SATSE: Part 5: Future Directions

It's hard to make predictions, especially about the future - Yogi Berra

"Who the hell knows?"

So said Edie Falco, who plays Carmella Sorprano in the TV The Sopranos. With refreshing honesty, let's be honest, celebrity comment is not known for the truth, Falco told Reuters in April 07: "Who the hell knows?" answering a question about what happens next after the demise of The Sopranos which ran for six series. As with careers, so with marketing, communications and technology. Woody Allan has a great expression and a personal favourite of mine: "If you want to make God laugh, tell him your plans for the future."

So if it's all so difficult, why bother wondering? Terrible things might await us in the future. Prognosticate at your peril. Well, peril does not scare me, in fact, I embrace peril, I yearn for peril. And so, fixing a safety helmet to my head and proudly displaying my boy scout badges on my chest, I plunge into the future. To examine the trends and bring forth a view (and many may challenge it) of how service delivery will evolve over the next few years. For the sake of keeping this relevant I'm taking an evolutionary approach rather than a revolutionary approach to the future, so no flying cars in this book, but even evolving what we have today may well seem pretty revolutionary.

Before looking forward, I want to look briefly back. Can you remember the first time you heard the word 'blog' or the word 'podcast'? Couple of years ago? Perhaps a small, small number of you can go back four years, but that's the point. Four years. We have photobloggers and videobloggers, mainstream TV programmes like BBC Breakfast doing a daily breakfast podcast. For the vast majority of us it didn't exist four years ago. It's a true phenomenon. A publishing milestone. A great, great moment in history - thought, type, publish. Not 'thought, type, look for a publisher, get fed up, give two fingers to the establishment, return to work in the call centre'. No, think, type, publish. Now. And millions of people are doing it. If that can happen in four or so years, what can happen in the next four, never mind the next forty. Flying cars anyone?

Chapter 17: Service Technology Options

• Moving to IP


• Video Contact Centre

Business Intelligence

Business Intelligence - you may snigger at the phrase and place it next to other great oxymorons, such as "airline schedule", "unbiased opinion", "political science" and "People's Republic of China". "Business Ethics" is another good one. Whilst I can see the joke, for business often presents itself as effort without intelligence, there is another meaning to the phrase. That being "the collection, organisation and distribution of business facts to improve decision making". The facts might relate to current stock positions, stock refresh rates, stock-keeping unit (SKU) growth rates, for example, where data generated from the business operations are used to record progress against targets and to make decisions about future activities.

Management Information (MI) is another common (and related term) found in call centres. BI is perhaps a more technical source of information for managers, relying on what can be measured and is focused on actual performance. MI is a wider term for intelligence (including opinion and estimations) that might inform management decision makers. For the purpose of this book, I am going to assume that BI is a subset of MI in that the representation of facutal information can be used as a key input (but not the only one) to management decision making. That said, the terms are often used interchangeably. I cannot count the number of times a client has asked for their MI, when what they actually mean is BI, but there you go.

Business Intelligence has some common attributes:

• metric - the name given to a particular measure of BI, say, abandon rate, call duration, etc.


• aggregation - taking large numbers of transactions together to create more representative figures, for example, average call duration


• granularity - being able to look at different levels of detail, for example, first call resolution percentage at across the floor, or by team or by agent


• filtering - setting criteria to reduce the volume of transactions studied to those of interest for a particular purpose, such as, average wait time for calls placed before 9.00am


• calculation - often some core data is used to calculate new pieces of data, such as percentages or averages which are better at highlighting variance in data observations


• exception - if some event happens, report it, such as if abandon rate exceeds 2% of calls presented, report it


• compliance - showing a target and an actual performance for a particular metric, say average handle time, which was expected to be four minutes (the target), but the actual for the period being measured was four minutes and thirty seconds, or a variation of plus 30 seconds, or 12 1/2 per cent

The point of BI is to inform, to allow better decisions to be made and ultimately to improve some business driver - customer service, profitability, etc. The value that BI has to the business manager is the extent to which it aids their decision making. A few years ago, I had a very cynical MIS (Management Information Systems)lecturer who told a story, possibly apocryphal, of two managers in a large corporation who receive their copies of the monthly MI, delivered in paper form (this was well before MI systems developed dash-boards and online query tools). One manager immediately bins the heavy report but the other says "I find it much more valuable" and proceeds to remove the elastic band from the papers, retains the band and bins the paper. The managers had clearly discounted the monthly MI as being too cumbersome to be of any use. Of all flaws that MI can have, this is not the greatest but it is a bad one - it's as much use as no MI. The worst sin, however, is inaccuracy. This can be very serious as interpretations made from incorrect data can have a devastating effect on business performance.

Generating too much information is a fault either of poor MI design or of user request overload. I once had a client that asked for a monthly dash-board - a quick snapshot of business performance in the contact centre. To start with the dash-board had one page with four quadrants, each representing a key metric graphed out for legibility and with room for a brief comment underneath. Over a period of six months, this "dash-board" evolved from one page to twelve pages. A twelve-page dash-board. Talk about oxymoronic. The eleven pages that followed the original page were filled with large grids of data and were text heavy. The last page contained an explanation of some of the codes that appeared in the previous pages. Despite the client being one of the very best to work with, in this regard, they had gone completely nuts and gone overboard to the point that the dash-board was not fit for purpose and met none of the common attributes of a dash-board: brevity, simplicity or clarity.

Software as a service / hosted applications

Chapter 18: Modeling & Customisation

I like golf. Playing it and watching on TV. Doubtless many others do too. I also like skiing, but not watching it on TV. I'm sure many others do too. What about enjoying golf, watching golf, enjoying skiing, visiting cities, photo-blogging, keeping rabbits, reading The Onion, The Economist and The First Post, visiting America and Italy and trying to make the perfect salsa. Starts to reduce the number of people like me from millions to, possibly just a few. We're all like this, there's lots we have in common with people but we are also very individual. For a call centre, this presents a problem, how to you best represent yourself to people with widely differing perspectives and interests?

Well, generally a contact centre is not setup to handle calls from skiing, golf watching, rabbit keeping, Economise reading photo-bloggers. Contact Centres are typically arranged around simple transactions, like a home broadband help desk, a charity donation line, a utility payment centre, etc. For this reason, contact centres tend to avoid individualisation and concentrate on very simple segmentation of customers, usually on one dimension, for example, value. So, "select" customer (those that spend at the highest level) get one level of service and the great unwashed get another (inferior) level of service. This is easy to enable, simple to communicate and operationally, easy to manage, however is a short-sighted approach, if not actually fully wrong-headed.

These days a boadband provider is rarely just a broadband provider (likely they provide phone service and perhaps mobile communications too), a charity is rarely about just fundraising (issue promotion and member get member, spring to mind) and utilities are rarely just utilities (infrastructureless organisations offering gas, electric and other services such as extended boiler warranty, home care, etc). This is the era of the multi-faceted brand, of service extenstion and massive cross sell. So, is the same tone of voice to all consumers still looking like such a good idea?

Of course, going too far the other way leads to operational inneficiency, you can't have a team setup just to offer a great service to golf-loving, Economist reading, rabbit-keeping photo-bloggers, just in case one should call. You might wait a long time. Still, there is value in looking at improved segmentation and tailoring service and offers more closely to the real issues people care about. For example, as hard as EDF Energy try, I really can't get excited about Gas and Electricty offers that involve Nectar points (the combined loyalty scheme with diverse members such as Sainsbury's, BP, etc) however, perhaps if they spoke to me about something that I am interested in, relating their meat and two veg product to something that I care about, might that make a difference?

Segmentation works in two ways. For service and for targeting

• Managing data across channels


• Profiling schemes, MOSAIC, etc


• Heat Maps


• RCV, CLTV, Scarcity Allocation Models


• Managing massive customisation

Chapter 19: New Distribution Models

• Asset Digitisation


• Centralising or Dispersal of Assets?


• UGC/Web2.0/Social Networking - What does it mean for Contact Centres?


• RSS feeds and podcasts & blogs - What happens when advisors self-publish?

Copyright 2007, Robert A Innes