STEP 1: The why, what and how
The customer
The customer still tends to be seen as an external party, to be treated, indeed, with some caution! What a risk we take if we telephone a customer to enquire if everything is all right – we might get a complaint! No news is always good news! If this example sounds familiar it is probably because your company is ‘chairman-focussed’ rather than ‘customer-focussed’. Have a look at your organisation chart. Does it start with the chairman or managing director at the top and then branch downwards showing the different departments – sales, finance, production, despatch, etc.? If so, where does ‘the customer’ appear on your organisation chart? Organisations such as these, often, are not customer focussed as they would like to believe.
As a customer, you know when you are dealing with a company like this. You telephone to place an order and are put through to the sales order clerk. Having placed the order, you then ask a question about the balance on your account. “Hold on”, says the sales order clerk, “I’ll put you through to the accounts department”. Pause, music, clicks and possible delays because the relevant account clerk is on another call. Eventually, you are put through and get the required information. You then say, “Oh, by the way, can you confirm that the order I placed last week will be delivered today, as promised?”. The accounts clerk then has to decide whether this is a query for the sales order department or the despatch department. Either way, you are treated to more delays and hassle.
Processes cut across traditional, departmental boundaries. In chairman-focussed organisations, no one feels a sense of ‘ownership’ of the customer.
The customer-focussed company organises itself around its processes. It understands that processes need ownership. It recognises that the sales order clerk can be trained to read a sales ledger account and check despatch records, and much else besides. With the advent of modern, object-oriented software systems there is now no excuse for companies not to be customer-focussed.
How would you describe your company – chairman or customer focussed? If more like the former than the latter, then a programme of process analysis and improvement should be undertaken as a matter of urgency.
The purpose of a survey
The purpose of a survey is to gain knowledge. Knowledge is the light that illuminates the road ahead – the more clearly you can see that road, the more certain, faster and safer your journey will be. This metaphor may seem self-evident at first sight but stop for a moment to consider the nature of knowledge.
There is a fundamental problem with knowledge – knowledge is more than just facts! Perceptions often count for more than facts. You may know that you have done a good job for a customer but if that customer thinks you have done a bad job, which is going to determine whether the customer buys from you again – the fact or the perception? The more you know what people are thinking, the more successful your company will be.
There is only one way to find out what people are thinking – ask them! Some companies do not find this easy. Historically, management used to assume that it knew best and to ask the views of the staff, let alone the customers, was considered to be a sign of weakness. Thankfully, times are changing and management has come to appreciate that the people at the ‘coal face’ often have the best ideas. Management now recognises that the knowledge of the business (its ‘Intellectual Capital’) is held diversely through all the people working in the company. It is quite common for companies to consult with their staff and encourage feedback to improve the performance of the company. Conducting internal surveys of staff is an effective way of discovering how people feel about working for the company. Unfortunately, this revolution within corporate culture that has resulted in management and staff working together in a spirit of partnership has not yet extended to embrace the customer to the same extent. Those more enlightened companies that treat their customers as ‘part of the family’ are reaping real benefits. Their customers are proving to be their best salesmen!
The customer satisfaction survey is a two-way communication. It asks your customers for information but it also makes a statement to them. The obvious statement is that you value them as a customer but you are also making an implied statement that you will take action on the responses you receive. This action has to be real and prompt, and communicated back to your customers, otherwise you will lose credibility and risk making relationships worse than they were before you conducted the survey.
Customer satisfaction surveys should not be considered as a substitute for internal monitoring and control. You should not be relying on your customers to tell you what is wrong with your company. You should know this already. Your internal systems should be telling you the levels of late despatches, product failures and returns, time taken to process orders, telephone call backlogs, complaints, etc. In the best-run companies, customer satisfaction surveys should be a regular and normal part of monitoring procedures and updating performance indicators. Performance indicators are more than just financial numbers, ratios and statistics. They are also about the quality of products, customer service, delivery, management, staff, administration, culture and care.
The purpose of a customer satisfaction survey should be to test your own beliefs about the quality of your customer service with the perceptions held by your customers. If your company is truly customer-focussed, the satisfaction survey should not produce any major surprises.
Survey styles
There are two styles of survey – qualitative surveys and quantitative surveys. In most surveys these characteristics will overlap but one will usually predominate over the other. It is this predominance that determines the methodology for conducting the survey.
The qualitative survey is used where the sample population is small and the subject area is specialised or subjective. Consider, for example, a customer satisfaction survey to be conducted for a company that sells high value, scientific equipment to a small number of customers in different countries. Such a survey would best be done by telephone discussion or even personal visits, simply because there is likely to be wide variability in the factors that influence opinion. The way in which the company manages each project is likely to be significantly different from one customer to another. Cultural variations from one country to another can influence customer perceptions. Qualitative surveys can be very expensive to conduct because of the time they take to perform and analyse.
The quantitative survey is used where the sample population is high and there is reasonable commonality in the products or services. An electrical wholesaler, for example, with 2,000 customers, whilst selling several thousand different items, is likely to process orders, despatch goods and deal with complaints in the same way for most customers. For such a company, a quantitative survey is more appropriate because one can build a questionnaire with many standard questions that will apply to most customers.
Survey types
There are two types of survey – controlled surveys and speculative surveys.
Controlled surveys are those that are sent to pre-selected panels of customers both in terms of numbers and selection criteria. They are delivered to respondents within invitation letters or e-mails or conducted by telephone or personal interviews. Controlled surveys address specific issues and are terminated after a period of time or when the required number of responses has been received. Controlled surveys may be repeated periodically to establish trends and to generate data for performance indicators.
Speculative surveys are those for which the target population is unknown. Such surveys typically are embedded into websites or left, for example, as paper questionnaires in hotel bedrooms. The sponsor of the survey has no influence over who will respond, or how many replies will be received or the timescale over which they will be received. Speculative surveys tend to be continuous surveys with no set time limits.
Pros and cons of survey delivery methods
Mail is slow, expensive, labour intensive and error prone because all the data has to be scanned or keyed manually. It is difficult to follow up on incomplete or illegible responses.
Telephone gives quick results. Contact with the respondent gives the opportunity to follow up points but it can be difficult to consolidate results. Not all people have telephones or are in when called.
Personal visits enable the interviewer to build a relationship with the respondent and to follow up points but the process is slow and expensive.
Online gives you instant results. Conditional logic can be built into the questionnaire to make it more efficient for the respondent. Costs are low. Non-replies can be followed up automatically. The system ensures that data entry is complete. Virtually all businesses have e-mail addresses. Most consumers have e-mail addresses or public access to the Internet. Companies, however, have been very slow to recognise the value of holding e-mail addresses for their customers. Remember to obtain and record your customers’ permission to contact them by e-mail.
Conducting a survey
There are four principal stages to conducting a survey:
- Setting the objectives
- Building the questionnaire
- Selecting the target population
- Processing the results.
STEP 2: Setting the objectives
The objectives of the survey must be clearly stated, understood and agreed. Firstly, identify the problem you are trying to solve, and then decide what information you need to solve it. Is your problem, for example, an increase in the level of complaints about product quality, delivery, after-sales service or a decline in sales? The nature of the problem will influence the structure of the questionnaire and the choice of customers to whom you send the survey.
Consider contacting a small sample of customers with whom you know you have a good relationship or who are typical of the problem you want to solve. Telephone or, ideally, visit them and discuss your concerns, get their perspective on the problem and invite them to suggest the issues that you should be addressing in your questionnaire. Customers will see this as a positive step on your part to improve customer service and will be willing to help you.
If you have a really serious problem or are having difficulty in defining the problem, then consider forming a focus group with a small number (5 to 10) customers. Get them together to help you to understand the issues and formulate your survey objectives.
If you are not aware of any specific problems but want to introduce surveys as a regular means of monitoring perceptions and performance, then define the information you need from the survey in a way that will enable you to establish meaningful performance indicators. You will then be able to monitor trends and measure the performance of your business as you conduct successive surveys.
Wherever possible offer an incentive for completing the survey. Offering two incentives is a good idea. The first incentive could be an entry into a prize draw (magnum of champagne or a case of wine, for example); the second could be a voucher for a specified sum or discount off the next purchase. If you are conducting the survey online you should be able to get the system to manage these incentives for you. Do remember to advise your customers of the winner of the prize draw (with the prior permission of the winner) – it gives you another reason to contact your customers!
STEP 3: Building the questionnaire
Begin by explaining the purpose of the survey. Be sure to stress that you care about your customers and are seeking to give them an even better service (or solve whatever the specific problem may be). Give them an indication of how long it will take to complete the questionnaire. Describe the incentive you are going to offer them. Tell them that you will give them a summary of the results. Thank them in advance. Provide a contact name and telephone number they can contact with any queries.
When building your questionnaire, try to use a variety of question types. This will help to make the survey visually more interesting. The standard types of question are:
Single choice – such as the ‘Overall, please rate your level of satisfaction with our Customer Service’ question below. Here, only one answer is logically possible.
Multiple choice – such as the ‘For what reasons have you contacted Customer Service?’ question below. Here, more than one answer is logically possible. Always provide an ‘Other’ option, with a text box for the respondent to give a description.
Ranking – the respondent is asked to rank responses in the order 1,2, 3 up to a specified number.
Matrix – such as the ‘Please rate your level of agreement…’ question below. Avoid having more than 6 or 7 statements in a matrix question – otherwise they become difficult and tiring to answer. It is better to have several, small matrix questions than one long one.
Yes/No – designed for closed questions such as ‘May we contact you again in about 6 months’ time…?’ as shown below. Note the use of the word ‘may’. This is a soft and courteous way of asking a question and is much more likely to get a ‘yes’ than the ‘can’ word!
Data request – used for gathering a specific item of information such as a telephone number, postcode or an e-mail address.
Be careful to avoid bias in your questions. Don’t lead the respondent to the answer you would like to hear. ‘How do you rate the improvement in our customer service this year?’ presupposes that there has been some improvement. ‘How do you rate our customer service this year compared with last year?’ makes no such assumption and is likely to get more accurate responses.
Be careful with the use of open questions. If you ask ‘What do you like best (or least) about our customer service?’ you are less likely to get useful replies than if you ask the respondent to express a view over a number of specific aspects of customer service.
Start with general questions and become progressively more specific. By adopting this approach you will be able to test the consistency of the customer’s responses when you start asking more detailed questions later.
Your detailed questions should cover the three main elements of customer service: reason for contact, speed and result of response, competence and professionalism. You should also invite the customers to make their own suggestion for improvement.
Design your questions so that the respondent only has to point and click, or tick a box on a form. For example, if you want to ask ‘When did you last contact Customer Services Department?’ don’t give a free format text box for the answer. If you do, you will get the answer in many different formats (“Last week”, “22nd March”, “Recently”, “12/4/02”, for example) and it will be impossible for you to collate the data. It is far better to provide a series of date ranges and ask the respondent to tick the one that applies. Here is an example of a single choice, date range question:
Choose date ranges that are appropriate to your business. If, for example, you reorganised your customer service department on a specific date, or appointed a new manager, you may wish to discover if the customer contacted the department before or after that date, to see what differences there are in answers to subsequent questions about the quality of the service received. Note in the example above that the date ranges get progressively longer – this recognises that memory becomes less precise over time. Note also the need to give an option of ‘Never’, to cover the situation where the question is irrelevant – this can apply to other questions, where ‘not applicable’ or ‘no opinion’ (or similar wording) must be included as an option. This is particularly important with mandatory questions in online surveys.
Here is an example of a multiple-choice question. Note that ‘Other’ is offered as an option and can be combined with one or more of the other choices. The online survey system from which this screen-shot is taken is called surveySwift and displays a text field only if ‘Other’ is selected.
Here is an example of a matrix question. Minimise the use of matrix questions. These are questions that contain a number of statements and ask the respondent to express a view across a range from ‘very satisfied’ to ‘very dissatisfied’, or similar words. The matrix question adds variety but it takes longer to complete.
Here is an example of a yes/no question. In an online survey you have the option to route the customer to different parts of the questionnaire, depending on the response. Indeed, with surveySwift you have this facility to create logical branching for any answer to any line within any question.
Be sure to test your questionnaire thoroughly before sending out the survey. Check the logic and spelling and make sure that your questions will elicit the answers you need to satisfy the objectives of the survey. If you contacted a sample of customers when defining the objectives of the survey, ask them to test the questionnaire for you.
STEP 4: Selecting the target population
You could send your survey to all your customers but it might be better to test the survey first on a representative sample – say, 10%. This will give you the opportunity to discover if you have missed any key questions and adjust the questionnaire before you send it to all your customers.
It is also important to segment your customers by how active they are. This will enable you to look to see if there is any correlation between changes in buying patterns and perceptions of customer service. For example, are those customers whose orders have been decreasing over the last 6 months reporting higher than average dissatisfaction with customer service?
STEP 5: Deciding on the method of delivery
If you are selling mainly to businesses, then the ideal delivery method is an online survey embedded in an invitation e-mail. If your customers include consumers, it may be possible for you to obtain e-mail addresses at point-of-ordering or point-of-sale. If you are a retail outlet with a customer service department then you will probably have to use a paper-based questionnaire. You may wish to provide a prepaid envelope if the questionnaire is long. If the questionnaire is short, the customer can be invited to hand it into the customer service assistant or to place it in a sealed collection box.
There is now an alternative to paper based questionnaires in retail or other public locations. Kiosks can be set up at exit or other strategic points with touch-sensitive screens that enable customers to complete the questionnaire online. Customers find kiosks more fun to use and the data is consolidated and available to view instantly.
Some companies find that they get better results if they use an external agency to conduct the survey, thereby assuring the customers that their responses will be anonymous. If you feel that customer relationships are generally poor, then the anonymous survey may be preferable. If you know that your relationships are generally good and the main reason for conducting the survey is to monitor and sustain service quality, then anonymity may not be an issue – why not ask a few customers for their views on this.
STEP 6: Using the results
Don’t be afraid to publish the results of the survey within your company. In a truly customer-focussed business, everybody should be customer facing in his or her thinking, attitudes and actions.
Use the data to establish yardsticks of performance. Measure the trends by repeating the survey periodically. Set up cross-departmental (process focussed) teams to decide what lessons can be learned from each survey and take appropriate action. Implement this promptly. Report back to your customers, describing the actions you are taking and the benefits they will receive.