13.3. Making Technology Decisions

We have often discussed the dangers of creating a CRM infrastructure component without heavy involvement of the business functions that are already working closely with customers and delivering customer experiences. But few people outside the IT profession have any confidence in their ability to make technology decisions. The truth is that some elements of the IT solution require little, if any, perspective of the business, and others require a great deal. With a good understanding of the business organization and where work gets done, the IT team has a lot of the information they need to design the overall network architecture and select the appropriate hardware components that will best satisfy the business requirements.

13.3.1. Application Software—It's Always a Tradeoff

Application software tells the computer what information is needed to make decisions and how to perform activities based on that information. Picking the right software to automate processes and capture and store business information requires functional knowledge and expertise from the people who know the work that needs to get done. How do we get and keep the business engaged during the critical step of making software application decisions? The bottom line is this: the functional team needs the information about the business impact of the different options available for some of the key decisions. There are many decision-making tools in general use, and one of the simplest is the decision matrix.

As we discussed in Chapter 12, many companies hire expensive consulting firms to help them select the best software solution, but it can't be totally done for you by a consultant any more than your own IT team can do it alone. The business experts must be involved comparing the business requirements to the software capabilities and make any necessary tradeoff decision. Consultants and Information Technology can help a lot, but they can only help.

The decision matrix is a useful tool for comparing two or more alternatives across multiple factors. The simplest version looks like Table 13-2. Here's how you create a decision matrix. First, list all the alternatives across the top, and then list the important decisions factors down the side. For each cell, indicate how well that alternative meets that factor (0 for “not good at all,” 1 for “adequate,” 2 for “good to great fit”). The highest score wins.

A good way of getting started on a decision is to put a brief comment in each cell that describes the characteristics or impact of the alternative. Give this to all team members, and have them input a score. This can be done via e-mail, interview, or in a team meeting in which each alternative or factor is discussed in detail. More sophistication can be introduced by adding a weight to each factor. Sometimes, these are called simply “Musts” and “Wants.” If an alternative can't meet a must-have factor, it is eliminated. Alternatively, the factors can be grouped into importance ranks from high (5) to low (1). The final score is determined by multiplying the score by weight. All of these decision making variations work well, depending on the complexity of the decision and the time available. One of the most important results is the discussion that is generated through the weighting and scoring process.

Making sure that a package has adequate functionality is not the same as insisting that it accomplishes the function exactly the same way you do today. It was important that XYZ's call center had the tools and information necessary to create a customer quote and take an order. It was not important that the quote and order be created using exactly the way it's always been done. And finally, don't get stuck in terminal consensus building. Sometimes somebody just has to be responsible and make the decision. Be sure your sponsor and steering committee are up to this.

Build or Buy?

One of the first decisions that must be made is whether to buy a product off the shelf or build a customized solution. In the early days of computer automation, little software was available on the open market; everything was custom-built. This turned out to be a very expensive proposition, and many companies today prefer to buy solutions whenever possible. As more options become available for business activities, your choices continue to grow. Let's look at a simplified version of XYZ's decision-making process for Valencia, as shown in Table 13-3.

From the decision matrix, it appeared that buying is a better choice for XYZ, but there was one problem. Because of all the work that had been done, XYZ knew what it needed to consistently identify customers, but they had not yet found a product on the market that met their needs. Because buying software was so advantageous, the company decided to wait and do a more in-depth analysis of alternatives available, including a benchmark of what companies with successful customer relationship programs were doing.

Single Solution or Component Best in Class

This is another tough decision, and it's related to the “build or buy” choice. If you get a single, totally integrated solution, you should be able to minimize the amount of customization that needs to be done. This should be (and is) much cheaper than having to build and maintain all the connections between bunches of different tools. But there's a catch; it is not possible for any total solution to be best at everything that needs to get done for a successful CRM program. Most are outstanding in one or two areas, but often barely adequate in the rest. Even if acceptable for the majority, it's tough to remain in the lead position for everything. Remember that we want to keep the information and replace the tools as better ones become available.

The big problem is that most integrated software solutions are proprietary. Their database structure is built into the application and is not easily made available to other vendors' products. Nonetheless, XYZ's basic system strategy was built around integrating the best-in-class modules. XYZ took the approach of minimizing the number of connections that were required by connecting systems through hubs, rather than linking every application to every other application.

There's another problem, too. If you decide that you will only select a product that provides a total solution for everything you need, then you are forced to evaluate/compare all the functions at once, instead of being able to take small steps. It is difficult, if not impossible, to make such a big decision in a reasonable amount of time. Even if you plan to implement in chunks, how do you compare across all the activities you will need. Sometimes, companies will select a vendor who promises that its solution can be tuned to fit all requirements. Be forewarned; in most cases, one size fits none.

And finally don't forget your current (legacy) environment. You can't replace everything at once, so it will be a long time before you are actually running on one single total solution. If you are a small company that's just starting up, you may be able to buy a solution off the shelf and just operate your business the way the product works. The first time most of us learned to use an electronic spreadsheet, we just learned that typing “c” allowed us to copy data from one place to another. Changing that familiar task was difficult because we didn't even think about what we were doing. Even if a new product was better, at first it was frustrating and made us much less productive. This is the same situation. Of course people can learn to change, but it takes planning, resources, and effort to make change successful. So despite all the hype and promises, forget it! There is no silver bullet. This leads us finally to the question, what happens if we decide to integrate components?

Component Integration

There are several integration issues that must be faced if the organization opts for integrating best-in-class components. Of course, unless you can replace everything at one time, you will be faced with these issues anyway if you are fitting new tools into your existing environment.

The decision you make will be based on the cost of the integration efforts versus the value of being more independent of technology single solutions. You also have to ask whether your IT department has the knowledge and experience to manage complex integrations.

Evaluate and Select Your Tools

This is what you've been waiting for, right? This is where you find out exactly which CRM software products you should buy. Sorry, it just doesn't work that way. Even listing names of applications and vendors is so time-sensitive that, in a year or two, it becomes be meaningless, and forget keeping up with all the changes in what they do or how well they do it. Look at the example in Table 13-4, which presents the results of a survey done about two years apart. Each survey identified the CRM product leaders at the time. Look at all the changes.

By the time you can buy this book, some of these applications will have disappeared, others will have expanded functionality, and new ones will have emerged. What we can do is help you learn how to make your own decision as to what's right for your company and the time you need to make a decision.

Remember that different people learn in different ways and need different types of information to make a good decision. Some people learn by reading; these people want scads of product documentation. Others learn by hearing and want a visit from a sales team with product expertise. And some learn by doing; they want to get their hands on the product and try it out. Be sure you can meet the needs of all three types of learners.

There are several places you can go to help your company make the decision that is best for your business:

• The Internet: This is probably the best source of information because it is easy to find current information. From the Internet, you can find lists of products to consider, decision methodologies to apply, consultants to hire, and tools to use. Use your favorite Internet search engine, and enter your criteria (e.g., “CRM software selection”). One of the best sources of information that I found was a very practical and information-rich source at www.softresources.com.

• Consultant: Consultants bring a wide range of knowledge about different applications and experience about how different companies are using them.

• Books: The ones that tell you what to buy are not my favorite because of the aging issue. But if you're in a hurry and don't have time to do lots of investigation (and you can find some recent results), the $900 price tag on some of the customer satisfaction survey books (i.e., Consumer Reports for CRM) may be worth it to you. They do a very decent job comparing features and reporting results. However, my preference is for the books that tell you how to make your decisions such as Dick Lee's straightforward (read that “blunt”) and highly readable book about surviving CRM. Lee's Survival Guide presents a ton of common sense and practical guidance. He says, “NEVER, NEVER, NEVER let CRM software dictate how you do stuff,” and I wholeheartedly agree.

So, what we really have to do to make the decision that is best for our organization is to again create a plan and follow a process. This is how you will be able to make the best choice based on the information you have already generated as part of the project. Every step builds on previous steps; there is no starting over. These are the application selection process steps:

Table 13-5 shows the high-level summary of XYZ's RFP results for the database marketing process. Of course, the RFP itself divided activities into lower-level work tasks such as the various campaign management steps like the contact plan, execution, and results tracking. The complete summary for the project also included the data management and sales functionality.

Based on these results, XYZ decided to eliminate SAS as an alternative because (at that time) SAS didn't offer campaign management capabilities. Even though it had the strongest analysis and reporting capabilities, the team needed campaign management capability integrated into its database marketing solution.

Notice that although all four products had databases, none had clearly met the customer database (ID) requirement. Even though all four vendors had indicated that they had database and customer identification capabilities, it wasn't clear what “yes” meant. SoftResources has determined that, on average, vendors answer yes to 85 percent of RFP questions. They claim that “Yes” can mean anything along a whole continuum from having exceptionally robust features in that area to having the ability to modify the product to meet the requirement (as long as you're willing pay us lots and lots of money). Sometimes it even means “I don't know what you're talking about and I don't think you do either, so I'm just going to answer, 'Yes'.” That's why the next step is so critical; XYZ set out to do due diligence by actually testing the applications in-house and walking through the company's operation. Even more valuable is talking with companies who are currently using each of the products.

13.3.2. Where to Start—On or Off the Web

The reason there's a dilemma about where to start is that most companies even today get very little of their data from web sources. Although that's where everyone hopes to be, the web sources provide only about 15 percent of new data captured. Considering that many companies have been collecting customer information for years, the web probably accounts for less than 5 percent of a company's total customer information. Although we expect and hope that this will change over time, it will be quite a while before web sources overtake non-web sources as data generators.

So why is this imbalance such a problem? Web-captured data is likely to be pretty high quality because presumably the customer himself or herself has supplied the information directly to our web site. But few companies have enough web data to really answer the important questions they have about their customer relationships. To get more information, we are forced to supplement web data with older (legacy) data. Now, if we integrate all this old stuff into our Internet application, we are giving customers online access. They will be able to see all that horrible old data we have about them. It's just plain embarrassing!

The way to get around this problem is to begin your integration efforts offline, using either web or non-web data as appropriate to your project. Make data cleanup and quality a large part of the project effort. (We'll discuss this more in Chapter 20.) After you have sufficient customer information at an acceptable level of quality, you can begin to integrate your online access capabilities. Whenever possible pick the alternative with the lowest risk.

13.3.3. Putting the “I” Back in Information Technology

Remember that customer information is a company asset and must be treated as such. The IT department is responsible for more than just doing technology right. The business functions are responsible for identifying what information is important to them. Information Technology has an equally important role: They must capture the information the way the business users describe it—not because it's easier to implement something else. If the system doesn't capture the context of the information, we don't have any. Then it's just data.

Now we're ready to look at the fourth component of CRM – the people component. Without a doubt, this is the most difficult component to manage, and it will make or break your CRM program.