Sympathy vs. Empathy

I read a great piece by Dave Trott this morning on the difference between sympathy and empathy, and what it means for advertising. I think it’s at least as relevant for customer experience.

There’s an extensive academic literature about how exactly to define empathy, which is a rabbit hole I fell into hard a couple of years ago. I did my best to provide a summary in this post. But for now, let’s not worry about whether Trott is using the words exactly right, but look at how he defines them.

Sympathy, he argues, is when we give someone what we like in order to make them feel good. Empathy is when we give them what they like.

When it comes to advertising, that means that professionals should put to one side the idea of ever saying “I like…”, because whether or not they like an advert is irrelevant.

Instead they should focus on what works, and understanding why it works. With good advertisers, according to Trott,

“There was nothing random or subjective in their work, everything was there for a reason.

Consequently, absolutely every aspect of it was open to question.

Because every aspect could be explained logically.”

This is absolutely true for good customer experiences as well. I’d suggest two rules:

1. Don’t design an experience that would work for you, design one that works for customers
2. Design every aspect of the experience deliberately

That second one is so important. You need to know why you’ve chosen a certain font, or music, or store layout, or proposition, or channel mix, or lead time, or whatever, and that decision should never be because you like it.

How do you gain that empathy? That’s what customer research, especially qualitative customer research, is all about.

To design good experiences for customers you need to understand what customers value and why it matters to them. You need to forget about your own likes and dislikes. You need empathy, not sympathy.

Choosing a customer loyalty strategy means thinking differently about profit

You’ve probably seen us say that our approach to business is founded on the idea of a customer loyalty strategy.

In a nutshell that means that your strategy is to find and keep the right kind of customers, customers who will reward good experiences with loyalty behaviours such as retention, recommendation, and increased share of spend, becoming very profitable in the long term.

One stumbling block that prevents many businesses from fully embracing a loyalty strategy is that it means you have to think differently about one of the fundamentals of business — profit.

Products aren’t profitable, customers are

To understand a loyalty strategy, you have to accept that products aren’t profitable, customers are.

This is a point that was made very powerfully in the classic book “Driving Customer Equity“. In it the authors argue that chasing profitability can be fatal. They showed, for example, that many organisations enter what they call a “product profitability death spiral”, by focusing on profitability at a product level.

The logic seems compelling: you look at the profitability of your products, and then eliminate the ones that are least profitable to focus on those that give you the best return.

Sounds sensible, doesn’t it?

The problem is that products aren’t profitable, customers are profitable. If you remove products that are important to customers, then they’ll go elsewhere to find them, and that means that you lose sales on your profitable products too.

A similar logic applies to aspects of the business which are seen as costs, like the contact centre, and which matter very much to customers even though they don’t in themselves add to the bottom line.

Profitability has to be understood at the customer level, and that’s why customer lifetime value is such an important tool for companies who want to pursue a loyalty strategy. It helps you to understand the ways in which loyalty leads to profitability (often increasing profitability over time) and identify customers with the potential to be loyal and profitable in the long term — ”barnacles” not “butterflies”.

Good profits and bad profits

This might seem anathema, but not all profits are good for business, at least not if you take the long term view. The failure of business-themed gameshow The Apprentice to understand this is one of the reasons that I can’t watch it without shouting at the TV.

Fred Reichheld coined the phrase “bad profits” to point this out, but it really should go without saying.

It’s easy to turn a short-term profit by mortgaging the future of your business. Conduct a fire-sale of your assets, sack all your staff, promise customers things which you know you’re not going to be able to deliver. Short-term profits will look great, but acting like this is obviously not sustainable if you’re hoping to do more than avoid getting fired for one more week.

“Good profits” are those that customers give you willingly, because you’re creating value for them, and those that do not come at the expense of future profits by weakening the ability of your business to meet customer needs.

All of this leads to a simple principle, if you want to pursue a loyalty strategy (and we think you should):

Don’t chase profit, chase profitable customers.

Are you falling into the all or nothing trap?

I should have gone out for a run today.

I wanted to do 6 miles, but there wasn’t time, so I didn’t do anything. I couldn’t be bothered to go out and do the 2 or 3 miles that there was time for.

This is an example of something I call the “all or nothing” trap, which is one of the most pernicious mental traps that we, as individuals and organisations, tend to fall into it.

When it comes to the customer experience, we often decide that if we can’t do exactly what the customer wants us to do then there’s no point doing anything.

Or we decide that making the ideal change to a system or process is too difficult, so we stick with things as they are.

This kind of black and white thinking is a trap because there is almost always something you can do that will help the customer experience. Even if it’s not perfect, even if it’s not quite what the customer wanted, something is a lot better than nothing.

Do something now

After a couple of decades of working with clients on improving the customer experience, I can almost always tell during the final presentation whether or not a client is going to take the results of the survey and use them to make things better for customers.

If you want to be like those that do, my advice in three words is: do something now.

Right now, today. Do something that you believe will improve the customer experience, even if it’s only a tiny thing.

If you’re worried about finding budget or building a business case for something that’s going to need investment, then find a way to test it on the cheap. With a bit of creativity it’s almost always possible to prototype something well enough to prove that it works, even if you can’t roll it out to all customers.

Many of the best CX improvement ideas are free (or very cheap) and end up saving the organisation money. One simple example is the power of checklists, which I’ll talk about more in my next post.

Buy a copy of Tom Peters’ The Little Big Things and you’ll find, according to the cover blurb, “163 ways to pursue excellence”. Most of them are small, many are cheap, and one of them must be applicable to you!

Avoid paralysis by analysis

The biggest warning sign of an organisation that’s not going to make any significant change is that the questions they ask show that they are looking for reasons not to act.

It’s very easy to get stuck in “paralysis by analysis”, endlessly debating what customers meant by their score, or which priorities for improvement should be chosen, or whether it would be more accurate to re-run the analysis with 12 customers removed.

None of that is going to make things better for your customers.

Once again, this is where a “test and learn” or “fail fast” mentality works well. Rather than spending ages debating which approach to take, or whether something will have a positive or negative impact on customers, try it out and see.

Pilots, prototypes, and tests will teach you more than any amount of discussion.

Don’t fall for “too early to tell, too late to change”

Edward Tufte, the “godfather of information design”, had a brief career in consulting, which he gave up because of one particular style of resistance to change:

Products existed only in two states: either too-early-to-tell or too-late-to-change.

Edward Tufte

The more you test, and the smaller and more agile the scale on which your projects work, the less likely you are to be susceptible to this.

Doomed projects lumber on because no one can face dealing with the consequences of them failing, when a small-scale test at the beginning would quickly have highlighted problems.

Resist complexity

It’s easy for plans to become overcomplicated. One reason that the “do something now” mantra is effective is that it’s almost impossible to overcomplicate at such short notice.

How many times have you given up on a complicated workout plan because you didn’t have all the right gear, or you lost track of which exercises you were supposed to be doing each day?

One climber came up with this daily exercise routine: do 1 pull-up, 3 press-ups, and 5 sit-ups; then do 2 pull-ups, 6 press-ups, and 10 sit-ups; then do 3 pull-ups, etc….and keep going until you collapse. It’s easy to remember, you can do it pretty much anywhere, and it’s surprisingly effective (it neglects your legs, but that’s climbers for you!).

Is it perfect? No. Is it better than nothing? Absolutely!

You want to turn making changes to improve the customer experience into a habit, and that means you have to keep it simple. Build small ideas and practices into the everyday, whether it’s meetings or individual ways of working.

Don’t bet it all on investing in a new CRM system, but make something better for one customer, right now.

Learning from your happy customers

“Your most unhappy customers are your greatest source of learning.”

Bill Gates

This quote from Bill Gates often turns up when people are talking about the importance of customer experience, and the power of surveys to help you fix problems.

Is it true?

I think it’s a simplification that reveals a mental trap that all of us are prone to — we assume that improving the customer experience is about fixing problems. The truth is that fixing problems can create a good experience, but it probably won’t create a great one.

In this post I want to talk about what you can learn from your happiest and most loyal customers.

Who are they?

The first question is “Who are they?” These customers are obviously a good fit for you, and they’ve shown that they will reward good experiences with loyalty. That’s a great basis for refining your strategy.

In fact the basis of a customer loyalty strategy is to identify this type of customer, make sure that your proposition is tailored around their needs, and then do everything you can to attract and retain them.

Why are they so happy?

What did you do to make them loyal? Your research should give you insight into why your happiest customers are so satisfied and loyal. Often what you’ll find is that there is a foundation of reliable delivery — doing what you said you would do — but that on top of that is something extra.

It might be helping them out when they needed something quicker than normal, building a relationship of trust over time, or delighting them with an unexpected extra.

By definition these things are often very individual. They don’t simply scale up, but by taking the time to notice and review them, you can understand where the potential lies for building these stronger links with customers.

William Gibson once said “The future is already here — it’s just not very evenly distributed.” and what I’m arguing is that the same is true of excellent customer relationships. They already exist, but you need to build more of them.

Are you missing the most important part of driver analysis?

There was a time when key driver analysis was an expensive add-on to your customer survey, an exclusive offering from swanky consultants with bona fide statistical chops.

During the early 2000s it was democratised, and increasingly seen as a fairly standard part of the customer survey process.

Unfortunately, somewhere along the way, the world lost sight of what key driver analysis is really all about.

What is key driver analysis?

I should start by explaining what key driver analysis is.

The techniques used can vary, but key driver analysis is all about identifying which aspects of the customer experience are most strongly linked to customers’ overall feelings as measured by overall satisfaction (or sometimes likelihood to recommend or another outcome measure).

Targeting our efforts to improve at the key drivers means that we can focus on the things which will make the most difference to the overall experience. There are some caveats, which I don’t want to get into today, but the basic idea is sound.

How well does your model fit?

Key driver analysis was traditionally calculated using multiple regression. Nowadays people who know what they’re doing will use one of a number of more sophisticated techniques that build on multiple regression while overcoming its flaws (e.g. principal components regression or relative importance analysis).

In each case we’re building a model that attempts to capture the relationship between a set of drivers and an outcome variable, which gives us two key bits of information:

• The strength of the link from each driver (the Beta coefficients in multiple regression)
• The success of the model as a whole (the R²)

Why don’t we pay more attention to R²?

Time and again I see the results of a key driver analysis reported with almost no attention paid to the model R².

To understand why it matters, we need to know what this “coefficient of determination” tells us. It can be interpreted in a number of useful ways:

• How well does our model fit the data?
• How accurately are we able to predict what overall satisfaction score customers give, based on their other scores?
• How much of overall satisfaction is explained by the items on the questionnaire? (You can think of this as a percentage, so an R² of 0.64 means we can account for 64%)

All of these are slightly different ways of saying the same thing, but it’s the most important thing about your model — how good is it?

What don’t you know?

Something obvious, but often overlooked, is that the R² also tells us how much of overall satisfaction we can’t account for. If we can explain 64% then that means we can’t explain 36%.

That seems like a pretty important thing to think about.

What can account for that missing explanatory power? There are two main explanations…

The questionnnaire is missing something important

If your R² is low, then it may be the case that you are missing something that is having a big influence on the customer experience. You need to try to figure out what it is.

When did you last use qualitative research to find out what matters to customers? Have you updated the questionaire to reflect their current priorities? Have competitors recently launched a new innovation or way of working?

Speak to customers and see if you can figure out what your questionnaire is missing.

There are distinct groups of customers with different priorities

The other common explanation is that you have groups of customers with very different priorities. For example you might have “price-driven”, “quality-driven” and “convenience-driven” customers.

Multiple regression assumes that it is working with a homogenous group of customers. If this isn’t the case then, even if your questionnaire covers all the drivers that matter, you can end up with an overall R² that seems low.

It may help to try fitting your driver model on subgroups of customers such as segments and demographic splits, but often it turns out that these different priorities correlate only loosely to things which you can measure. Latent class regression can help, if so.

What R² tells you

Key driver analysis is a great tool, but if you don’t pay attention to your R² then there’s a danger that you’re missing the most important information about your model.

It will tell you whether your model is doing a good job, help you make sure your survey is complete, and flag up the possibility that different customers have importantly different priorities.

If you only judge your model on the driver weights, then you may be working hard on things which aren’t the biggest driver for any of your customers.

Have you tried explaining it to a rubber duck?

There’s a popular technique for programmers called “rubber duck debugging“.

If your code isn’t working the way you’d expect, you get hold of a rubber duck and explain to it, line by line, what your programme is supposed to do.

At some point you’ll reach a step in the explanation where what you want to happen next is not what your code is actually doing, and you’ve found your problem.

Something very similar often happens when you get people from different departments together to talk something through (like, for instance, in a service blueprint session).

Being forced to articulate, in simple terms, what each step in a process is supposed to achieve can expose places where there’s a disconnect between what you’re doing and what you should be doing.

Even better, you have to explain why you’re doing what you’re doing. That’s often how these cross-departmental conversations turn up myths and misconceptions in the ways we work together.

It’s not uncommon to find out that steps in the process are more complicated and time-consuming than they need to be because they’re based on a misunderstanding of what another department needs, or outdated beliefs about timescales.

So, if your process isn’t working the way it should, have you tried explaining it to a rubber duck?

Sometimes questions reveal more than answers

I’ve been reading the AMSR report “How we’ve changed“.

It’s a great read if you’re interested in thinking about some of the ways UK society has changed since the middle of the last century, and perhaps even more to show how the ways in which we use research to understand those trends have evolved as well.

One thing that struck me was that the questions and response options tell us as much about the society of the time as the answers do.

Questions reveal our values

We often see research as a neutral activity, disinterestedly reporting on the opinions and behaviour of the audience of interest, but the truth is that the research itself reflects the biases and values of the people who design it.

This idea of “reflexivity” is something qualitative researchers have long been used to dealing with, but quantitative researchers haven’t always been so open to the idea that their techniques are susceptible to biases that they bring to it.

There’s no doubt that they are.

Quantitative researchers, and anyone else using data to derive insight, need to think about three big tests for any piece of research:

Have we spoken to the right people?

One of the biggest invisible problems with many research projects is the biases built into who responds, and who is invited to take part (two separate questions).

The first is a matter of non-response bias, and can be addressed with pre or post-weighting. That’s an article for another day.

What about who we invite to take part?

In an interesting article about the changing food habits of people in the UK, the report notes in passing that a TGI survey switched in 2002 (!) from asking their questions to the “main shopper” rather than to the “housewife”. What a world of assumptions are baked in there!

These kinds of bias are very difficult for us to see when they reflect norms taken for granted in our culture, but leap out when we look back at them.

How do our questions affect the answers we get?

Researchers worth their salt will already be very careful about leading questions (e.g. “Would you agree that…?”) or biased rating scales (e.g. “Excellent, Good, Average, Poor”).

We can all make mistakes from time to time, which is why it’s good to get your questionnaires checked properly, but these have never been seen as good practice.

More pernicious are the questions which include assumptions about society and the way people live their lives. We’ve already seen some interesting assumptions about “housewives”, and littered throughout the report are examples of attitudes to topics like race and sexuality which would not be seen as acceptable today. As the report says,

“Pre-conditioned cultural narratives have embedded these biases into societal constructs making them easy to reinforce and more challenging to remove.”

Judith Staig

Another sources of bias is the list of options that we offer to respondents. If we’re asking people which celebrities they admire, as a 1993 MORI survey did, the choice of who to include in the list may speak volumes. Lisa Stansfield came top, by the way!

Is your question measuring what you think it measures?

Surveys can turn up apparent contraditions. Some data from 1984 showed that 81% thought immigrants should be treated like other Britons, whilst 47% thought race relations legislation should be abolished because “too much is done to help immigrants at present”.

What seems like a contradiction comes about because people are complex, and attitudes are (to some extent) formed at the point the question is asked. Everyone likes to think of themselves as fair, whatever their views about immigration.

The contradiction only appears if we make the mistake of assuming that the response to a single question can reveal whether someone has a positive or negative view of immigrants.

This comes down to validity. In other words: are your sure your question is measuring what you think it measures? Don’t fall into the trap of thinking you know why people gave a particular answer unless you dig into the detail to find out.

Will we do any better?

It’s easy to look back and judge the past, even the relatively recent past, but will we do any better?

I think it’s important to consider the biases and assumptions that are inevitably there in our research, and do what we can to minimse them.

Technical matters like leading questions and biased scales are one thing, but what about those baked-in assumptions, and the angles that certain questions invite customers to take?

One obvious solution is to improve the diversity of the industry. If more voices and perspectives are represented in the design of the research, then there’s a much better chance that no one set of assumptions has the opportunity to distort our findings.

When in doubt, make a map

I love maps. From John Speed to the OS, and best of all the beautiful Swiss maps, I’ve spent many happy hours pouring over sheets of paper that bring the world to life.

Maps are also the secret to a trick shared by workshop facilitators, qualitative researchers, and dataviz types: if you want to think through ideas, look for patterns, or identify contrasts, make a map.

A map has two dimensions: up/down and left/right, and plotting objects against those dimensions allows us to understand them more clearly because we can take advantage of what our brains do well (seeing) instead of relying on things they do badly (thinking).

When we think of maps, we tend to think of geographic maps. In this case the dimensions are North/South and East/West, like this:

We can look at that and immediately see that the position of cities in the plot reflects their position in the real world. It allows us to see which cities are close together, and to judge the relative distances between them fairly accurately.

Maps aren’t just for geography

But the reason maps are such a powerful tool for thinking is that the two dimensions can be anything we want them to be. We can map any variables we like onto those dimensions, and when we do so we’ve invented the scatterplot! We could choose, for example, population and average house price:

Once again the position of the cities in the plot lets us see which are close together, which are far apart, and to judge just how far apart they are. That corresponds to real facts in the real world, and visualising those relationships in this way makes them easier to think about. Scatterplot “maps” are, to my mind, a chronically under-used tool for looking at relationships, spotting patterns, and identifying outliers.

But we’re not limited to quantitative maps, those dimensions can as easily be based on judgement as measurement.

A consensus map

Maps are one of my favourite tools to use in workshops, because those dimensions can be literally anything. Working together we can agree where items should lie along each dimension without needing any kind of quantitative measure.

It’s a great way, for instance, to help people prioritise ideas based on how achievable they are versus how impactful they would be:

This is a simple tool that I come back to again and again because it works so well in almost any situation. In fact it’s one of a handful of exercises that I keep in my back pocket knowing that they can inject life into a workshop that’s flagging.

It’s a good illustration of how easily visual thinking can be incorporated into your day-to-day work of thinking or facilitation, without any need to worry about whether or not you can draw. Simple frameworks like this allow us to tap into our visual skills, and create a shared picture of the world.

A map of your thinking

And there’s no need for collaboration, maps are just as useful when it comes to getting your own thinking in order, and especially when it comes to explaining that thinking to other people.

Let’s say I want to summarise different research methods in terms of the tradeoff between the amount of insight you get from each person (“depth”) and the cost/difficulty of getting a large sample (“scalability”). I might come up with something like this:

Clearly there’s nothing quantitative about this. Each of those points could be put in a different place based on different (equally valid) opinions and criteria, but it helps to quickly communicate how I think about the subject.

Maps, as I hope I’ve convinced you, are one of the most powerful tools we’ve invented to help us think about, communicate, and document the relationships between things.

They have a place when we’re trying to articulate our own thoughts, when we’re trying to achieve consensus to make a decision, and when we’re analysing and presenting data; and I think we don’t use them anywhere near as often as we should.

When in doubt, why not try a map?

How to decide what to do to improve the customer experience

As I’ve talked about before, “actionable insight” is a phrase that needs to be unpacked before it starts to mean anything:

Insight = Findings + Interpretation
To achieve insight, you need to combine the right research findings with interpretation to make them meaningful.

Actionability = Insight + Collaboration
To make it actionable, you need to combine insight with the knowledge of your colleagues about the way the experience is created.

Change = Actionability + Planning
To achieve change, you need more than good intentions, you need to plan concrete steps to address those actionable insights.

Picking priorities

One of the key aims of any customer survey should be to identify what we call “priorities for improvement”. This may seem a bit basic, but it’s amazing how common it is to see customer research that fails in this regard.

Your priorities should not be based simply on where the lowest satisfaction scores are, but on a combination of importance and satisfaction. In practice with our clients we use about 5 or 6 criteria when choosing priorities, but this idea of “doing best what matters most” to customers is the foundation.

But customer insight isn’t just about identifying where the problems are; it can guide how you address those problems, and help you identify what to do to improve them.

What kind of priorities?

Depending on your current performance, you’ll find that your priorities may fall into three categories, which you’ll be familiar with if you know about the Kano model, although the terminology varies. It’s crucial that you identify which category each priority falls into before you start to take action.

Givens/Satisfaction maintainers
These are core business items which everyone does fairly well, which means that you don’t get much credit for doing them better. It also means that, as long as your performance is at an acceptable standard, customers won’t really notice it, and they’re certainly not a reason to choose one supplier over another.

What you’re looking out for here is the proportion of dissatisfied customers. If you find that more than a few percent of customers are giving low scores for a given, then you need to figure out what’s going wrong and do something about it. Failure on a “given” drives customers away quicker than anything else you can do.

Your aim is to take low scorers and persuade them to score you 8 out of 10. Here’s a 3 step plan:

• Identify problem areas. Find out who your low scorers are, and what makes them different from other customers. Is it a type of customer whose needs you misunderstood, or who were sold something that is unsuitable? Is there a particular branch that has a problem?
• Read the comments. Use verbatim feedback from dissatisfied customers to understand what went wrong, look for root causes, and suggest solutions.
• Emphasise consistency. Make sure your messaging to staff is clear – on givens you’re not looking for new ideas, you’re looking for totally reliable, consistent, delivery.

Delighters/Hidden opportunities
At the other end of the spectrum, you may sometimes choose as a priority something which is a “delighter”, in other words which customers don’t notice until it is significantly different from what everyone else is offering. These can be a compelling source of differentiation, but be aware that customers won’t put up with problems in other areas just to secure some “nice to haves”.

When it comes to improving customer satisfaction here, you’re looking for opportunities to build differentiation. It might be innovative solutions, ways of working that no one else is offering (can you make customers’ lives easier?), or stronger relationships. Don’t let these distract you from doing the basics consistently well, but they might help you make the step from “good” to “great”.

This time what you’re looking for is ways to turn 8s into 9s and 10s. Again, here are 3 things you can do:

• Focus on the true believers. Flip the given logic on its head and look at who’s scoring you 10 out of 10. Who are they, and what makes their experience different from everyone else? How could you do more of that? What can you learn from the Account Managers who go above and beyond for their clients?
• Steal ideas. Who does this really well? Usually you’ll have to go outside your sector to find ideas, but if you can be the first to spot the opportunity to change how a sector works the potential rewards are enormous.
• Build empathy. The more you understand your customers, the more you’ll see opportunities to improve the way your product sits in their lives. How could you make it easier or more useful for them? This is where a design thinking lens can take your approach to the next level.

Satisfaction enhancers
These are the “more is better” attributes from the Kano model, where satisfaction is linearly related to performance.

This means that it makes sense to look for systemic improvements that will impact all customers (rather than focusing mainly on the extremes at the bottom or top end of the scale), as well a trying to eliminate dissatisfaction and grow top box satisfaction.

So with so much to go at, where do you start? This time your 3 things are:

• Zero defects. If dissatisfaction is more than 10%, start by taking a “givens” approach to diagnosing and eliminating poor performance.
• Good to Great.If you have a lot of 7s and 8s, then take a “hidden opportunities” approach to figure out how to get more 9s and 10s
• Systems. Take a step back to view the system that creates the experience that customers are scoring. What are all the influences, feedback loops, and unintended consequences that get in the way of a higher score?

It never ends

Improving the customer experience is a journey without an end – the goalposts keep moving!

One thing you’ll notice over time is that the priorities migrate. They may start off as hidden opportunities, but as your competitors start to copy your innovations you’ll find them morphing into satisfaction enhancers. In time they’re likely to become givens, as everyone in the market reaches a level of competence in that area.

As long as you understand that different types of priority demand a different approach, and as long as you have the research you need to understand them and identify actions to improve them, you have the tools to keep improving the experience.

There’s always a trade off

“There’s always a trade off…Good design is not maximization of every response, or even compromise among them; it’s optimization among alternatives.”

101 Things I Learned in Engineering School

We use research because we want to understand customer decisions. We want to know why.

Why did a customer choose Product A over Product B? Why did they defect? Why did they recommend us?

When customers make those decisions, they’re making a judgement based on their perception of what will make them happiest. A classical economist might talk about maximising utility (“this one is the best value”), and a behavioural economist might focus on heuristics like satisficing (“that’ll do”).

Modelling trade-offs

Conjoint analysis is a research technique that organisations sometimes turn to when they want to predict or understand how customers make decisions. It tries to replicate a real consumer choice by asking respondents to trade off attributes at different levels.

We’d all like a safer car and a more fuel-efficient car, but if it really came down to it which one would you choose? How much more would you be prepared to pay?

Sometimes these attributes are traded off in pairs, and sometimes (rarely nowadays) as a full profile, neither of which seem to me to do that good a job of replicating how customers really go about making decisions…but that’s a post for another day.

What I want to talk about is not the trade-offs that customers make, but the trade-offs we make when we design products and experiences.

Trade-offs in design

I started with a quote from 101 Things I Learned in Engineering School, which is a great book. I always think it does us good in the world of Customer Experience to learn from designers who have to worry about people dying if they get things wrong!

What does this particular quote have to teach us? That when we’re designing customer experiences we have to recognise two truths:

• Yes, there will always be trade-offs. You can’t give your customers more information and make the document shorter.
• The answer is not compromise (average amount of information, average length of document), but choosing the best overall option.

But how do we know what the best option is? What if we end up designing for a hypothetical “average” customer that doesn’t exist, and end up with an experience that doesn’t work for anyone? That’s the problem with the compromise approach.

Imagine we’re designing a mortgage application journey, and think about the needs of first time buyers versus serial doer-uppers. One group has never done this before, and is going to need lots of hand-holding, the other wants everything to be as streamlined as possible. How do you design an experience that works for both of them?

Designing flexibility

This is where your research can help, with tools such as personas are a great way to reveal the different needs of different customers (or the same customers at different times). But even when we have a good understanding of what those different needs are, it’s not always obvious how to design around them.

Those choices need to be made holistically, looking at the entire experience, and we need to design in flexibility to make sure we accomodate different needs. One size rarely fits all!

In person to person interactions a lot of this flexibility can be provided by a member of staff with the right knowledge, emotional intelligence, and freedom to use it.

In digital journeys, it can be effective to have help or additional information on hand for those that need it, whilst making the journey streamlined for those who are confident with what they’re doing.

These are both great ways to design flexibility into your journeys, but don’t make the mistake of presenting too many choices to customers. As The Paradox of Choice showed, too much choice can make us less happy, not more. It increases the effort we have to make in the moment, and it can cause us to second-guess the decisions we’ve made.

Your proposition can help with this by making it clear what type of experience you’re aiming to create, and who for, but you will still need to offer some flexibility.

Great customer experiences are those that feel tailored to our needs, whether or not they really are, that give us the opportunity to be in control, but don’t overwhelm us with choice. Not a compromise, but an optimisation among alternatives.