The three questions you must ask when building a CX strategy - and how to answer them

A compelling customer experience (CX) strategy requires more than Voice of the Customer research, journey mapping and NPS or CSAT results as inputs.

The quality of CX strategy depends on the quality of insights underpinning it. Insight development – defining what research is undertaken, what data is collected and what models are developed – is a whole strategy in itself. Understanding how the different forms of insight help you answer the key questions (see red boxed area in Figure 1 below) enables you to determine priorities.   

This is the fourth article in a series describing the end-to-end process for developing valuable customer experiences. The first one outlined the overall process as described in Figure 1 below, with the two subsequent ones (see here and here) focusing on how CX professionals need to reflect business strategy in customer experience design.

Figure 1: CX strategy overview (click to enlarge)

The three questions to be answered

There are three high-level questions that you need to be able to answer when creating a CX strategy:

1. What do customers need and value and what will they buy from us?
2. How well are we serving customers and how can we serve them better?
3. How do we turn value created for customers into value created for the business?

The first question looks at desirability and credibility – what creates value for customers and what problems they perceive your brand can help solve. The second looks at your current performance against your brand promise and your customers’ expectations of what you will deliver, also identifying where improvements can be made. The third ensures that customer experience doesn’t become an exercise in trying to delight customers for the sake of it and there is a commercial rationale for what you are providing.

Sources of insight

Similarly, there are three primary sources of insights:

1. Research and analysis – research undertaken into customers and competitors; also internal analyses to identify root causes of problems, improvement opportunities and risk areas. 
2. Performance reporting – covering how well value is being created for customers in objective terms (Customer Performance Indicators or CPIs), how that is perceived by customers (using CSAT and NPS) and then how that flows through to performance on the metrics that matter for the business (customer acquisition, retention, etc.). 
3. Predictive analytics – creation of models to help optimise allocation of internal spend to where it is most effective, also to help predict what customers want to achieve so they can be better served.

Figure 2 seeks to map the sources (including their most important elements) to the three key questions, highlighting the role each source plays and its importance.

Figure 2: Key questions and sources of insight (click to enlarge)

What do customers need and value and what will they buy from us?

The most valuable insights into customers’ needs comes research and analysis. Customer research comes in a variety of forms – qualitative or quantitative, prompted (e.g. questionnaire-based) or unprompted (e.g. contextual enquiry), group- or individual-focused, etc. All of these can play a part but some deliver more value than others.

When the objective is discovering customer needs, qualitative research gives participants the opportunity to explore, identify and reveal their priorities, with one example being customer journey mapping. This requires identifying all the jobs that customers need to complete to achieve their desired outcome, the decisions they need to take, and which jobs are most important with high level emotions invested in the outcome (the moments of truth). This map can then be used for evaluating current performance (see next section).

Journey mapping can be undertaken in workshops with groups of customers or using ethnographic techniques where researchers observe a target customer in their home environment to see how they interact with existing services (including any workarounds they have created), asking questions to understand why they are behaving as they are if the reason is not immediately clear. 

Both approaches generate valuable inputs but they are expensive, meaning only a small sample of customers can be interviewed, thereby increasing the risk that those taking part are not representative of the customer base. Further testing with a bigger sample is valuable once the insights have been transformed into concepts, at which point brand fit can also be assessed. 

Quantitative techniques become valuable once potential propositions have been defined and the aim is to test how desirable they are to customers and how credible the brand is as a provider. A powerful research technique in this context is conjoint analysis which asks participating customers to choose between different packages made up of the same attributes but different levels within those attributes. By forcing customers to make 20-30 different choices, with feedback incorporated so the trade-offs become harder and harder, the relative value of each attribute and each attribute level can be quantified. (For a more detailed explanation of conjoint analysis, see here).

Competitor analysis also provides insights into what creates value for customers. Tracking competitor offerings and monitoring each competitor’s share of purchases or usage (if customer uses multiple providers) can be very revealing of customers’ real (rather than stated) priorities.  

In the context of understanding customer needs, predictive analytics helps organisations understand what customers want to achieve in online journeys so that the next best action can be suggested. Analysing journeys taken by previous customers (start and end points, routes taken, most frequent next step) enables you to predict with a high degree of likelihood the customer’s desired outcome so that they can be automatically guided to their desired conclusion.   

How well are we serving customers and how can we serve them better?

Customer satisfaction (CSAT) and net promoter score (NPS) research play an important role in understanding how well a business is doing with customers. And their value increases when the results are incorporated into a causal reporting framework.

Reporting increases exponentially in value when it is focused on causation enabling you to understand what factors are determining the observed results and what levers you need to pull to achieve the results you want to see. Causally-focused reporting links how well the business is serving customers in objectively measured terms (using customer performance indicators or CPIs) with customers perceptions of the service provided (using CSAT and NPS) with how that translates into the behaviours that drive KPIs for customer growth, retention, acquisition and cost to serve.

Such an approach enables you to understand how improving performance on CPIs improves CSAT / NPS scores and the commercial metrics that matter most to business leaders. Without an understanding of causation, decisions on how to improve performance are limited to guesswork and gut instinct.

Figure 3 below provides a simplified illustration of what a causation-focused reporting tree might look like.

Figure 3: Causal reporting tree (click to enlarge)

On the left are the targeted financial outcomes – increasing revenue and decreasing costs. Feeding them are the primary CX strategy outcomes and KPIs, with the operational metrics that drive KPI performance to the right again. In this simplified example, customer acquisition is driven by referrals from existing customers, the success of campaigns in generating leads and lead conversion. And for cost to serve, the key operating metrics are the number of interactions (by type per customer) and the average cost of each interaction type.

On the far right are the CPIs. These are the customer-facing equivalent of company-facing KPIs. KPIs track performance on creating value for the business, CPIs track how well you are creating value for customers.

The CPIs that are appropriate will depend on your value proposition – whether your are seeking to differentiation through delivering superior convenience, responsiveness, choice and flexibility, service quality, cost savings, risk reduction or simply lower prices. If the value created for customers can be measured in monetary terms, then this is a great measure. For example, the comparison service MoneySupermarket tracks the savings it delivers to each customer on insurance, loans, energy and TV/broadband renewals versus staying with their existing provider. In total these savings amounted to over £2 billion in 2018.

Creating such a reporting system has two major requirements – the development of causal hypotheses (if we do A, then B will happen) by the CX team and the creation of a data architecture that enables customer-level data to be captured and stored. The former defines what should be measured and the assumed relationships between variables, the latter ensures that differences in treatment and resulting behaviours can be observed at a micro-level and aggregated as appropriate.

While reporting can answer the question ‘how well are we doing with customers?’ the second part – ‘how can we do better?’ requires research and analysis.     

CSAT and NPS scores provide a temperature check on customers perceptions of the service they are receiving, but improving those scores requires more detailed analysis, particularly textual analysis of explanations of why those scores were given (alongside textual analysis of complaints, social media comments, etc.) Being rated 3 in NPS terms will trigger a desire to improve, but unless there is an understanding of the rationale for that score being given (online forms were confusing and difficult to complete, waited a long time to speak to someone in customer service, passed from one department to another, etc.) the ability to improve is limited to guesswork. 

Similarly, the volume of issues and complaints are indicative of service quality, but reducing the number of them requires root cause analysis – repeatedly asking ‘Why?’ something is happening a minimum of five times or until the root cause of the problem is revealed. Each ‘why?’ peels away a layer to reveal the underlying problem. 

Let’s take an example of your company having a surge in the number of complaints from customers about the difficulty of completing an online form. At first glance this may look like a simple web site fix but asking why reveals that the form does not reflect the current product line, suggesting that form updates should be incorporated into product release processes. Asking why there is this disconnect reveals that the product development and service teams won’t co-operate, and that they don’t co-operate because they have conflicting performance management objectives, and that there is no linkage between their KPIs because the culture of the organisation is highly siloed. As the true problem emerges, the solution requirements evolve.

Improving customer journeys also requires digging deeper than reporting in the form of customer path analysis and customer journey reviews.  

CSAT and NPS scores provide a temperature check on customers perceptions of the service they are receiving, but improving those scores requires more detailed analysis.

Customer path analysis uses customer data from the web site, CRM system and transaction processing systems (any data that can be linked at the customer level by a common identifier) to show journeys across different company touchpoints. It reveals the frequency of routes taken by customers – where they start and end, which result in a transaction or other positive outcome, which journeys are working well and where there are high drop-offs.

Customer path analysis is particularly useful for identifying incomplete journeys (where customers don’t complete their mission) and duplicated effort – for example, where a customer first went to the FAQ section of the website and then called the contact centre – which creates a poor experience for the customer and increases the cost to the business. In both cases the fixes are often readily apparent. 

The journey map created as part of understanding customers wants and needs can also be used with research participants to capture the as-is experience. This looks at which customer jobs are currently painful (i.e. time-consuming, effortful or stressful) to complete, how decisions on what to do next are made, what is the prevailing emotion at each stage versus what the customer would like it to be and what steps create emotional extremes, with particular focus on how well the moments of truth are being addressed. The primary focus of customer journey reviews is to find where you are falling short. But as they look beyond the scope of your service, customer journey reviews also identify where there are gaps in complementary services – gaps that you may need to help address to improve the overall customer experience.  

How do we optimise value for the business?

How reporting helps answer the question ‘how do we optimise value for the business?’ is described above as it should be integrated with creating value for customers. And as with improving the value created for customers, improving the value created for the business requires research and analysis to understand how performance can be improved.

Market and competitor analyses play a key role in prioritising customer strategy objectives. When markets are growing very fast, prioritising acquisition over retention, cross-sell and reducing cost to serve makes sense. When markets are more mature and competition has consolidated, acquisition becomes harder. Competitor analysis identifies those with fewer resources and less compelling propositions so that acquisition activities are focused on customers of these weaker competitors rather than all competitors generally (the latter can quickly become a zero-sum game). Competitor analysis also plays a key role in mitigating churn – identifying which competitors are most likely to steal your customers so relative weaknesses in offering can be addressed.

Similarly, share of wallet analysis identifies the opportunity to grow revenue with a particular customer. It is critical in B2B markets where customers source similar services from multiple providers but is still relevant in B2C markets such as financial services and retail. Identifying how much customers are spending on services you can provide, how that expenditure breaks down and the share of expenditure that you and others have enables you to take a targeted approach to increasing both products and revenue per customer.

As with improving the value created for customers, improving the value created for the business requires research and analysis to understand how performance can be improved.

Pricing analysis enables revenue yield optimisation – a key driver of profitability with the impact of price changes flowing quickly through to the bottom line. It ranges from simple analysis of prices charged to customers against volumes to check adherence to pricing corridors – particularly important in markets where sales managers have some discretion over discounts offered – to more complicated analyses that look at price elasticity (something which conjoint analysis can also help with).

Channel value analysis plays an important role in helping reduce cost to serve. It calculates the cost of managing an event (e.g. inbound query, purchase) by channel (e.g. call centre, live or asynchronous chat, self-service, etc.) and matches that to the value added by handling that interaction in that channel (e.g. conversion rates, future cost avoidance). It enables you to identify where low value interactions are being handled in high cost channels (also when and with whom) so that interventions can be designed to better balance value and cost.

Achieving channel shift also requires predictive analytics – calculating channel use propensity scores to identify which customers can be encouraged to use lower cost channels. Predictive analytics also directly impacts the KPIs of customer acquisition, retention and growth through identifying the most likely candidates to buy, defect or cross-purchase.

Customer lifecycle analysis also plays a role in cross-selling through identifying next best actions – which customers are most likely to deepen their relationship with the business without any intervention, which ones will deepen it if nudged and which ones will never deepen it, no matter how much you market to them.

Customer lifecycle analysis helps you to optimise marketing spend – ensuring that marketing dollars are spent where they will be most effective. Customer profitability analysis plays a similar role. Matching marketing spend to customer profitability – or customer lifetime value (which blends customer lifecycle analysis with customer profitability analysis) – ensures decisions are made with business value in mind.

The benefits from this type of analysis are both significant and quickly realisable. At an online retailer I worked with, customer profitability analysis showed that they were rewarding customers whose behaviour made them intrinsically loss-making (for example, buying the same product in three sizes and returning two of them) by offering them loyalty discounts which made them even more unprofitable! Ceasing these inducements led to an almost immediate uplift in profits.

How these different analyses can be combined to help you decide which customer strategy objectives you should prioritise will be the subject of a subsequent article in the series.