As customer experience is a key competitive differential, what is more important to managing customer experience: data or process?
Process. Why? You can get away with bad data if you have a good process. The reverse is not true.
Bad data won't address your process. And even if your data is good, how will you rely upon it and trust it without good process?
Good process will be resilient to bad data, and allow you to trust in the data when it is good... (what ian said, basically ; )
In various contexts not given, I would agree with any of the above answers, but I would say, "neither."
If I understand the question, most important to good customer experience management is a good customer design process.
One that yields understanding of what those users need, not what our perspective tells us, then explores alternative designs.
Good process without good data feels like bad process. If I'm applying for a credit card with my bank, but my onboarding app doesn't know anything about me or my existing account—that's just irritating. Or I'm ordering a burrito, and I order the same thing every time, but the app doesn't remember that—annoying. So ultimately, bad process ≡ bad process but bad data ⇒ bad process.
A bad process can succeed with good data, better than a good process can deal with bad data.
Example: no matter how good your process is, if the credit card number, or the customer address, is wrong, it simply won't work. However, if the credit card number and customer address are correct, a very poor process might indeed succeed.
The answer, unequivocally, is YES.
I don't care how good your process is, if your data stinks -- forget about its suitability for analysis, how about even just basically ACCURATE -- then it doesn't matter how good your process is.
Example: the time I took over the payments for a relative's natural gas service, and TWO YEARS LATER I discovered neither my relative nor I had been receiving bills. Turns out the company updated the address but not the town (bad data), and suddenly sent me an invoice for the whole amount (good process).
Similarly, I don't care how good your data is: if your process stinks, then you'll never be able to properly leverage that great data to any kind of positive service effect.
Example: "Hey, I see you qualify for our next level of frequent flyer perqs! [good data] But I can't upgrade you, so you'll have to call back and ask for the right department. [bad process]" Yeech.
In both cases, I, as the customer, LOSE!
We’ve all had the client from hell – the one who tries to create exactly the old process in new software. We probably also worked out the solution – to inspire them with a vision, reframe the question and set new goals and milestones for the redesigned process to aspire to.
We’ve had the team member who won’t listen to reason. Insists their way is right. Often it ends up in a head-to-head pitting his credibility against yours. The skilful change manager, of course, doesn’t let it get that far – he doesn’t pitch in with a solution, he shares the problem and lets the data guide everyone to the best answer.
So we have a false construct here. What is most important is the process to gather, distribute and use the data.
The great thing about software is that it yields up data – lots of it – as soon as you get it up and running. So the first step is to create a Minimum Viable Product. You can use almost anything as your initial process – the existing process simply mapped out in the new software or even your client’s pet idea. Once it is running, the data shows everyone where the time is lost, the resource is stretched, the resilience is non-existent etc.
So how do you use this data? You can examine it in depth, apply Lean and Six Sigma principles and work out a stunning new process. No-one will thank you for it – indeed it is a recipe to set up the head-to-head. And even if you bulldoze it through and impose it, the new data will show up new opportunities and pitfalls.
Or you can involve everyone in a big experiment. Show the data visually, in real-time if possible. Explain what all the bits mean – latency, percentage utilisation etc. etc. Encourage them to think about what each means in real-life and inspire them to start thinking about how they can make the numbers go up (plan the graphs so that up is a good thing). Perhaps even give each person a project to lead a team to optimise each of the key parameters, coming up with ideas for everyone to test.
In stage 2, show them how they are inter-related – if one goes up, often another goes down. Let them see how one department affects another. Kick off a big debate about which tradeoffs are acceptable. And how often there is a third way which isn’t optimal for either, but actually works out better overall.
In stage 3 you have some models everyone can try. So you set everyone to work out how you will measure success. How to ensure you are only testing one thing at a time. How to statistically work out whether differences you see are meaningful.
Most importantly, you’ve built a team. Everyone wants to know what works best. Itching to create new ideas. And, most importantly, prepared to concede some of their personal turf for the good of the group. This team has a host of things to try and levers to pull, impatient to see results - they see the benefits of making it work and have invested themselves in finding the best way forward. They will pull doubters into line themselves. And arguments over whether BPM is the right solution have been completely forgotten.
So far we’ve only talked about employees and stakeholders. Once they’re experienced at building processes then you can introduce customers into the debates. Their employee created process has their ideas of what customers want to the fore. But at least you’ve now got that to work.
Challenge them to benchmark themselves against competitors. Who is providing the best customer service? Who has the best reputation with customers? Why? And how can you measure it?
Then challenge them to come up with something better than anyone else out there. You’ll be surprised at how well they’ll connect with customers and empathise with their concerns and how they can do better.
Because what is truly important in providing a good customer experience is engaged employees.
"of course it is X, because if X sucks, Y doesn't matter". I don't know how this proves much.
But to the specific question: I believe the more likely answer to this is data.
Customer experience management is first about understanding the problem (customer's pain points, needs, wants, aspirations, context) and second about understanding the solution (escalation flow, triggering actions and events). Sometimes the mere existence of the right data can render a perfect customer experience process useless, simply because the problem description does not match the solution at hand.
To give you a basic example: as we are working on our first start-up, we analyzed out competitors and all of them focus on a desktop solution, with accurate processes and forms and good tehnical integrations. However, our customer interviews show that they are always on the move and most of them are technologically illiterate (some of them couldn't even answer a call on a smartphone). They do everything by personal interaction and by phone and probably only half of them have a smartphone. No wonder our competitors have very little traction, and mostly around the corporate version of our customers (which we do not target).
So, knowing this, how do you bring, through technology, a relevant solution to their problems? Because for sure prescriptive workflows are not going the help them much and when they do, it's just basic flows. It is mobile-first - this is clear. Is it data-driven, event-driven, flow-driven, content-aware, context-aware - or a bit of everything? We'll see :)