“How do I ask my customer?”
Think back to the days when you were about 12 years old. You really wanted to have some candy before supper, but your mom said no. If you were anything like me, the very next thing you did was run and ask daddy to see if he would say yes. Why is this important to you today?
When we have a question to which we know what we would like the answer to be, we have the skills and abilities to fish for it. However, what happens when you objectively want to know something? The truth is that you can make survey results say anything. How do you get the truth from your customers and prospects without influencing the outcome?
Who do you ask?
When we talk about surveys from the standpoint of statistics, the first consideration is always that you have to ask the right people. When you are able to ask everybody, you are using the population. Most of the time, however, you do not have access to everyone, and even if you did, not everyone would answer your survey. This is why we use a sample. You will often hear the term “representative sample” to describe a situation in which you take a piece of the population, and the results would be the same if you were to question the whole population.
If you know who your customers are, you can make sure that when you get your surveys back from a few of them that the responses represent all the non-respondents accurately. There are many biases that can alter these results. Let’s say, for example, that we wanted to know what our readers value about our magazine. Ideally, we would get responses from all of our readers. If we were perfect, we may even analyze non-readers to understand what is missing for them. Instead, we send surveys to everyone signed up to our email newsletter online. This would likely get us responses that indicate that our customers value online content. Perhaps, there would be a high percentage of individuals who work in prepress within the sample. That is, however, not a representative sample for our magazine. There may be many individuals who only receive and enjoy the print version of our magazine. In this case, we would never find out what they like about us.
Sampling goes back to the idea of asking mom first, then trying dad, except that we don’t recognize that we are altering results. It is very important to make sure that you understand as much about your population as possible. Make a list of customers that are likely to do business with you. Start sorting into business size, location, revenue, etc. At the end, you can say something along the lines of “my population (as best as I can describe it) is a small-to-medium size real estate company in the Winnipeg area.” From then on, every time you conduct a survey you think about how closely you match that sample. While there are statistical methods to help shape the sample to your whole population, you must also think about how the sample might affect your results.
What do you ask and how do you ask it?
The most important thing to understand about surveys is that people will not give you more than 10 minutes of their time to finish them. This means that the questions must give you the most bang for your buck without going past that amount of time. There are different types of questions you can ask your customers – but they should all be relatively painless to answer, otherwise you’ll get no responses, or even worse, inaccurate responses. For example, “how much did you spend on paper last year?” might work. “How much did you spend on recycled paper with 30% or more post-consumer waste?” will at best give you a guess. Unless you are paying your customers to fill out the survey, they won’t look it up.
Simple questions that require the customer to rate something on a scale fare very well. If you are using a Likert scale, or any kind of scale for that matter, it is best to tell the customer what the numbers mean. As a simple example, you can say 1 = strongly disagree, while 5 = strongly agree.
As a tip, if you know who the clients are, don’t ask them to re-describe themselves. Do the work to look it up for yourself. For example, don’t ask what they print, unless you don’t know. If you have a reasonable idea and have seen most of their work, don’t waste their time asking the question. In addition, be careful with dichotomous questions. As an example, we often ask people if they are male or female in surveys, while the answer may in fact be “other.” Creating your survey with a group of people really helps in that regard. The great thing about customer surveys is that we are all customers. Think about the things that you would want to be asked, and ask them.
In today’s market, client data is very valuable and surveys are highly cost effective. There is no need to spend thousands, or even hundreds of dollars, to survey your clients. When you are wondering about your population, you may look to purchasing market survey reports (ex. NAPL). When you have a good grasp of who you would like to know about, and have access to them, you can use online tools that are very cost effective. No more difficult than setting up an email account, you can use software to ask your questions. The list in the table above is just some examples to get you started. They are all user-friendly and cost effective. You can begin by practising in the office, then moving the surveys out to clients.
Happy surveying as I leave you with my favourite stats joke of all time: “75% of statistics are made up on the spot!”