19.1. Evaluating Performance Metrics
In Chapter 18, we set some very specific and measurable goals for the project offer as shown in Table 19-1 (specific, measurable, and time bound).
| Objective: Increase Profit | Performance of | Metric |
|---|---|---|
| Sell 150 total units of either Series 825 or 827 systems to the targeted customers within four months | Orders | Toll-free number orders for 825 and 827 products |
| Reduce by 35% the number of 815 and 817 systems owned by targeted customers and under support contract. | Number of total 82x purchases by target customers | |
| Leads | Number of inquiries | |
| Overall | Number of calls |
The project team wanted to know how many calls were received, how many inquiries were received (and later turned into qualified leads and finally sales), and how many customers did exactly what the offer asked them to do (buy via the toll-free telephone number). Maybe most importantly, XYZ measured all the orders for 82x products placed by targeted customers during the six-month project period. Of course, it can be valuable to look at new results periodically. The period should be based on the average buying-decision time for the product, but a good rule of thumb is to check twice more over a period equal to the original project period (e.g., for Valencia two more checks at three-month intervals is a good plan).
The purpose of performance metrics is to measure the impact of the specific campaign in two areas: the overall company performance, and customer satisfaction and loyalty.
19.1.1. Company Results
As we know, the end goal of CRM is to increase profit by either increasing revenue or reducing costs.
Increase Revenue
We counted the number of orders that were received during the project period, so now we can easily calculate the increased revenue, and by subtracting the cost of sales, we know the impact on profit. But this ignores the reality that many campaigns don't deliver immediate results. We also need to measure sales that might be gained through the influence of the project on future product purchases:
Leads that were converted to buyers who purchased after the four-month project window
Orders resulting from referrals and offer information that was passed along to customers not in the target segments
Orders resulting from success with the original offer (test purchase)
And the influences go on, depending on your project situation. For example, there is the “top of mind” result from receiving a recent (and appealing) offer. In fact when surveyed, 65 percent of HP's customers complained that they didn't receive enough information! It is difficult to measure indirect influences, but there are ways to identify the source of the order that we discussed in Chapter 18.
Reduce Cost
For the Valencia project, there was a major cost savings that contributed to company profit. Product support contracts on the old 815 and 817 models were no longer moneymakers. Over time, the equipment had become prone to failure, and some of the replacement parts were not standard and needed to be special ordered, so they were very expensive.
Not all projects will have this particular cost savings feature, but there are other ways of reducing company costs. The Valencia call to action encouraged the use of the call center to place the order. On average, it took about two hours to identify the proper product configuration, prepare the quote, and enter the order. A very conservative rate for a sales rep (fully loaded) is $500 per hour. For a telesales rep, it's about $150 per hour. It is safe to subtract about $700 from the average cost of sales for orders taken by phone. Here are some other examples:
A new system for configuration is easier and more efficient to use.
The telesales team has an automated knowledge base, so configuration and quote generation takes less time.
Targeting the right customers means fewer wasted offers.
Remember to look at all the cost saving opportunities that impact profit, not just the increase in revenue. Even though increased revenue is the true benefit to be derived from CRM, it can be difficult to show adequate short term profit results.
Return on Investment (ROI)—Reality or Myth
ROI is another performance metric that is often used to compare the cost of delivering a project to the financial results achieved. Financial return on investment is calculated by dividing the total cost of completing the project by the total cost savings and sales (minus the cost of sales). It's a very sound measure of the worth of the value of a project, and it has long been favored by Finance. But it's not completely clear that a CRM project can recover in the short term the cost of building the new infrastructure. Certainly, we can measure the ROI on the offer (cost to produce and deliver message divided by profit that resulted from the offer), but to justify all the computer and network costs, programmer salaries, and training and education of the project team and the new end users is another story. To fairly evaluate the whole CRM project cycle, we need to consider returns over a period of years. Of course, no one wants to wait five years to see whether an investment was worthwhile. Generally, we measure the results of the initial offer(s) and project what we'll expect to gain by using the same infrastructure to repeat the offer over a period of years. The time to recoup all the expense is called the payback period and depending on the company's financial situation, about three years is generally considered reasonable.
Arthur Hughes prefers Lifetime Value (LTV) measures to ROI for just this reason. Time framed measurement approaches like simple ROI have, according to Hughes, the problem that although
ROI is clearly the way to measure the immediate result of any direct marketing effort…LTV [Lifetime Value] calculations include such factors as the retention rate, the referral rate, and the (long-term) spending rate, rather than just the response to an immediate promotion.
Calculating the payback period is much like LTV in that both look at the contribution over a number of years.
19.1.2. Customer Results (Loyalty and Satisfaction)
Another area, in our customer-centered organization, that is important to measure and understand is the impact on the customer relationship. At the very least, check a sample of customers for satisfaction and/or loyalty changes after every project. Remember, as long as you perform these tests on a fairly regular basis, the trend in what you find is more important than whether or not you have the perfect survey and perfect tools to product an absolutely reliable score. We know that satisfaction is a necessary but insufficient condition for customer loyalty, but it is a good place to start.
Satisfaction
Satisfaction can fairly easily be measured simply by asking your customers how they feel about the contact they just received from you. Of course, we don't want to have the survey cause dissatisfaction; too many interactions can be a bad thing. Satisfaction surveys should be used judiciously in the context of an overall plan for customer contact.
I recently bought a new car. As I was leaving the dealership, my salesman and his manager both told me that I would be receiving a survey regarding my satisfaction with the buying experience. Unless I could answer that I was satisfied with the experience, I was told, my salesman would be in trouble. Hmmm, I got the point—but did the auto company? What did they think they were measuring anyway? Did they know what was happening in the dealerships? If they did, were they serious about really understanding customer satisfaction or just serious about being able to publish high satisfaction scores? Unless I absolutely despised every moment of the car purchase (in which case I probably wouldn't have purchased anyway), I was going to indicate that I was at least very satisfied if not delighted with the experience. I didn't mind helping out the sales guy, because he seemed to need the support. But doesn't the auto company really want to know if their dealerships are satisfying customers? Frankly, the whole experience turned me off – but of course, I didn't tell them that! Now I just share the experience with clients as an example of what not to do.
There are so many better alternatives if you really want to know what your customers think. Why didn't the car company just send me the survey without any coaching? Better yet, they could have contracted for a telephone survey from an independent telemarketing company. If you want to avoid bothering too many customers, remember that you don't have to sample everyone. You can gain significant information from a small, random sample of customers
In all our relationship management efforts (from interaction to follow up to satisfaction surveys), we should always remember the meaning of the Hippocratic oath: First, do no harm. The oath not only makes good moral sense, it makes good practical sense too. Campaigns that damage customer relationships (for example, unsolicited e-mail, which is always perceived as spam) should be halted immediately and never, ever repeated. At least you've shown your customers that you listened to their concerns even if the original practice was unpopular. Remember that we discussed that fixing a customer problem can actually increase satisfaction. Never making any mistakes is about like never borrowing any money; it doesn't hurt your “credit” rating, but it doesn't do it any good either.
As we've discussed, satisfaction is most valuable when you also ask the customer how important each characteristic is. Don't spend time improving areas that customers don't care about. An example of such a two-pronged question is shown in Table 19-2.
| What is your impression of your salesperson's knowledge of the product? | ||
| Poor | 1 2 3 4 5 | Extensive |
| How important is it that you have a knowledgeable salesperson when making a purchase? | ||
| Not Important | 1 2 3 4 5 | Very Important |
Finally, satisfaction is more important as a customer relationship trend than as an absolute score. Compare current satisfaction scores to the results from previous surveys to identify the current experience as a relative measure, rather than an absolute one.
Loyalty
In many cases, loyalty is a blind emotion, but customer loyalty is based on the customer's value perception and is measured by behavior. Satisfaction is an indicator, but true loyalty is evidenced only over time. For some products (like refrigerators or, in my case, cars), the buying cycle is very long; it may be years before the dealer knows whether I will be a loyal, repeat buyer. One good measure, though, is where I get the car serviced, which also can impact overall loyalty. Loyalty must be measured over time, but satisfaction is often used as a loyalty indicator.
Share of Wallet
Market share measures the percent of purchases from your company compared to all purchases made in your market. In the customer relationship world, Peppers and Rogers popularized the concept of “share of wallet.” Share of wallet represents the percentage of what a customer spends in total in your market compared to how much of that money comes to your company.
Share of wallet is another good indicator of loyalty. Of course, the calculation is made based on a budget estimate (size and type of company, income and address of consumers). Few customers will tell you how much they have to spend on computer technology or on dining out over the next year.