Bookmark and Share

Tuesday 20 October 2009

Why are contact centre desktops so broken?

My colleague, Toby Baker has created this guest posting based on his observations from working with Corizon's customers and partners applying enterprise mashups to integration problems in contact centres.

I spend a lot of time in contact centres, analyzing applications in use on the desktop, and see the same challenges and issues again and again. This is my top ten:
  1. Too many applications, each with their own login, UI and screen layout
  2. Applications that time out when they are used, forcing another login
  3. Applications with too many screens, each in a tab, sub-tab or sub-sub-tab
  4. Applications that require way too many clicks to get the job done
  5. Applications that show you the customer name on the first screen, only to not show it on all subsequent screens, forcing the user to write it down on paper
  6. Applications that perform so badly the agent has to become an expert in ad-libbing
  7. Screens where the information is way down below the bottom of the screen, or all the way across to the right
  8. Screens that are so crowded with information, it’s hard to pick out what you need
  9. Screens that are so visually unappealing it makes them hard to use
  10. “Customer Records” that are split across many applications and so require a large amount of copy/alt-tab/paste to see the whole picture
Everyone knows that a unified desktop is a good idea, and fixing the ten problems above is a good thing. So why is it so hard to do? Let’s look at the problem in more detail:

First of all, the applications - why are there so many of them?
  1. Companies merge, acquire and get acquired. Technology stacks don’t get integrated quickly, or at all.
  2. Core CRM applications have huge programmes associated with them to implement front- and back-end CRM processes. It takes years to get round to doing “the other apps” so we end up with original apps as well as a new CRM front end.
  3. To stay ahead of the competition, enterprises launch new products and services before the core applications can really support them, so new applications are built.
  4. Agents build their own applications to help navigate through the complex desktop. These are known as “Shadow IT” or “Guerrilla” applications.
  5. Technology is “improving” all the time. There are new widgets and frameworks and standards and approaches that IT tries to keep up with.
Secondly, why is the data all over the place? Why is the thing that I want buried deep, across multiple screens or a few scrolls away? Why are apps so hard to use?
  1. Databases organize information in a way that works for databases. Customers have orders, addresses, contact details. Orders have statuses, products, dates. Products have prices, catalogue entries. All too often, applications use the structure of the database to drive the structure of the screens. Mostly, users don’t look at information in the same way as computers, and nor do customers when they are asking for it.
  2. The tools used to create applications make it easy to use the data layout to drive the screen layout.
  3. Tabs, tabs, tabs. It seems like the right way to organize information, but demands change, so it rapidly becomes the wrong way to organize information.
  4. It’s sometimes easier to add something to a screen than it is to create a new screen. You then get a crowded screen with loads of unnecessary fields on it.
  5. The people designing the applications understand about data. The people who use the applications understand about the servicing of customers and their products. There is a mis-match here!
  6. The people building applications don’t sit down with call centre advisors enough. If they did, they would not build apps in the way that they do!

Thirdly, why do I have to look in lots of places? Why aren’t the sequences of activities I need to go through - the "user processes" - catered for?
  1. User processes change faster than apps do. Customer service organizations manage sales processes, retention processes, diagnostic and fix processes. They need to learn based on what works and what does not work. They need respond to market conditions, competition, regulations. Importantly, they need to do this without changing systems. Unfortunately, as user processes are all too often wired in to applications (or on paper), this is too hard to do.
  2. Sometimes it’s easier to create a new application than modify an existing one.
  3. “One size fits all” does not work in a customer service organization where you have a variation in process across each user group. Maybe you have retention, inbound, outbound, first line and second line. Maybe you have organized teams based on product specialty. Either way, using traditional development techniques, it’s almost impossible – or very costly – to create applications that are tailored to the needs of each user group.

Any more thoughts? Please leave a comment. Toby will be back with more on addressing these problems in a future blog.

Labels: , , , , ,

Friday 16 October 2009

Quick thoughts from Oracle OpenWorld

Just come back from Oracle Open World 2009, where the atmosphere was very different from last year: the key focus was on making things works with a lot less hype. Most of the Oracle keynotes I attended insisted, demo after demo, on the progress made in integrating different applications from the portfolio with fusion middleware. The CRM team did a great job in demonstrating new features and functionality delivered in the latest releases. Fusion Apps were announced but remained relatively low key. One key differentiation from the existing apps seem the clear focus on user experience. Early analyst feedback seem to confirm the success achieved there ( see Ray Wang or Paul Hammerman)

The feedback from customers was also focused on integration and innovation to take costs out. One key theme was "instant gratification": looking for technology and projects that would deliver very short term benefits at low risks but without building new legacy. Retiring legacy applications seem high on the agenda as well.

In walking the Exhibition Floor, it seemed less product companies were exhibiting than last year with SIs and other specialist consulting firms having replaced them. One exception was the presence of with a large booth and keynote to boot. Talk about extreme coopetition! The scale of OOW was as impressive as last year and the organization flawless which was no small achievement considering that more than 40,000 people attended. The power of OOW remained in the great potential for networking and meeting customers, partners and analysts. We had many very productive conversations in a few packed days which made the trip more than worthwhile.

Many good reviews were written about OOW. I have selected a few below that I thought were particularly worthwhile reading:

- Paul Greenberg
- Denis Pombriant
- Esteban Ekolski
- Michael Fauscette

Labels: , , , , , , ,

Sunday 11 October 2009

Thoughts on Forrester Business Technology conference: lean times and enterprise mashups

Last week David and I attended the Forrester Business Technology Forum in Chicago. The main theme of the conference was adopting “lean” approaches to IT – which seems a particularly appropriate term for the lean times we find ourselves in.

The lean concept applies across the full range of activities performed by an IT department, but I was particularly struck by the discussion on its use in application development in John Rymer and Dave West's session*. Four key ingredients were seen as essential to lean software:
• Making sure deliverables are “fit for purpose” for the business problem at hand
• A clear focus on hard, measured value
• Simplifying the delivery platform – moving away from overbloated infrastructure stacks if they aren’t needed
• Allowing efficient evolution – avoiding stovepipes without creating undue overhead

Overall, this combination of getting quickly to the right, “just enough” solution and then improving and reusing, while measuring all the time is both powerful, and with the benefit of hindsight obvious! However, it is sobering to remember how far from this approach so many enterprise application development and integration projects are.

Much of “getting to lean” is about methodologies, management techniques and people. However, the right technology for the right problem can provide powerful enablers if chosen appropriately – another key point from the sessions. Discussions at the conference confirmed that - when it comes to integrating applications for people - process based enterprise mashups can provide a key to a leaner approach:
• Encouraging building directly from the desired user requirements under the control of the business and process owners
• Allowing user and process execution to be clearly benchmarked and measured to show improvements
• Pragmatic, light-touch, standards based integration
• A first, “just good enough to make a different solution” in days, followed by iterative improvements and spinning out reusable “building blocks” as a natural by-product

Lean is an important theme. Using enterprise mashups won’t get you there by itself, but it encourages and helps you to work in a lean way – which can’t be a bad thing!

* How can lean software enable you to better serve the business by John Rymer and Dave West, Thursday 8 October 2009, #BTF09

Labels: , , , , , , ,

Thursday 8 October 2009

The contact centre dilemna

Recent studies have highlighted the current dilemma most companies have with contact centres: they are both critical in maximising customer value and considered poor in meeting customer expectations. A YouGov survey conducted in the UK showed that phone and email were by a significant margin the most likely channels used by customers to contact a company for customer service (75% and 70% respectively). Self service comes third with 43%. At the same time, consumers were unhappy with the quality of service provided: 83% were frustrated with the interaction and 60% of agents agreed with them!

This matters enormously because “on average, 40% of customers who suffer through bad experiences stop doing business with the offending company” according to a recent survey published in the Harvard Business Review by Dave Dougherty and Ajay Murthy from Convergys. They go on to define the same reasons for the frustration as identified in the YouGov survey: lack of knowledge of the agent and incapacity to resolve the query on the first call. With increasing access to relevant information online through the internet and social networks, I expect customers to become more and more demanding of the contact centre agent in the future. With the simpler queries being increasingly resolved thorugh self-serivice, interactions will become more complex and more critical and the cost of failure will increase.

So what stops contact centres providing the right customer interaction? The main reason is the proliferation of applications on the desktop. The Corizon Contact Centre survey of 90 Contact Centre Managers shows that agents have to use on average 5 different applications during a call. It is not uncommon to find more than 20 on each desktop. In such an environment, it is proving very difficult to provide to the agent, cost effectively, the right information at the right time and, crucially, the tools that allow them to resolve the query in a fast and efficient way and stop customers needing to call back.

Fixing this problem has proved very difficult as demonstrated by the generations of legacy applications one can find by walking in a contact centres, from ancient green screen applications to the most recent CRM systems. Resolving this issue cost effectively requires a new approach. This is where process mashups come in. They deliver a step change in agent productivity by providing a dynamic UI that guides the process and gives agents the right information at the right time. Agent desktops can be integrated using a step by step approach, fixing the most broken processes first and delivering immediate improvement. The process mashup approach enables the easy creation of fit for purpose desktops, changing the economics of integrating applications for people by extracting “mashable” components from existing systems to deliver an initial solution in days. Once initial process hotspots have been fixed the business can iteratively work towards delivering the ‘perfect’ application for the contact centre agent that can constantly evolve with the business needs, all within a governed and secure IT infrastructure.

Labels: , , , , ,