One of the most important pieces of research any shop owner can do is customer satisfaction research. The data it provides can show you your strengths, highlight your weaknesses, and give you an overall picture as to how your customers see you as a business. With this information you can then capitalise on your strengths in your promotional material, identify focus points for improvements in your business and improve how your target audience views your business.
If you’ve never done any research before, it can seem like a daunting and complicated process, but these hint and tips will help demystify the research process and help you gather meaningful data.
Picking your sample
“Sample” is another name for the group of people you want to participate in your research. You want the sample to be relevant, current, and able to provide answers to the questions you’re asking.
To give an example for this, let’s say I was conducting a nationwide student satisfaction survey, asking people how satisfied they were with their experience at universities in the UK. The most obvious requirement for my sample would be that they would all have been students, but I would also want to take into account that some people may not necessarily remember all the details of their university experience if they attended university many years ago. I would also want to take into account the fact some people may not have studied at a UK university. This would mean that my sample was UK residents that have studied at a university in the UK in the last 5 years.
Using this same process for picking a sample for a customer satisfaction study, the obvious requirement is that they have made a purchase from your online shop. The order will need to be fairly recent to be fresh in the customer’s memory, so you may only want to sample customers that have made a purchase in the last 6 months. You may also want to exclude very recent customers as they may have not yet received their purchases and so have not yet had the complete shopping experience from your online shop. These requirements would mean that your sample is customers that have completed an order under 6 months ago but over 1 week ago.
Choosing a method
When choosing a method, there are a number of considerations that need to be kept in mind. First and foremost is the type of data you want to gather. There’s two distinct types of data: Qualitative data and Quantitative data. Qualitative data is used to gain an understanding of underlying reasons, opinions and motivations. For this reason, it most commonly uses open question interviews, focus groups and detailed observations as its methods for gathering data. Quantitative data is information that can be measured and written down with numbers and commonly uses multiple choice question interviews, questionnaires and online polls for its data gathering methods.
When it comes to measuring customer satisfaction, you’ll likely want a mix of qualitative and quantitative data; you predominantly want to know the number of customers that are satisfied or dissatisfied with their experience on your shop, but it’d also be useful to know why they were satisfied or dissatisfied with their shopping experience.
These requirements leave you with questionnaires containing mainly “closed” (multiple choice) questions with some “open” questions (where participants can leave a comment) as the best method for gathering data on customer satisfaction. The final decision is then how you present the questionnaire to your potential participants.
Questionnaires can be conducted face-to-face, over the phone or online, and each presentation method has its own drawbacks and advantages. If you have a bricks and mortar store, you may find conducting the questionnaires face-to-face much easier than if you don’t have a bricks and mortar store, but it can also be time consuming and might be targeting the wrong audience, as in-store customers will not have necessarily used your online store before.
Telephone interviews can also be time consuming, especially if your sample is large, but have the advantage of allowing you to go off script and probe further if there’s a lack of detail in an answer or an interesting topic you want to know more about comes up. Online questionnaires are the most flexible and scalable way to gather quantitative data as they can be filled in at the participant’s’ convenience, are the easiest to send back once filled in and can be sent to hundreds of people with minimal effort and cost.
Constructing your questionnaire
If you decide to use an online facility, you’ll generally find that they have very easy to use question builders that mean you only have to type in your question and the possible answers. Google Forms is one of the simplest to use and will also give you some basic analytic tools once you start getting answers. Another facility (and, as it happens, the one I use for ekmPowershop questionnaires) is SurveyMonkey, which also has a question bank of pre-written questions and templates for a range of questionnaires that may prove useful for someone new to creating questionnaires.
Whatever facilities you use, there’s some basics to keep in mind. First of all, ensure your questions are not in anyway biased. This is where the question itself is leading the participant to pick a particular answer, and this leads to unreliable data. For example, if you wanting to find out how a participant would rate the overall shopping experience on your site, asking “Did you have a great shopping experience on our website?” would be an example of a biased question, as it’s presuming the participant had a positive experience. A better question would be “How would you rate your shopping experience on our website?”, as this wording of the question doesn’t contain a possible rating in the question and is completely neutral.
The second tip to keep in mind is keeping everything as simple as possible. Given that you’re asking people about how satisfied they were with various parts of their shopping experience, you’ll likely be using questions that have a rating scale as the answers. Making sure you use the same rating scale in the same order as much as possible will ensure the questionnaire is easy to follow and easy for a participant to fill out. In general, our questionnaires use a rating scale with 5 possible answers. So, for example, if the question was “How would you rate your shopping experience on our website?”, the possible answers would be “Terrible”, “Bad”, “OK”, “Good” and “Fantastic”. These then cover an extreme negative, a slight negative, a neutral, a slight positive and an extreme positive opinion. If I were then to ask “How would you rate your experience regarding the delivery of your order on our site?”, you could use the exact same answers in the exact same order without the answers becoming irrelevant to the question.
My third tip would be to keep the questionnaire as short and concise as possible. Making the questionnaire too long will put people off participating and will lead to less data. For a customer satisfaction questionnaire, you ideally want 10 questions or less; one asking how frequently they’ve used your website (an optional but useful demographic question that will allow you to break down your sample in analysis), one asking about the overall experience, one asking how they’d rate the site in terms of ease of use and why, one asking how they’d rate the site in terms of customer service and why, one asking how they’d rate the site in terms of delivery and why, one asking if they have any other comments or feedback on their experience and a final question giving them the option to leave contact details if they’d like to discuss their feedback further. This covers all the areas of the shopping experience, allows the participant to explain their answers if they wish to and allows them to identify themselves, but only if they want to.
This links to my final tip, which is only make the absolute essential questions obligatory. This is especially true in the case of the identifying question, as some potential participants may be put off if they have to identify themselves. Allowing participants to remain anonymous generally leads to a higher response rate, especially if you also inform potential participants that their answers will be fully confidential.
When it comes to deciding which other questions should be set to require an answer and which questions should not, I would recommend making all the rating questions required and the questions asking why they have given that rating optional. This prevents the potential discouragement a respondent may experience if they have no specific reason for giving a certain rating, whilst allowing respondents who have specific reasons to express them.
Once you have a questionnaire and sample, you then need to distribute the questionnaire to your sample. If you’ve determined your sample from the customers list in your online shop, you’ll more than likely have email addresses for them, allowing you to email a link to the questionnaire out to your sample. If you do this through an email client or browser based email service like Gmail, you’ll want to ensure that you put your own address in the “To” field then add your customers’ addresses in the “BCC” field; this will ensure you don’t distribute everyone’s email addresses to everyone the email is being sent to.
An alternative to mailing out the questionnaire directly through your own email service is to use an email marketing tool. This will also ensure recipients do not see all the other recipients’ email addresses and will allow you style the email a little more than is possible in an standard email from a mail client or service. It will also allow people to unsubscribe if they don’t want to take part.
Once you’ve sent out the link to the questionnaire, you’ll have done everything required and could leave it there and await responses, however there is a chance people will miss the initial email, or it may get lost in the usual daily emails, so I would recommend following up with some additional emails as an additional prompt. In the case of emails sent from ekmPowershop relating to ongoing research, I try to send a “chasing” email once every three days, making them more frequent towards the date I’m closing off my data collection ready for analysis, but the number and frequency of chase emails you send will depend on your own sample.
If you’re selling a physical product, you could also advertise the questionnaire on any paperwork you send to your customers alongside the items they’ve ordered, as anyone who’s received an item from you is likely to be eligible to take part. Advertising the questionnaire through social networks sites like Twitter and Facebook is also possible, though requires some attention as it is possible for people who are not actual customers to like or be active on your profile on these sites.
If you’ve marketed your questionnaire but still feel it hasn’t had enough participants, you may want to offer an incentive, such as bonus loyalty points on your site or a chance to win store credit. It’s worth noting that this does increase the chance of people responding without putting thought into their answers because they just want the incentive. It’s also worth mentioning that you should not be too disheartened by a low response rate; whilst it’s very dependent on the sample you’ve aimed the questionnaire at, the quality and length of the questionnaire, and numerous external factors, an average response rate for an external questionnaire is 10-15%.
I hope the tips in this article inspire you to take the plunge and start conducting your own customer satisfaction research. Just like most things you do when running an online shop, it can take some time and effort to get right, but the insights that come back from it can be invaluable to a growing business. If you do decide to start your own customer satisfaction study, or if you’ve done one in the past, we’d love to know any additional hints and tips you may have or hear how you found the process and what you learnt from it.
Conducting research for your shop- Q&A
We hosted a live Q&A session about this blog post. If you missed it, don’t worry – we’ve saved it for you. View the Q&A