We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Competing on (Customer) Decisions

Originally published April 22, 2008

Customers’ expectations are on the rise, demographics are changing, business is more complex and yet the pressure to reduce costs continues unabated. Organizations must think about the customer experience to retain and grow existing customers. One key, clearly, is improving the service that can be delivered by front-line staff such as those in stores or call centers. Another is to ensure that every interaction – not just with staff but with your emails, letters or websites – contributes to the customer experience in a positive way. Personalization – making customers feel that you know them and are responsive to their needs – and the loyalty of customers to your organization are also critical. Yet you cannot do everything for everyone, so the resources to be applied must be applied where they make the most difference. And that might be different tomorrow from how it is today.

Competing by focusing on customer treatment decisions should be a core element of a customer-centric strategy. To compete on customer decisions, you must adopt the core principles of enterprise decision management or EDM.

It is important not to focus purely on the opportunities for replacing people with software, something many customer service departments have done in the past. The opportunity for decision automation to empower front-line staff to deliver better customer service, while also improving the experience customers have when interacting with systems, is huge. The ability to replace a complex, potentially stressful decision that must be completed in a short time window, such as deciding who is a good customer, and instead make an automated decision can result in dramatic improvements. Getting really personalized can boost response rates while making your customers “feel the love.” All the customer treatment decisions you make (through staff, through the website, through the options presented to a customer) should reflect what you know about your customer. You must, however, know what your objective is. Perhaps it is to reduce cost with a minimum of impact on customer service. Perhaps it is to improve the customer service of your best customers. Maybe it is about improving customer retention.

EDM can only help if you are clear as to your business objectives – it will only be successful if you define what “success” is before you start. EDM is a powerful tool for improving the customer experience.

One of the most important metrics in call centers is first call resolution.

“When you call a toll-free number to order clothing, get help with your computer, or book a plane flight, you want the call to be quick and conclusive. You want the operators or agents who handle your call to accept your order, answer your question or make your reservation before you hang up. In the contact center world, that's known as first call resolution, and the companies that you call want it too. If a call takes too long or leaves your issue unresolved, the company's costs go up and it may lose you as a customer.”1

There are a number of ways to use EDM to deliver on this promise.

  • The first, and most obvious use of EDM, is to manage approvals. Many calls are not resolved first time because the person who answers the phone is not authorized to do something – refund something or expedite it. Yet, all too often the data required to make the decision is available at that point. The status of the order, the customer and pricing rules are all known. The automation of approval decisions using business rules is straightforward and allows the person who answers the phone to immediately approve anything that passes those rules. A rules-based approach also helps with complex dialogs as it can walk the call center representative through the various steps, in the right order and so deliver the right answer.

  • One of the things that can make it hard for call center representatives to close a call quickly is their lack of information about risk. It might be the retention risk of a customer, the credit risk of a prospect or even the risk that a customer will buy without a discount the representative can choose to offer. Displaying information to help with this is one approach, but the reality is that risk calculations can be complex and so representatives are hard-pushed to make them quickly and accurately while a customer is on the phone. Automating risk predictions, using predictive analytic models, and embedding those predictions into automated decision services can thus materially improve the ability of the customer representative to close the right “deal” on the first call.

  • Customer segmentation so that the right script can be used or so that the call can be routed to the right representative can also reduce call transfers. Effective segmentation, linked to the behavior of customers and their likely needs, can be linked to rules to make sure the call center handles each call as efficiently as possible. Combined with the approval and risk automation, and with pre-targeted offers and answers, this can make first call resolution much more practical.

  • Customer-focused organizations are increasingly trying to ensure that their call centers deliver the best next action – what is the treatment that will most develop, extend and improve this specific customer relationship. Automating this decision and making sure it is embedded into the call center process takes EDM and can both improve first call resolution and improve the customer relationship.

  • By automating these decisions, you also get the advantage that third parties acting for you can resolve calls as effectively as you can. Companies with multiple layers of partners and channels cannot assume that those resolving problems work directly for the company. They must support a complex ecosystem of responders.

  • Automating these decisions using a technology like business rules that allows you to rapidly and accurately change the logic used to respond to calls also delivers increased business agility. If your call resolution rules are handled by training and policy manuals, then changing them is likely to be slow and expensive. If the rules are externalized from your system in a way that allows business users to manage them, then they can change as soon as the business sees the need for change.

In addition to these direct improvements, automation of the decisions using EDM can also improve the quality of interaction. After all, the person answering the phone no longer has to worry about the technicalities but can instead focus on the human interaction. This pays off in terms of customers feeling more connected to and more valued by the representatives.

There are many views of loyalty, but one of the most prevalent and compelling is the idea that large, multinational, multichannel organizations must generate the same feelings as local stores and local branches do – the sense that the people with whom you deal know you, value your business and tailor their response to you. Richard Hackathorn wrote a great article on the same topic called "Forward to the Past" in which he said, “As we fast forward into our global economy, realize that we often direct our technology to achieve the common things of the past.” One way to do this is to try to give your customers personal contacts, named staff, who become familiar faces and voices and, by virtue of dealing with the same customers over and over again, are knowledgeable about their needs. This might work, and EDM can certainly help you with it, but what happens if that employee leaves, especially if that employee leaves and joins a competitor? Will you find that your customer is loyal not to your brand or your products but to your employee? Instead of trying to build loyalty to individuals, you need to create loyalty to the store or bank or brand. To do this, to have an ongoing process of building and sustaining loyalty without transferring it to a specific employee, requires EDM without a shadow of a doubt. Think about what makes someone loyal to a small store or branch, or to a specific employee. Perhaps that person got something done for them or perhaps the local store made them a really compelling offer based on what they purchased in the past. Perhaps that person knew the customer well and so dealt with them appropriately. Maybe the rewards being offered were just what the customer valued. Certainly, there was not a lot of referring to management or generic scripting of responses going on. What these situations have in common is a focus on the decisions that customers want made (pricing, refunds, shipping prices and times, offers, loyalty rewards) as well as those you want to make about them (cross-sell, up-sell, retention). Because EDM focuses on and automates and improves these decisions, EDM can deliver on organizational loyalty. Think about it:

  • By automating decisions, you can ensure that customers are not referred around the organization unnecessarily. The first person they speak to can act immediately because the system can deliver the answer without having to get a supervisor on the line. They won't become loyal to the individual who can "work the system" if the system empowers everyone.

  • By embedding best practices rules in a decision, every call center representative can respond like the most successful. The difference between the best and the worst customer service representative will be smaller, making it less likely that one particular representative will become a customer's favorite.

  • Using predictive analytics and statistically significant segmentation rules, customers can be treated more appropriately and in a more targeted fashion – you can align your treatment with their behavior. Their wants and likes can be more accurately included in decisions made about how to treat them. Automation of this means that everyone can see the trends in the behavior and value of a customer and know what action to take. Trying to use dashboards and reports means that only the most analytically sophisticated representatives can do this, risking the transference of loyalty to them not the company.

  • Automation means that the same decision can be delivered through any channel – at the store, on the web, in the call center. Consistency of treatment builds loyalty (assuming the treatment is not obnoxious) and knowing that you can get the same treatment from anyone keeps that loyalty linked to the company, not to an individual. Instead of always coming to the branch to see a particular person, they will know that they can use any channel and still get great service.

  • By targeting loyalty program offers, you can use those loyalty program dollars more effectively – not just to reward but also to change behavior. Making a proactive offer to those who are on the cusp of becoming loyal customers, for instance, might be more effective than rewarding those who become loyal.

An EDM approach is not going to help with friendliness or surprising customers from time to time, but I do believe that higher volume, more self-serve oriented businesses must adopt this approach to really deliver customer loyalty. Using business rules management systems to implement a decisioning backbone for consistency while retaining the ability to respond to changes and injecting insight using predictive analytics lets you recreate the corner store feel and benefits while still delivering the transaction throughput and response times a modern business often needs.

Personalizing your business is one of those things that never really seems to go out of style and that remains one of the most powerful ways to use EDM. Making a system respond in a reasonable, yet personal way to a consumer or other customer is widely, and correctly, seen as a way to bring better service to bear in a scalable way. Personalization has some synonyms too like "behavioral targeting,” “precision marketing” and “micro segmentation." Personalization means more than just scripting responses. It means providing the best response in each situation for that specific customer. Technology has been evolving steadily to help in personalization. Content management systems and more dynamic websites make it easier to display personalized content. CRM and call center applications support more dynamic scripting and displays and work faster so that a call center representative can use them while in a conversation. Analysis of unstructured text and even voice, and better support for newer channels like text messaging (SMS) and email all contribute. Without an EDM approach to decisions, however, all this will be for naught. Providing “extreme” personalization requires:

  • Fine-grained, analytically derived segments for targeting so that customers feel that responses (offers, messages, style) are really aimed at them and that their needs and wants are understood.

  • Rules to implement policies and regulations to make sure that every channel, every representative, interacts appropriately and legally every time.

  • A customer’s own rules and preferences must be taken into account. When a channel is picked for a communication, it should be based not just on effectiveness but also on their preferences. Customers must be rewarded, with more personalized service, for providing you with the data you want from them about their preferences.

  • It is not enough to measure response. You must move to prediction of responses. You must be able to get a reasonably good estimate of how a particular customer (or at least a micro-segment of your customers) will respond to a particular message or offer. People will not feel that you have personalized your interaction with them if you keep offering things they don’t want.

  • Effective correlation of lots of data inputs so that all your information can be brought to bear on the personalization. No matter how customers have interacted with you (email, store, web), you must reflect that total set of knowledge in your interactions.

  • Rapid response despite the complexity of all this. Personalizing a response only helps if it can be done seamlessly and that means quickly so that the customer sees the same response time for a personalized response as for a generic one.

The best way I know to do all this is to adopt enterprise decision management for the critical decisions as only an EDM approach will let you truly personalize these decisions and thus your business. Adopting EDM gives you the business rules platform you need for best practices, regulations and customer preference rules. Business rules management systems also allow business managers to bring their expertise to bear directly by empowering them to manage the rules. Predictive analytic models ensure that the data you have gathered is turned into something executable, not just reported on. The forward-looking aspect of predictive analytics also makes for better personalization based on what you will want/need, not what you used to want/need. Decision services act as a focal point to bring all your data together and then deliver the personalized decision across all your channels. Finally, adaptive control gives you the kind of test-and-refine approach you need to make sure that your personalization is working for your customers. Personalization must deliver value to customers, and they must appreciate it. Otherwise, they will not participate; and without their participation, your ability to gather the information you need to personalize effectively will be limited. EDM helps you move personalization from the trivial to the effective.

Before wrapping up this article, I want to use one of my favorite examples – interactive voice response or IVR systems. Think about the IVR systems you use or the ones your company runs:

  • Do all customers get the same option or do you decide what options to offer each customer based on what you know about them?

  • Are the options the same when you get a call from a prospect as when a customer calls?

  • Do you make any attempt to predict what your customer might be about to do based on what you know about them so you can prioritize those options?

  • When you add an option to your IVR, ATM or other self-service channel, do you think about each customer and decide if they will want to see or hear that option?

  • When your automated systems refer a caller to a person to make a decision, does that person do anything to make the decision that the system could not do?

Most companies would have rotten answers to these questions. Why? Because they are not thinking about these decisions at a micro level – they are thinking about them only at a macro level. The trouble is that each customer assumes, consciously or unconsciously, that these decisions were taken specifically for them. If you make them listen to the same 5 options every time, even though they always pick one of two, then they assume you don't know them or don't care. Every time you or one of your systems interacts with a customer, you make a decision about that interaction. If you choose to make it the same way every time, then that's your choice but you are still making that decision. Personalization should mean changing this so that every interaction feels personal.

One last thing. Seth Godin once had a post called Blow up your home page. Part of personalizing the experience your customer has comes from your web presence, and when a customer or prospect visits the "home page" of your site, it feels to them like you made an explicit decision to display that content they see. But did you? Probably not. Probably you made a decision about everyone who visits or perhaps you made a decision for everyone with a little bit of space reserved for logged-in users to get something special. But the whole page - its design, its content, its arrangement – can be driven by what you know about a customer, a previous visitor or even typical new visitors. This is extreme personalization, taking that hidden decision (what should this customer see on this page at this time) and making it explicit.

While the primary focus of applying EDM to customer service should not be cost containment or reduction, there is still a role for EDM in managing costs. For instance:

  • With 25% churn and complex IT systems, training costs are a huge problem for call centers. By automating decisions, EDM can reduce the complexity and thus cost of training as a call center representative simply needs to know how to invoke the decision and pass on the answer to the customer, not how to make a potentially complex decision.

  • Unnecessary calls to agents because the IVR system is unmanageable increase staffing costs and wait time. A better IVR system, personalized to a caller and designed to allow automated actions, means that fewer customers will want to “get human,” meaning fewer staff and shorter wait times.

  • Because the website is brochure-ware and does not let customers do anything, more customers call in than is necessary. A website that supports true self-service, with automated decisions, means that more customers will self-serve and fewer will call in. Automating cross-sell and up-sell on the web even means you won’t lose a sales opportunity.

  • Time spent by agents reviewing customer records or querying business intelligence (BI) systems to try to determine if a refund can be paid costs real money as well as reduces the quality of experience for customers. Automating the decision, not the display of information, speeds response times and so reduces costs while making it easier for call center staff to get the “right” answer.

  • Automation that is hard to verify for compliance with regulations can cost money for fines and manual audits after the fact. Using business rules to manage decisions makes for easier compliance as business owners and legal experts can directly review the rules being executed. Fewer errors, fewer fines, less costly review cycles.

  • Customer service systems require constant change to cope with changing demographics and customer expectations. These costs too must be considered part of the cost of customer service. EDM helps build agility into the system, making it easier and cheaper to change pricing, delivery, refund and other rules without spending on coding or, indeed, on training.

There are more than just these, but you get the drift. EDM does not just personalize and improve customer service while building loyalty, it reduces costs too.

EDM delivers value across the decision:

  • Increased precision – more targeted offers that focus on smaller, tighter segments of your customer base

  • Improved consistency – across channels, across representatives and over time

  • Cheaper and easier – agility with externalized decisions that are easy to change, independent of systems or processes

  • Increased speed – rapid decision making enabled by automation reduces call length and reduces the number of calls, reducing wait times

  • Reduced cost – less training, fewer representations, less time on hold


  1. Definition from CRM Daily
  • James TaylorJames Taylor

    James is the CEO of Decision Management Solutions and works with clients to automate and improve the decisions underpinning their business. James is the leading expert in decision management and a passionate advocate of decisioning technologies – business rules, predictive analytics and data mining. James helps companies develop smarter and more agile processes and systems and has more than 20 years of experience developing software and solutions for clients. He has led decision management efforts for leading companies in insurance, banking, health management and telecommunications. James is a regular keynote speaker and trainer and he wrote Smart (Enough) Systems (Prentice Hall, 2007) with Neil Raden. James is a faculty member of the International Institute for Analytics.

Recent articles by James Taylor



Want to post a comment? Login or become a member today!

Be the first to comment!