Customer relationship automation: a viable business objective or an oxymoron?

In business circles, the acronym of the moment would have to be CRM.

Corporations the world over are spending tens (or often hundreds) of millions of dollars apiece in the name of these three unassuming little letters.

Of course, those letters stand for customer relationship management — a term that basically refers to technology and business systems that together automate the management of customer relationships.

The promise held out by CRM is an appealing one, but before you grab your chequebook and attempt to collar a software vendor, it’s worth pausing for a moment to consider whether or not it’s actually possible to automate the management of customer relationships.

Or are, as many media reports suggest, these countless millions of dollars being squandered in a fruitless quest for an impossible dream?

If customer relationship automation is, in fact, a viable business objective, it would be nice to have a road map to follow to maximise one’s chances of arriving at the ultimate destination unscathed.

This article examines the viability of customer relationship automation, and then presents such a road map.

Why automate relationship management?

The recent interest in the automation of customer relationships can be attributed to two factors:

1. Market pressure for continual improvement in organisational efficiency.

2. The availability of technologies (particularly in the areas of telecommunications and the Internet) that at least hold out the promise of the automation of relationship management processes.

Market pressures make automation a priority

Automation increases the throughput of a process.  Of course, throughput is good because it increases the utilisation of infrastructure, drives down costs, and inflates profits.

Manufacturing processes were the first organisational processes to be candidates for automation.  This is because, at the birth of the modern corporation, manufacturing was the bottleneck in organisational throughput.  (Demand exceeded supply.)

The best known example of early manufacturing process automation is Henry Ford’s employment of the assembly line in the Model T manufacturing process.

For most of the second half of the last century, the attention of organisations was on the automation of manufacturing processes.

As a result, we are now in the position where most organisations have removed the bottlenecks from their manufacturing processes.  Accordingly (at least from a manufacturing perspective), supply typically exceeds demand.

Because manufacturing is no longer a bottleneck, improvements in efficiency in this area will have little impact on organisational efficiency as a whole.

Just think, if you wanted to improve the efficiency of a machine, you’d first apply oil to the parts that squeaked.  Adding oil to well-oiled parts would have little effect.

In order to extract further efficiency improvements, managers have had to focus on the new bottleneck in the organisation.  This new bottleneck is the front-end of the organisation — or the interface between the organisation and its customers.

New technologies promise relationship automation

In manufacturing terms, the two factors that make the greatest contribution to efficiency are throughput and quality.

In the management of customer relationships, efficiency is a function of throughput and intimacy.

It’s worth noting that, in both cases, these factors appear to be mutually exclusive.  One would assume that quality (or intimacy) would decline as throughput increases.

In a manufacturing context, the reality is that the technologies that enable economies of scale, also enable economies of quality.  (Both throughput and quality benefit from the elimination of variation — the basis of mass production.)

But what about throughput and intimacy?  Are they not mutually exclusive?

Well, until recently they were.  If we look at retailers, a high volume retailer has always provided less customer intimacy than a low volume one.

Let’s compare a corner store with a traditional department store.

The former provides both geographical proximity and familiarity (the retailer remembers your name and your preferences).

But to access the lower prices offered by the latter, you have to drive further, and sacrifice the benefits of familiarity.

Recently, however, technology has enabled these two key components of intimacy to be scaled.

Relational databases now allow organisations to remember names and preferences (familiarity), and telecommunications and the Internet provide customers with access to organisations from the comfort of their own homes (proximity).

Of course, the best example of the scalability of intimacy is Amazon.com (perhaps the world’s most mature practitioner of CRM).

You can shop at Amazon.com from anywhere in the world where you can find a telephone jack and an Internet Service Provider.  And Amazon.com remembers your name, your preferences and your purchasing history.  (Amazon.com can also use this information to e-mail you — surprisingly accurate — recommendations of books that may be of interest.)

The value of personal interaction

It’s one thing to use technology to deliver (at least some of) the components of customer intimacy, but it’s another to expect customers to maintain an ongoing relationship with a computer.

Or is it?

The key question here is: how critical is personal interaction to the maintenance of a customer relationship?

If it’s absolutely critical, the requirement for personal interaction is always going to be a constraint on the automation of relationship management (a bottleneck).  Accordingly, the key promise of CRM may not be deliverable.

The answer is that the requirement for personal interaction is a situational thing.

The requirement for personal interaction varies from customer to customer and from touchpoint to touchpoint.  (A touchpoint is a point in the customer relationship management process where a customer comes in contact with your organisation.)

In some situations, personal interaction can obviously add value.  And in others, as Amazon.com demonstrates, it may be unnecessary.

But in some situations, personal interaction may actually subtract value.  This may be because your customer would prefer the anonymity of an impersonal interaction (would you rather have a person or a computer advise you of your bank account balance?).  Or it may be that personal interaction unnecessarily complicates an otherwise simple transaction.

Before you automate

The news that it is, in fact, possible to automate at least some of the touchpoints in your customer relationship management process is potentially very exciting.

For many organisations, the customer relationship management process is dependent upon significant amounts of infrastructure.  This certainly applies to organisations with retail distribution channels.

Again, this point is well illustrated by Amazon.com.  In 1998, Amazon.com enjoyed an increase in revenues of approximately $300 million.  The additional infrastructure that was required to support this increase in sales cost Amazon.com just $100 million.

Now if Barnes & Noble, (Amazon.com’s principal offline competitor) had increased its sales by $300 million over the same period, it would have had to open 100 new stores.  This would have required an investment in infrastructure of approximately $1 billion.  (In other words, Amazon.com’s customer relationship management process appears to be ten times more scalable than that of Barnes & Noble.)

But, as is often the case with business process reengineering, the automation of a customer relationship management process is rarely as simple as it looks.

The problems we most often see relate to organisations thinking locally, but acting globally — instead of the reverse.

By thinking locally, we mean that organisations suboptimise their customer relationship management processes.

And by acting globally we mean that they attempt to automate everything or, boil the ocean.

Boiling the ocean

We have seen many troubled CRM implementations that amount to a valiant attempt to do that which is quite clearly impossible.

While, technically, it may appear feasible to simultaneously automate a sales force, build a call centre and replace a number of legacy customer databases with a unified, e-commerce-enabled CRM solution, organisational inertia will obviously doom this project to failure.

It is important to begin any reengineering project with a global view of the desired end state.  However the scarcity of resources (particularly management attention) and organisational inertia mean that it is better to focus on those initiatives that are likely to yield the greatest (and quickest) results.

As suggested previously, these initiatives should be identified by looking for the bottlenecks in your customer relationship management process.  (Remember, if it doesn’t squeak, don’t waste oil on it!)

Suboptimisation

Suboptimisation can occur when you optimise a component of a process without regard for the impact of your changes on the process as a whole.

This is a danger for customer relationship automation projects because it is so easy to underestimate the importance of personal interaction.  (Your interactive voice response system may reduce your transaction costs, but is it disenfranchising high value customers?)

Ask yourself, if the technology is readily available, should a recruitment firm automatically create a web-based clearing house for its candidates (as Monster.com has done)?

To a technologist, the obvious answer is yes.  From a technical perspective, it makes sense to automate the matching process and to remove the barriers between employers and candidates.

However, if this recruitment firm services an executive clientele, this automation is almost certainly not in its best interests.  An executive clientele will look for a recruitment firm that deliberately maintains barriers between employers and candidates.  In other words, the inefficiency of this organisation’s matching process is actually a design feature.

To avoid suboptimisation, it is important to:

1. Maintain a clear understanding of the objective of the particular process on which you’re working.

2. Examine the impact of proposed changes on a range of customer profiles over a range of possible situations.

Automating your customer relationship management process

The following four-step process enables you to plan the automation of your customer relationship management process, without falling victim to either of these problems.

The first two steps focus on the process as a whole, and the last two, on the touchpoints that require automation.  (It should go without saying that the bottleneck in a customer relationship management process is almost always a touchpoint.)

1. Define your objective

2. Identify bottlenecks

3. Identify touchpoints

4. Optimise touchpoints

Define your objective

It's sometimes easy to lose sight of exactly why a process exists.

Even if the answer to this question is obvious, it should still be top of mind for the duration of the planning process.

A failure to do this can easily result in process suboptimisation.

Your definition of the process objective should be broad enough to discourage this suboptimisation, though not so broad as to be meaningless.

For example, if you were considering the automation of a help desk, you might define the object of this process as to provide remote resolution of customer problems, thus avoiding their escalation to field technicians.  If you were to define the objective as to provide telephone resolution of customer problems, you may overlook the fact that it is possible to migrate a significant percentage of your help-desk functionality to your Website, as Microsoft has done with its Knowledge Base.

Identify bottlenecks

The easiest way to identify bottlenecks is to flowchart the touchpoints in your existing process.  Once you have done this, ask yourself which touchpoints are constraining the throughput of the process.  If the answer is still not obvious, ask yourself, if the throughput of this process were to double over the next six months, which touchpoints will not scale?

You will often discover that the bottleneck is not where you might initially have assumed it would be.

We often encounter sales managers who are considering using Sales Force Automation software to improve the productivity of their sales forces.  In this situation, we question whether or not the sales force is, in fact, the constraint in the sales process.

More often than not, we discover that salespeople are operating nowhere near their peak capacity — accordingly, they are not a bottleneck.  The real bottleneck is the process that generates sales opportunities.  (Hence the importance of our Relationship-centric Marketing process.)

Map touchpoints

Once you have identified the touchpoint (or touchpoints) you wish to optimise, you can plot it on the Touchpoint Optimisation Matrix below.

This matrix enables you to specify appropriate communication channels types by mapping your touchpoint into one of four quadrants.

It then enables you to narrow your selection by considering the trade-off between the scalability and the effectiveness of different channels.

[click to enlarge]

To plot your touchpoint on this matrix, simply ask yourself the following questions:

1. Should this contact be initiated by your organisation (push) or by your customer (pull)?

2. Is personal interaction of neutral or negative value to this touchpoint (transactional)?  Or does it add value (inter-personal)?

Optimise touchpoints

Once you have determined the quadrant to which this touchpoint belongs, you will be faced with a choice of communication channels.  Typically, there is an inverse relationship between the scalability of a channel and its effectiveness.

For example, an outbound telephone call is more scalable (less expensive) than a face-to-face customer visit from a salesperson.  However, if the touchpoint is the point at which the decision is made on a $100,000 purchase, you will probably be prepared to pay a significant premium for the increased effectiveness of the more expensive channel.

Accordingly, your choice of channels should take into account the significance of the touchpoint — in isolation, as well as in the context of the ongoing relationship with the individual customer (or customer profile).

Now that you have your choice of appropriate communication channels, it’s important to give careful consideration to the design of the touchpoints (and the environment surrounding these touchpoints).

This is particularly important when you’re dealing with inbound communications, where the choice of channel is generally your customers’.

To explore this point, let’s return to our help-desk example.  We’ll assume that this help-desk has three communication channels:

1. A web-based knowledge base (self-help)

2. A web- and e-mail-based job system (asynchronous assistance)

3. Telephone support (synchronous assistance)

Customers' initial inclination will be to use the telephone support service for all support problems.  Because this channel is difficult to scale, this organisation’s challenge will be to migrate customers to the more scalable channels.

While an obvious solution to this problem is simply to charge for telephone support, this heavy-handed approach may well result in the sub-optimisation of the customer relationship management process.  (Charging penalises all users — including those for whom this is the most appropriate channel.)

A more realistic approach is to design the touchpoints (and the environment surrounding them) so that the preferred channels are perceived to be the path of least resistance.

The following initiatives might be appropriate:

• Advertise an e-mail and a Web address rather than a support telephone number.

• If customers have access to computers, arrange for support staff to train them in the use of the knowledge base at the same time they are solving their problems.

• Have support staff enter all jobs into the e-mail-based job system — so that customers receive automated e-mail updates.  These updates, can promote both the knowledge base and the job system.

• Arrange for your job system to automatically query your knowledge base when jobs are entered and present results as possible solutions, prior to logging the job.

This example should illustrate that the design of the touchpoints is just as important as the technology used to automate the touchpoint (if not more so).

On the subject of technology, it is a mistake to assume that those touchpoints where personal interaction is necessary cannot be scaled.

The reality is that, these touchpoints often benefit from the same technology that is used to automate the purely transactional touchpoints.  The only difference is that this technology is used to empower the staff responsible for direct customer contact, rather than being used by the customers themselves.

Not an objective, an obsession!

This article began by asking the question, is customer relationship automation a viable business objective or an oxymoron?

It should now be clear that the words automation and relationship are not mutually exclusive.  (In fact, as the banks and Amazon.com have demonstrated, we sometimes prefer those relationships that don’t require personal interaction!)

As far as the first part of the question is concerned, the automation of customer relationship management should be more than a viable business objective — it should be an obsession.

In an environment where supply typically exceeds demand, the future of your organisation depends on continual innovation and seamless integration with your customers.

Your manufacturing department has just passed you the process improvement baton.

Go to it!

This article was researched and written in conjunction with Mike Smith at Simbient Solutions.  Simbient Solutions specialises in integrating Web interfaces with back end solutions (i.e. CRM and ERP systems).  Simbient’s clients include E*TRADE, Fairfax, General Electric and Shopfast.

[Technology Index]  [top]