UK Agent Attrition

An interesting article on silicon.com about UK contact centre agent attrition increasing for the 5th year in a row.

I actually wrote a post on the survey (and HSBC's response to agent attrition) last week "The contact centre agent experience - First Direct", but the silcon.com article has a good quote from the analysts Contact Babel.

"Steve Morrell, principal analyst at ContactBabel, said in the report: "The lack of growth in agents' salaries is certainly a major factor in producing high levels of attrition. Businesses should be working to move low-value interactions onto web and phone self-service channels, and to use the savings created to pay higher salaries to their agents. This will attract and retain high-quality staff, a move which will immediately and permanently benefit both the business and the customer base.""

All very much thoughts this blog has been discussing in terms of moving from call centre to contact centre and making contact centre part of a multi-channel strategy. I'm not sure, though, this will necesaarily translate into higher UK salaries as I see high-skill offshore like South Africa being a significant factor in the market.

I've always stressed the importance of moving to converged IP because it makes managing multi-media in a multi-channel environment so much easier. Silcon.com aren't making the connection as explictly as I do, but they do produce the quite interesting statistic that:

"...pure IP infrastructure will be commonplace in most UK call centres within the next two years, with 41 per cent of respondents saying their operations will be fully IP-architected by the end of 2009.
Currently only 17 per cent of call centres have an full IP infrastructure, with a further 28 per cent using a hybrid IP/TDM network."

I suspect that one of the other big drivers to IP (besides multi-media management) is the ability to virtualise across borders. I know this is a message Cisco has pushed strongly, and I think it's the right messge for many enterprises. This is especially true in Europe where the continuing integration of the EU offers many opportunities to distribute work more efficently and more widely. Agent jobs (especially in areas like e-mail, where language skills and accent are not so crucial) will go to areas where the work represents a higher salary and on-shore will be left with higher skill, higher value agents.

SOA - bringing CRM, telephony and business together? part 1

It's a Friday, so perhaps time for thinking a bit further ahead.

One of the interesting things about the contact centre (and before that call centre) is that it has three areas that have rarely had more than a passing acquaintance with each other. By areas I mean the the three core functions of telephony (so that you can deal with customers remotely), CRM (which here means the delivery of data and interactions, rather than a specific packaged application) and business process (which is what the contact centre is built to handle and why customers call).

Traditionally self service, or IVR (Interactive Voice Response unit) has high-lighted this separation. It is probably more familiar as "...press 1for..., press 2 for...., etc..." and is the source of enormous customer frustration. This is mostly because it has come from the telephony world (which focused on scalability, responsiveness, routing, etc... ) rather than the IT or business world. As a result traditional IVR tends to be good at managing telephony but rather poor at integrating with CRM, business processes or being customer specific. Similar disconnects between needs and functionality can be highlighted in CRM or all those manual work-arounds in business processes.

It's into this world that SOA enters.

For those not familiar with IT trend, SOA stands for "Service Orientated Architecture". The idea is to build IT systems out of small components of functionality ("services") that can be put together by business processes to perform tasks. These services are not applications as we know them (though they could be) but are much smaller units of business functionality, such "calculate mortgage interest rate" or "identify & verify customer" and can be used by many different functions. These components can be part of existing applications (which increases efficiency from re-use of existing IT) or even come from third parties outside of the enterprise.

In call centre SOA has arrived first (and perhaps surprisingly) in the self-service environment of voice portal. A key factor was that the development of IP convergence meant that voice could be manged in the IP protocol world of the IT industry. A voice portal is not the the traditional IVR unit and from the proprietary telephony world. Instead is more part of IT and can be thought of an internet portal that presents information by voice rather than by speech. The SOA part of this is that if you have services for any channel (branch desktop, web, etc...) then you can present them in the voice portal. So for example, if you greet customers with a personal message when they log into your web site (e.g. "Welcome Mr. Smith, your most recent order shipped this morning"), the voice portal lets you do the same for the telephone channel. It also lets you present information both on a personalised basis and dynamically, which can dramatically increase your customers' use of self-service.

What voice portals are forcing is a coming together of telephony and CRM and SOA is the most sensible architecture for doing so in a way that aligns with business processes. Cisco, Avaya and Genesys are the clear leaders for voice portal, but in my next post I'll try and explain how things like tooling and service management make for big differences between what the vendors offer amd are critical to being successful with an SOA approach.

The contact centre agent experience - First Direct

There's a good overview on the BBC News website of what it's like to be an call centre agent at First Direct (the telephone and internet banking arm of HSBC). What's also interesting is that it shows how First Direct deals with the problem of agent attrition and under-motivated staff. All very relevant when at the start of the month the UK Contact Centre Operational Review found that agent attrition was up for the 5th consecutive year and now stands at an average of 32%. The idea that you loose a third of your staff in the year in which you hire them would not be sustainable in any other industry. If your brand depends on telephone service (and the staff that provide it) then it is not affordable and the BBC article shows what can be done to retain good staff and hence maintain the quality of customer experience.

I find First Direct very interesting as an example of what can be done with the contact centre. First Direct actually dates back to the pre-interent age and were set up by HSBC as experiment in telephone only banking. It was extremely successfully and has since then led a number of developments such as migrating transactions to internet and introducing fee based retail banking to the UK.

The thing that most struck me about the work environment was not the technology, but the focus on the agents. Once you've migrated simple transactions to the Internet channel, the phone becomes predominantly the channel for complex interactions. I've discussed this a little before why the people matter so much when offshoring ("Offshore - why I would go for South Africa over India") and when looking at e-mail ("e-mail in contact centres, process driven or response driven?" & more fully in "Call Centre to Contact Centre... or why worry about e-mail when the phone's ringing?"), but I think it's a topic I'll need to return to in more detail.

In the end, technology is critical for handling basic interactions and being efficient, but it's how the people in the contact centre use that technology that will determine the customer experience and the perception of the brand.

Speech market share - the role of non-European langauges

Some very good comments on the "Speech Market Share" posting I did in October, especially from Martin on Telisma and why they might have achieved such a large growth in European marketshare.

I was particularly interested to see that Telisma support Hindi and are working on another 18 Indian languages (including Indo-Arayan and Dravidian). This could be very interesting for a couple of reasons.

Firstly it's the established capability developed for the domestic market that I see as crucial for successful offshore operations. This is partly why in previous posts ("Offshore - why I would go for South Africa over India"), I've tended to rate South Africa ahead of India as a location for off-shore contact centres. If India can build a domestic capability then it's ability to take on offshore work will also improve. Also, and perhaps further off, the skills developed by Indian software developers working on speech for the Indian market could be very helpful for South Africa with its eleven official languages develop speech for its domestic market.

The second, and perhaps more unexpected use of speech is to serve migrants in Europe. These groups can be significant demographic segments and well worth offering higher service for. For example HSBC launched a UK Islamic bank in 2003 and it is easy to envisage it offering multi-lingual support for its customers. Perhaps further ahead is Canada. When I worked there I was expecting English and French telephone banking, but was surprised that Scotiabank telephone banking was available in not just English and French (as you would expect), but also Cantonese and Mandarin. It turned out that these languges gave them a competitive edge for winning business in the SMB market segment.

Given my experience of working with Canadian call centres, I am surprised that UK banks have not moved further with multiple language support. This may be a cost issue (given how cost concius the UK is) and perhaps the ability to offer automated support will change that.

Workforce Management - Part 2 Vendor Selection

I was interested to see that yesterday's posting got quite a bit of interest. One question that seemed to crop up regularly was "when should you buy a Workforce Management solution from your CTI vendor ...and when shouldn't you?"

The short (and unsatisfactory answer) is that "it depends on your circumstances". That said, here are some suggestions as to what you should consider.

• How strategic is your CTI supplier? - If you route calls on a relatively simple basis then you probably have a lot of freedom in your choice of WFM provider. If your CTI supplier is a critical part of your business (e.g. many, many skill groups, all overlapping spread across multiple sites) then it makes sense to take a WFM solution where you can export your skill categories straight from CTI into WFM.
• Are you wanting to monitor workforce skills, education and interaction quality? - If so, then you are probably looking to go the level beyond Workforce Management, which is referred to as Workforce Optimisation. This is best from a specialist provider, (like IEX or Calabrio) rather than your CTI provider.
• Where are you deploying your WFM solution? It may seem obvious, but unless it is localised, it will be difficult to use for non-English speakers. So for example in Europe, Calabrio (which supports English, French, Spanish and Portuguese, and should support German and Dutch by year-end 2007), Holy-Dis (very strong in the French market) and Teleopti (very strong in the Nordics) all have to be considered ahead of some of the large, but not localised US providers.

I'm sure there's plenty of other things to consider (all suggestions welcome!) but that should get you started....

Workforce Management - is it only for high end call centres?

Looking back at my posts I notice that I've written a lot on speech and self-service, a bit on CRM and a bit on e-mail. I've written nothing, though, on one of the major contact centre application areas of workforce management.

For those not familiar with it, workforce management is primarily about managing staffing levels so that they handle call volumes efficiently. What might appear a straightforward calculation becomes much more complex when call volumes peak and trough dramatically, calls can only be handled by agents with the right skills (in financial services, for example, this is a regulatory requirement) and the workforce (with these multiple skill types) might be spread across several sites and perhaps timezones.

For many call centres (especially the smaller ones that are the majority in Europe) the single largest workforce management tool is Microsoft Excel. This has great advantages in being familiar to most managers, relatively easy to use and a simple tool.

Once you get above 100 seats, or have a high degree of complexity in what your agents are skilled to do, I feel that workforce management tools become essential. As an example, a small stockbroker might have 100 seats. On Monday they need 70 or so of those seats to have trading skills to handle the call volume for market opening. Tuesday needs only 40 seats to have trading skills, but 50 seats need to be have qualified financial planning skills and 30 need to have trade settlement skills. Once you have multiple skilled agents (as in this example) it is good for the business, but it can quickly become very difficult to manage, especially where holiday planning and exceptional events need to be built into shift patterns.

The big players that I've worked with are Calabrio Work Force Management , Genesys Work Force Management and Aspect Workforce Management. What I find quite interesting is that the Aspect and Genesys offerings are traditionally pitched at the very high-end call centre, e.g. not the stockbroker in my example where a few agents handle a very complex possible set of interactions, but more the retail banking model where thousands of agents handle a smaller, though still complex set of interactions. Admittedly this high-end is Aspect and Genesys' heritage and where demand for management tools came from but I'm not sure it's the future, especially once multi-media traffic is added to the traditional voice traffic mix. I suspect that as Genesys has always positioned itself for highly complex call routing in very large centres there is a natural market for Genesys Work Force Management plugged into Genesys CTI. The problem for most European centres is that they are not large enough to need the Genesys scale of solution, but are too complex for a simple switch or a PBX and ACD.

One of the interesting things I've found in my work is that small call centres (especially in the B2B environment) can be highly complex and handling very large value transactions and are not just small call centres for small companies. As an example, I worked with a very large cosmetics firm who managed all their European wholesalers from a 70 seat call centre. That (small) call centre ran over 100 product lines, several $B of business a year and did so in six languages. In terms of European call centres, language is obviously a key agent skill and scheduling the right mix of languages is crucial before any other skill can be considered. It might have been a small call centre but it was highly complex.

One of the things I like about the Cisco Contact Centre Express is that it's restrictions are mostly around scale rather than function and there is the option to buy with built in Workforce Management. As an OEM from Calabrio it has a nice look and feel for the user and avoids a lot of the integration issues that happen for any call centre solution when a 3rd party workforce management application is used. There's a good overview here with some screen shots. Importantly, it addresses a call centre market segment (small but complex) that has traditionally been required to either take an over-engineered solution designed for larger customers or a simpler solution for small users that didn't meet their needs.

Is cost a contact centre issue or a symptom?

One recurrent theme in contact centres is that of cost.

Mostly, the complaint is that contact centres cost too much. I would argue that this is because the cost is calculated for the contact cnetre operations alone and with little reference to the value that the contact centre might bring for the wider enterprise. I've touched on the idea before that cost is not only operational metric to consider (mostly in the post "Offshore - why I would go for South Africa over India") and that doing the same function at lower cost is not going to solve many of the bigger and more expensive problems. I would argue that cost is only a symptom of problems in the wider enterprise, a symptom that is shown in call centre.

The CMP magazine is taking another approach and trying to change perceptions by having an award ceremony to recognise the top 50 call centers in the UK. The difference here is that the adjudicators will be customers rather than industry peers, and so recognition will be based on absolute customer service outcomes with no reference to the cost of achieving those outcomes.

This seems worthy, in terms of understanding and demonstrating how much contact centres can achieve. It also go me thinking, though, about something I've seen for a while. What often shows up as cost in the contact centre is actually a problem in a completely different part of the organisation. For example, it's obvious that badly made products lead to returns and complaints. The call centre of the organisation making those products would therefore have to be quite large, yet the problem isn't the call centres' ability to handle calls efficiently, rather it is the products that generated those calls. Similarly in financial services if a product is too complex for its target market that will generate extra service traffic. The problem there is not the cost of handling calls, but of insufficiently well designed products for a given market.

The cost conscious retort might still be "so what?". The products were sold, the revenue was booked, why worry about handling the calls on any level other than cost? Two good reasons come to mind immediately, but I'm sure readers will have many more. The first is that the cost of handling customer complaints well is far less than the cost of reputational risk and brand damage. Compare call centre costs with the marketing costs to rebuild a troubled brand and the sums are heavily in favour of good customer service. The second point is more subtle, but is about the ability of the organisation to adapt. If an organisation can understand what customers think of it in real time (that would be the contact centre that does that) then it can adapt and change in response. The cost savings from withdrawing a product or altering an internal process that is causing problems can be far larger than the cost of the call centre.

In short the value of the contact centre is in its ability to make the organisation responsive. That and the brand value it adds are key yardsticks against which to measure its cost.