Benchmark survey: customer service.

For most of the last ten years, software companies have managed to keep their customers at arm's length. Dealers, consultants, and other third party organizations have always handled the face-to-face selling, held the customer's hand, and built the long-term account relationships. Typically, the only point of contact between developers and user3 has been the tech support department, whose employees handle a narrow range of post-sale, product-related questions.

Lately, however, software companies have begun rediscovering--and recapturing--their own customers. Especially in mature product categories, "installed base marketing" has become the new buzzword. More than ever, companies are beginning to think about ways to leverage customer loyalty, to encourage word-of-mouth recommendations, and to boost sales of aftermarket products and services.

As part of this rediscovery of the customer, customer service groups have begun to play an increasingly critical--though often ill-defined-- role. To get a better sense of how that role is evolving, we recently put together a survey of industry practices and performance standards for customer service. Over a two-month period, we conducted telephone surveys of customer service departments among Soft-letter 100 companies (the hundred largest independent developers and publishers in the U.S.).

We completed 58 interviews from this group of companies, and the results--summarized in this special report--provide an overview of how the software industry currently handles customer service operations. This survey is unusual in one respect: Ordinarily, about half the companies in our Benchmark Survey sample databases fall below the $1 million level in annual sales. We realized very quickly, however, that responsibility for customer service in small companies tends to be shared by the whole organization; thus, these companies rarely have discrete data about staffing, call volume, or other performance metrics. So this time we decided to focus only on a sample of larger companies. About half of the responses in this survey (27) came from companies with annual sales over $10 million--the mid-point of this year's Soft-letter 100 rankings--and about half (31) from companies with sales from $3 million to $10 million. When we analyze the information these 58 companies provided, we see several important trends:

* Role in the organization: Although a fair number of companies still seem to think of customer service as synonymous with tech support, most larger companies have set up separate groups--one to answer technical, post-sale questions about product features, and another to resolve sales-related problems. Overall, 42 companies in our survey report that they now operate a dedicated customer service department or group, compared to 16 that don't have a separate group. Larger companies are slightly more likely to set up separate groups: 81% of companies with sales over $10 million have dedicated customer service groups, while only 65% of companies below the $10 million level have dedicated groups.

It's worth noting that not everyone believes customer service belongs in a separate organization; of the 16 companies that don't have dedicated groups, ten insist they have no plans" to create a separate customer service group.)

* Reporting channels: Among companies with dedicated customer service groups, 53% (20 companies) place that group in their sales or marketing organization, and 16% (six companies) have groups that report directly to the president or chief executive: Who oversees customer service?*

  Under $10MM Over $10MM      All
        Chief executive                   5         1            6
        Sales/marketing                  11         9           20
        Operations; Other                 4         8           12
        * Based on number of survey responses to this question.

* Customer service functions: Customer service groups tend to be a

catch-all for a wide variety of tasks, ranging from handling

inside sales (especially of upgrades) and routine correspondence

to customer newsletters, product returns, and ad tracking.

Often, the customer service group acts as a first-level call screening service, directing incoming calls to appropriate departments. One survey respondent described his department's role as the "customer's advocate," and another reported that customer service representatives "sit on product development teams and review all product literature." But perhaps the most common customer service role, according to our survey data, is simply to resolve "mundane" problems--for example, replacing a defective disk--that other departments aren't set up to handle. Ranked by frequency of mention, here are the primary functions that customer service departments say they perform:

What does customer service do?
                             Under $10MM Over  10mm        All
Solve mundane problems            28           20           48
Handle upgrade orders             27           17           44
Process/authorize returns         25           17           42
Provide presale literature        22           19           41
Sell aftermarket products         23           17           40
Survey customer satisfaction      23           17           40
Process inbound phone orders      24           15           39
Process product registrations     18           13           31
Run customer service BBS          11           10           21
Track advertising leads           13             7          20
Produce customer newsletter         8            6          14
Organize focus groups               3            8          11
Operations; Other                   4            8          12

Staffing levels: Like tech support, customer service tends to be people-intensive. Among companies with dedicated customer service departments, the median staffing level is 7% of total company employment (compared to a median of 12% for tech support employees). Moreover, size doesn't seem to yield significant economies of scale: Companies with revenues above the $10 million mark devote 7% of their staff to customer service; those with revenues below $10 million devote 7.5%. (By comparison, tech support staffing ratios are actually slightly higher among larger companies.)

Customer service and tech support staffing ratios

  Under $10MM Over  10MM        All
Customer service
  Median                          7.5%         7%           7%
  50% Range                       5%-11%       3%-11%       4%-11%
Tech support
  Median                          11%          15%          12%
  50% Range                       5%-15%       12%-18%      5%-17%
 *Percent of full-time employees vs. total company staff.

We also asked about the career backgrounds of typical customer service employees. Although the data is sketchy, it's clear that customer service is ordinarily an entry-level position (and thus presumably low-paid). Not surprisingly, companies tend to look for employees with sales skills and outgoing personalities, rather than strong technical backgrounds (in fact, several respondents said they believe tech support employees usually don't perform well in customer service environments). Call volume: Software companies have become increasingly sophisticated about measuring the call volumes for tech support, but we found a good deal of guesswork about customer service statistics. Overall, however, a typical customer service representative seems to be handling about 50 calls per day, averaging 3.5 minutes of "talk time" per call. Larger companies tend to achieve higher levels of productivity, probably as a result of greater automation and employee specialization.

Call volume and average talk time
                             Under $10MM Over  10mm        All
Median calls per employee         40/day      70/day      50/day
  50% Range                       20-50       40-90       30-70
Average talk time per call        3.5 mins    3 mins      3.5 mins
  50% Range                       2.5-5 mins  3-5 mins    2.5-5 mins

One surprising discovery: There's apparently no correlation between product price and the complexity of customer service calls. We split our database into two roughly equal groups-- those with flagship titles that sell for less than $300, and those over $300--and found that median call volume and talk time were identical for both groups.

Performance metrics: Measuring service quality has always been an elusive problem. We asked our respondents what yardsticks they use to quantify customer satisfaction or to measure the performance of the customer service department. Our interviews suggest that companies still rely heavily on informal sources of feedback--for example, comments on registration cards, customer correspondence, user group meetings, CompuServe forums, and the like. However, a substantial number of companies also conduct occasional customer satisfaction surveys, and many others monitor such formal statistics as product return rates and problem resolution.

How do you measure customer satisfaction?*
                           Under  10mm over  10mm      All
Formal customer survey          21          13         34
Product return rates            19          13         32
Percent of problems resolved    17          12         29
Number of calls per problem     15          10         25
Renewal rate on paid support    11          10         25
* Number of mentions among survey respondents.

Future revenue value: one way to judge the value of customer service is to measure the long-term "asset value" of a customer, in terms of potential upgrade purchases, aftermarket sales, and perhaps word-of-mouth referrals. A few companies have developed models that suggest an average registered customer represents as much as 500-$750 in future revenue; thus, investing in customer satisfaction (or simply keeping a registration mailing list up to date) makes a contribution to the company's bottom line that can be quantified with reasonable accuracy.

According to our survey interviews, however, only a small number of companies--13 out of 58 firms--currently have any idea of the revenue potential of a typical registered user (and even among these companies, most customer service managers don't know the specific dollar value their companies attribute to a customer). Similarly, only nine companies report that they have analyzed the return on investment from cleaning up mailing lists or keeping phone numbers current." This lack of hard data, we suspect, implies that software companies still view good customer service more in terms of image value than tangible payback.

Other statistics: As part of our interviews, we collected data on several topics:

Slightly more than half (53%) of the companies we interviewed have set up database systems that they use to keep track of customers (frequently, the same database is used also by the tech support organization).

About two-thirds (62%) of our respondents provide a toll-free phone number for customer service.

There's little consensus about how to provide customer service to overseas users. Twenty-two companies say they rely on local dealers to handle customer questions, 17 say they have in-country staffed offices, and two have set up regional support groups. However, the most popular method for communicating with overseas customers still seems to be the fax machine.

For readers who would like to explore our customer service database in greater detail, we've created a supplementary worksheet that contains all of our survey data (except for company names). To order, specify format (available formats include Lotus 1-2-3 3.5" and 5.2511 DOS or Excel 3.511 Macintosh). Price: $50, prepaid only.

POSTSCRIPT: As this report suggests, we feel strongly that customer service is about to become a critical topic for the entire software industry. We've already added a day-long customer service track to OpCon, our twice-yearly operations conference (Sept. 30Oct. 4 in Cambridge, Mass. and March 2-6 in Santa Clara, Calif.). And weld very much like to hear from readers who have developed innovative, state-of-the-art approaches to customer service.