Big Data. Big Deal?
A common theme that we pick up in every contact centre is that they track their KPIs. Assiduously. No, fanatically. In fact, a cynic might say that the contact centre is set up to serve its KPIs and not its customers.
Traditional theory is that if you hit your KPIs then you’re doing well in the contact centre. And once you’ve hit them, then let’s raise the bar a little more and we’ll be doing even better. Let’s add to that a high positive Net Promoter Score, and all’s well. Or is it?
We’re starting to see a new theme developing and it’s that of discovery. Yes, KPIs matter, but, to coin a phrase, you don’t know what you don’t know. So slicing and dicing what’s going on in the contact centre requires a different method of analysis.
Almost all KPIs are based on some benchmark of achievement for transactions tracked in a database. Mining this data requires a hypothesis of what to look for. For example, the shorter the calls in an inbound call centre, the more efficient you are at handling them. Right? Perhaps, but how do you cater for correlation of resolution of the customer’s problem with the call duration?
What most contact c managers have at their disposal is a set of standard reports. It might be possible to vary these slightly, but in the red hot heat of your average operational day, let’s be honest, you don’t get the time to go off piste and start to apply the grey matter. Instead, what happens is that assumptions are made against a theory about the cause of a problem, and poor decisions are made. The outcome of this is that many contact centres are constantly tweaking and changing but rarely do they get the source of a problem.
Even if you do get the time to do some proper discovery, good analytical skills of the sort needed to splice together many different sources of data are not easy to come by, and are generally not possessed by your every day contact centre management team. The result is that many organisations hire teams of financial analysts who take away the data, drop it into spreadsheets, and then a few weeks later they emerge with a report that tells you something that happened last month.
A perpetual guessing game against out of date spreadsheets really can’t be the way to run a multi-million, customer-facing investment, but for so many organisations that run contact centres, that’s BAU.
So what to do?
The Business Intelligence (BI) market has some amazing tools. They are well-established, proven, and backed by armies of men in grey suits with PhDs. Great empires have been built on them (Jim Goodnight, who built and owns probably the most technically advanced analytical vendor, The SAS Institute, is reputedly worth around seven billion dollars). But when you get started, how long before they actually deliver return on the not insubstantial investment? Even if you have the licence to use the products in your company already, what do these vendors know about running a contact centre? Not a lot. They can dazzle you with number wizardry that would make Dumbledore look like a remedial student, but the vast amount of customisation required to even get to first base really just doesn’t make good financial or operational sense. And even then, could they take you significantly upstream of your current measurement of KPIs?
Enter stage left, “Big Data”. But let’s be honest, is this any more than just a cool new epithet for the old analytical tools and data warehouses? In the majority of cases, probably not. However, where I do agree with exponents of Big Data is their assertion that when combining and analysing multiple data sources in new ways, often you can make startling new discoveries that you couldn’t possibly have seen before.
We’ve recently been working with a client, a Business Process Outsourcer. To them, KPIs are absolutely everything. They aren’t just fanatical, they follow KPIs with a truly religious zeal because if they don’t hit the right service levels, then they don’t get paid. When we first approached them, they said to us that they have their KPIs fully under control and understood. When pressed, they would admit that they had a team of people using Excel to assimilate the data, and when they finally received the reports, they were a month out of date. On the issue of contemporary reporting alone, they became interested in what specialised “Big Data for the Contact Centre” might reveal to them, so we embarked on a project to analyse just a couple of data sources.
Constrained by their fervour for the existing KPIs, we replicated those using our new method and tools. What followed, based on just a month’s data, was astounding. We were able to replicate the KPIs fairly simply, but what was fascinating was that we were also able to correlate data between different sources and pick up previously unseen causes of poor performance (down to a detailed level). They really could not have done this using the tools at their disposal today.
This one short project revealed three key pieces of insight, resulting in efficiencies of just a few percentage points, but monetarily, well into six figures. As we develop the project further, and unleash the remaining 80% of the product onto their data, we expect to be making annual savings into seven figures.
As just one Contact Centre in a UK market of about 6,000, I feel pleased that we can help them take their operation to a new level of productivity and service. But there’s much more to do on my mission to improve the industry. In the UK, “Call Centre Operator” is currently the single most common profession, applying to about a million people, not far short of 5% of the working population. The industry is horribly inefficient and comparable to the cottage industries of the 1900s when labour was cheap and the way to scale a business was to throw more bodies at it. What our industry needs is the application of understanding and knowledge to drive improvements in efficiency in the same way that the great engineers drove the industrial revolution. In that way businesses will rein in the spiralling costs of customer contact and maybe actually improve their service at the same time. And who knows, perhaps their KPIs might improve too?