction and business revenue up, churn rates and retention costs down, and average handling time often unchanged, just better used. Surveying and sensing customers is emerging as the leading approach to customer interaction – the irony is that as humans we have been doing this for ever, business and technology are just catching up.

By David Barrow,
Vice President, Vision, Solutions & Architecture,
Chordiant Software Inc

Web: http://www.chordiant.com
LinkedIn: http://www.linkedin.com/pub/dir/David/Barrow?trk=ppro_find_others

David Barrow is Vice President, Vision, Solutions & Architecture for Chordiant Software Inc, provider of Customer Experience (Cx) solutions to help leading global brands such as HSBC, Barclay’s, Capital One and O2 deliver the best possible customer experience. He is the CEO of the ‘decisioning’ company Chordiant acquired in 2004, he has nearly 30 years experience in developing innovations in predictive analytics, decisioning and the delivery and management of customer experience. For more information, visit Chordiant at www.chordiant.com.