Market Research for ISPs

Question: Large internet providers such as EarthLink can afford to do user surveys and focus groups. How should a small ISP with fewer resources find out what new products its customers want?

by Brock Henderson Principal, Henderson & Associates [September 18, 2006]

Question: Large internet providers such as EarthLink can afford to do user surveys and focus groups. How should a small ISP with fewer resources find out what new products its customers want?

Finding out what your customers want does not need to be a costly expenditure, but your data must be accurate. Let's face it, collect the wrong or useless information and you have wasted your efforts and your money.

So what do you do? Actually, it is brazenly simple: ask your customers. Just ask them. You may not be able to afford a professionally prepared survey or fancy focus groups, but you can create simple surveys and do focus interviews yourself. I will assume that you have three basic groups you want to research: individual consumers, business customers, and resellers (such as network consultants).

Your approach and questions will be similar but different for each group. In general you want to find out three things:

(1) What do they like about your products and services;

(2) What do they not like about your products and services; and,

(3) What products and services would they like to see offered in the future.

For residential customers, a simple survey could be prepared and e-mailed to them. No, this isn't spam since you have an established business relationship with them. If anyone does complain then be courteous and be sure to never send them another survey.

Send the survey to all of your residential customers, only a small percentage will respond, (1 percent to 5 percent), but it should be enough to give you a good look at your service and what you need to improve.

Some questions you might ask:

• Have you needed technical assistance in the past 30 days?
• Did tech support resolve your problem?
• Are you satisfied with the way you were assisted?
• Why, or why not?

The first three questions require a "yes" or "no" response; the fourth question gives you insight as to why the customer was or was not satisfied. This can be very important if the tech support person resolved the issue just fine, but was rude or condescending to the customer.

While it is important to find out what new products and services you should be offering, residential customers may not have a clue as to what could be offered. Thus, asking them, "What services could we be offering that we aren't offering now?" wastes both your time and theirs since they simply don't know.

It would be better to offer them a list of services and ask them to select the one or two services they would be most interested in seeing offered. In addition, you could also ask them if they would be willing to pay for these new services, and if so, how much. This would tell you not only what the more popular services are, but which ones you can charge for and how much you can charge.

Business customers can be given a nearly identical survey, modified to fit the business customer. However since you are probably dealing with a more technologically savvy respondent, questions about future services should include an opportunity for them to list services that might not be on your list of potential offerings.

Respondents who give insightful answers or who raise issues of technical failings should be contacted for a more in-depth conversation—especially those who raise technical issues. By acting on their concerns immediately, you demonstrate a desire to correct the problem and at the same time you are possibly keeping a customer who was considering switching service providers.

While a simple survey could be used with resellers, I suggest you sit down with them individually to discuss your services. This group probably has the best vision for potential new service offerings. While resellers might have the technical ability to create and offer a new service, they may not have enough customers to make the investment themselves.

You, on the other hand could take the idea, develop the service, and then offer it to all your resellers and make it profitable. Everyone wins: you, the reseller, and the business buying the new service.

Constructing a survey
There are two basic styles of questions, Open Ended and Closed. An open ended question would be like "What do you like about us?", allowing the respondent to provide a stream of consciousness answer. A Closed question would be one where you provide two or more choices for the respondent to select from, such as a "yes/no" or multiple choice type of question.

One of the most common mistakes in creating a survey is asking two questions in one, what I call the "double-barreled" question. Such as "Was tech support helpful and courteous?" Tech support might have been very helpful but rude, so how is a person supposed to answer? It is better to make this two questions: "Was tech support helpful?" and "Was tech support courteous?". This way, the respondent can answer "yes" to the first question and "no" to the second.

You can also ask scaling questions, such as "On a scale of 1 to 5, (where 1 is least favorable and 5 is most favorable), how would you rank your last interaction with tech support?"

Group similar questions together. Don't have a yes/no question, then a scaling question, then another yes/no question.

Customers are more likely to respond to a short survey than to a long survey; people simply do not want to take their valuable time to answer your questions. If you tell people that the survey will only take 5 minutes, then be sure that it will take 5 minutes or less. Don't waste their time with too many questions. If need be, divide the survey in half and send the first half to one half of your customers, and the second half of the survey to the second half of your customers. Your results should still be valid enough to give you significant insight into the minds of your customers.

Finally, in six to nine months, send the survey again. Compare the results. Did you fix any problems discovered in the first survey or are they still there? Have you improved, stayed the same, or gone backwards? By comparing the results, you will know whether your efforts have been successful or whether you need to work harder.

While I realize this industry is loath to do anything that even resembles Spam, I do not see how a customer can fault you for trying to improve your services. Call it the "How are we doing?" survey. I'm sure a large number of your customers will be happy to tell you the good as well as the bad. If you don't ask them what you are doing right and what you are doing wrong, you may never know. So ask!

—End