Financial Services Branches, IP and Contact Centre

One of the blogs forums I read regularly is Finextra, the financial services community forum. The UK financial services sector is such a major user of call centre technology that this is an excellent place to explore the environment that contact centre technology needs to interact with.

A good example is Michael Goldman's post on extending IP convergence to branch. His perspective is not that of the technologist (who have long argued for this) but that of the business person looking at processes and customer experience. I'm pleased to see that he's in favour of more IP convergence, but it's a different perspective from the technology vendors.

It's an area that I'm very interested in as consistent service across channels means (re)-joining the branch and contact centre. I say 're-joining', because historically the branch did manage the telephone channel and it was in response to consumer needs like 24hr service and the difficulty of queuing calls at branch that led to banks to moving the telephony channel from the branch to a centralised service point. This also allowed innovations with 'branchless' and single channel banks, like First Direct in the UK, long before the internet.

Technically, delivering contact centre functionality to branch (where appropriate) is very interesting as it allows a business a lot more flexibility in how it constructs the customer experience. It's an area where Cisco has perhaps a slight advantage from the scale of our IP Telephony deployments, but where Genesys, Nortel and Avaya also have interesting developments. It's probably a subject for further posts as something I'm doing a lot of work on at the moment.

Outbound, an explanation of the technology

I've had a number of requests from my previous post ("Outbound - industry reputation, branding and regulation"), for a better explanation of outbound technology. Some readers had assumed the silent calls were an accident or a technical failure rather than a deliberate business practice.

Unfortunately, the silent calls are actually a business decision based on how you set up your outbound dialler.

A dialler is simply a piece of hardware (or sometimes software) that can place an outbound call from a computer based list (usually a database). It may be obvious, but for the call to work there needs to be a connection at both ends, so the outbound call needs to be answered and then the dialler needs to hand the call to an agent or a pre-recorded message. It's a further small step to appreciate that many numbers dialled will not connect. The number may be engaged, no longer exist, or be for fax machines and when it receives the appropriate tone, the dialler will then not pass the call to an agent. Beyond this signal based intelligence, smarter diallers can recognise answer phones and voicemail

Generally speaking, a dialler can be set up in three modes:

• Preview Dialer - This allows phone agents to view the call information prior to the outbound dialler calling. The agent can decide not to initiate the call and when they do, the number may not be answered.


• Progressive Dialing - This passes the call information to the agent at the same time the number is being dialed by the phone dialer. The agent usually has a few seconds to view the call information, but cannot stop the call process. This method avoids engaged tones and (usually) fax machines, but will not avoid voicemail.


• Predictive Dialling is more sophisticated because the phone dialer automatically calls several numbers and only passes a call to an agent when a person has been contacted. This eliminates busy signals, answering machines, etc.
  

The important thing to note is that the first two result in the same number of outbound calls being made as agents. The third option means that more calls are placed than agents. How many more depends on some fairly complicated algorithms based around predictions of calls that will get an engaged tone or otherwise be unavailable. Ofcourse, this algorithm is not necessarily that accurate and that plus the error level (i.e. the number of calls placed successfully greater than available agents) are what result in silent calls.

The prime industries generating these calls (according to the UK government regulator, Ofcom) are telemarketing, market research, financial services (including debt recovery) and number scanners. The last group, which may not be familiar to readers, is where numbers are dialled at random to check which work, so that a list of 'clean' numbers can be sold to other telemarketers.

The problem these outbound businesses generate is that while cold calling may be unwelcome (and silent calls especially so), the outbound technology has beneficial uses that can get lost in the regulation. Getting a call back, rather than waiting in a queue or being pro-actively told information by a company are all much more positive uses. As so often, the problem isn't the technology itself but the uses is can be put to.

Outbound - industry reputation, branding and regulation

If you mention you work with call centres you can sometimes (!) expect a negative response, and you almost guarantee that this will be because the other person has been on the receiving end of an unwanted automated call.

It's an irritation that I understand and share.

My particular hatred is for the silent calls and the automated message calls. I'll explain the technology in a separate post, but these experiences are because of business decisions made by the calling companies, not because of technology limitations.

In my view, if the company concerned wants to treat customers like that, then I will not give it any business. If you are called by human it at least suggests that there is some effort on the companies behalf to value my business and make an attempt to engage with me. you get none of that with an automated message. Yet brand damage is perhaps not a sufficient deterrent, as most of the companies doing it are debt-consolidators, low-cost telecoms providers and other industries where brand is not a primary concern.

I would argue that companies which keep doing outbound calling on a lowest cost basis have already damaged the reputation of the call centre industry. Consumer pressure has already forced governments to act. In the UK in March last year (Ofcom, the UK communications regulator) increased the fine for silent outbound calls tenfold, to a maximum of £50,000 per instance. It also introduced three other new rules:

• Abandoned call rates must be below three per cent of all calls made in any 24 hour period for each campaign.
• All abandoned calls must carry a short recorded information message identifying the source of the call.
• Calling line identification (CLI) must be included on all outbound calls generated by automated calling systems. CLI allows people to dial 1471 and access the telephone number of the person or organisation calling them.

Already in the UK consumer can ask the TPS (Telephone Preference Service) to add them to a 'do not call' list but it is clear that this is not enough for many consumers. In the end the industry either needs to make consumers happy to risks further regulation. The pity in all of this is that legitimate and useful uses of outbound, such as companies calling to alert their customers of problems with (say) their bank account, or a delayed product shipping, may well be restricted. I

Offshoring and mainland Europe

I have to admit that most of my discussion of offshoring has been around the UK and Ireland experience, in posts such as: "Offshore - why I would go for South Africa over India".

Today I saw some research in TelecomPaper, "Spanish call centre staff numbers to fall 4% in 2007" that looked at the impact offshoring to Latin America was having on the Spanish market. This might seem a startling percentage drop for a single year, but there's some good insight on this at TMCnet giving figures from ACE, the Spanish call centre trade association. Part of the reason for the percentage shift is that compared to the UK or Ireland there are not that many call centres in Spain. In 2006, the number of call centers outside of Spain serving the Spanish market rose from 9 to 20, but more than 225 call centers remained in Spain. As TMC comments, and the UK's experience confirms, Latin America might win business initially on price and cost of labour, but they will have to meet quality expectations if they want to keep that business.

Of course, like India for UK companies, the physical distance between Spain and Latin America creates management difficulties. These aren't technology difficulties (an IP network can let you route calls anywhere in the world) but rather relate to the challenge of flying managers out and keeping expats overseas to manage the offshore operations. Interestingly, here the French have a distinct advantage. With much of North Africa having some French, many French companies have located in Morocco. This has the huge advantage of only being a few hours flight from France and in the same timezone. As long ago as 2005, Africa Investor had identified Morocco has having 55 French call centres with 6,500 employees. Today, numbers are harder to come by, but with major companies like France Telecom and SNCF having Moroccan operations and a dedicated annual trade show (SICCAM), the local industry looks fairly robust.

In other parts of Europe language will probably prevent widespread offshoring for voice (even in eastern Europe, German is not that widely spoken, for example), but I think e-mail may be another story. Without the need for a high standard of both accent and fluency, the non-time sensitive aspect of e-mail may lend itself to offshore management, but that is probably another post.

Contact Centre and the Environment

My attention was caught today by an article on the CCF website ("Contact Centres Going Green"). In many ways this struck me as a long overdue area of comment. Given the increasing focus of the European Union on the Environment, I'm sure that green issues will soon become a high priority for European Contact Centres.

My colleagues who specialise in data centre have been working on green issues for a while and their part of the Cisco website has a whole series of free videos on demand about energy efficient data centres. Similarly Cisco provides some good white papers on the use of technology for energy efficient facilities management. Interestingly, the knowledge for both these areas has come from Cisco's own efforts to reduce carbon footprint and you can find out more on Cisco's environment page.

Of course, datacentre has particular issues with power consumption, heat and hardware disposal that are not always directly applicable to contact centre. The CCF article rather highlighted this, by illustrating that unlike data centre most green initiatives involving the contact centre are enterprise wide initiatives. Even more interestingly, Voice over IP was highlighted as one of the key technologies, with up to 82% of respondents expecting to adopt VoIP and 83% looking to adopt workforce management tools. This highlights one difference between contact centre and data centre, there is a lot of scope in contact centre to become greener by using people more efficiently (through better workforce management and call centre virtualistion through the use of IP Convergence).

This confirms one of my suspicions, namely that for achieving green objectives in contact centre the focus should be on changing business processes and while IT should not be ignored, it should not necessarily be the starting point for change. My starting suggestions would be:

• Virtualise - By moving to an IP contact centre (and using a technology like Cisco's ICM) you can divide work more efficiently between call centres, which means fewer people and workstations.
• Location - It's traditional to locate contact centres in remote areas (cheap land prices). I've blogged previously on why this may not always be ideal ("City Centre Call Centers - A European quirk?"), but specifically on green issues, remote areas tend to require the workforce to drive while a city centre location allows shift workers to take public transport.
• Thin client - Delivering applications through a browser reduces your dependency on PC hardware for performance. Although you may still need to upgrade servers, that is usually a lot greener than replacing hundreds (or even thousands) of PCs.

Of course this is just the start of a list and I'm sure there are many more things to put there. Suggestions welcome.

Why is the agent desktop so often neglected in call centres?

I saw this interesting looking Calabrio Webinar ("The power of the unified desktop") coming up on CRMxchange, and it go me thinking. Why is the agent desktop so often neglected in call centres?

Speech technology, agent skill management, routing and so on are all perhaps seen as more exciting. This may be so from a technology perspective (though I'd argue that SOA has made things a lot more interesting when it comes to delivering applications to the desktop) but in terms of agent performance, the desktop remains crucial.

It seems obvious that agents can no perform no better than the information they are given. Developing this further, as well as data, the ease of navigation, integration of work flow and other aspects of a good agent desktop can make a big difference to both agent efficiency and the customer experience. Given that about 70% of a contact centres operating cost is agent costs. the desktop would seem an easy area for improved business efficiency.

Admittedly, the operating cost of an agent is partly why there is such interest in IVR and automating transactions. Yet the number of call centres with an integrated desktop still surprises me.

Part of the problem is running applications on the agent's PC or other client based framework is not appropriate for many legacy applications (such as 3270 and older) and a browser based portal environment is not always suitable either.

Especially for the smaller contact centre this can represent a real problem. A small contact centre may only have 70 agents, but it may be generating large sums of money in a B2B sales environment and may also have multiple back-end systems to use. Even something as simple as payments and shipping information can require multiple systems if the call centre has been grafted onto an existing business. Yet because the call centre is a small area of the business, it doesn't get the corresponding IT investment to do that level of integration. One of the things I like about the Cisco Contact Centre Express is that there is the option to include a desktop environment and one that integrates with the workforce management tool.

Yet I suspect that many of the problems are not that IT neglects the call centre (though it can do) or because of technology platform issues, but that many senior managers do not understand what their agents do. This lack of understanding could be a lack of focus on the customer experience or no real knowledge of the detail of the agent workflow (e.g. just how much does the agent have to do manually, such as data capture?) . The result, though is that simple areas like the agent desktop or virtualising telephony are neglected and more complex solutions are imposed that deliver less.

SOA - bringing CRM, telephony and business together? part 2, voice portals

It was a big subject that I tried to cover last Friday ("SOA - bringing CRM, telephony and business together? part 1") and of necessity I only covered it a high level, despite the length of post!

The main point to take away was that SOA can (in theory) be done at a number of levels in the contact centre. In theory, starting at the back end with SOA process choreography or data model rationalisation is as valid as starting at the front. In practice, though, starting at the front with either the agent desktop or the voice portal is much easier.

There's a number of reasons for this, but the main ones are risk (the voice portal technology is relatively proven compared to some other SOA approaches in call centre), scope (self-service can start with a relatively finite number of services) and benefits (even small increases in automation can lead to big cost savings).

The portal works for the voice channel in the same way as a traditional browser based portal provides a view for the internet channel. Obviously information has to be presented differently (sequentially for voice compared to concurrently for web) but the concepts are the same. In this environment it makes a great deal of sense to componentise services (such as identity and verification) and then make the same service available to web and voice channel. The benefits of a common ID&V service are not just efficiency from the re-use of code, but also a consistent customer experience, a consistent process regardless of channel and a better security system with easier management.

In my view Cisco Voice Portal (CVP) is the leader here with Genesys and Avaya not too far behind. It's not a view Gartner share but I feel they didn't fully appreciate the value of some of the deployment options CVP provides. A real strength of CVP is that it offers both centralised and distributed deployment options, allowing a business to provide self-service locally (in a branch or retail outlet) and thus queue calls at the edge of network before routing them anywhere in the enterprise that they want. Alternatively, there is option to deploy CVP in the datacentre or at the call centre. Video is another area of CVP strength. While video calls may be some way off for the mass market, a CVP based video call centre called Significan't is up and running here in the UK so that deaf and other sign-language users can access government services via a video based call centre.

I also find the Cisco tooling (which was Audium) very good for service creation and suited to SOA development as it is based on the Eclipse platform. This is not unique to Cisco (Avaya Voice Portal also uses Eclipse as its development environment) but the combination of CVP deployment options and open standards works well for the clients I've worked with.

In short, if you are going to bring SOA into the contact centre environment it makes a lot of sense to start in area of discrete functionality that can be re-used in other channels and to do so with a product which gives you the most deployment options.