Tài liệu Customer satisfaction measurement docx

Surveys are an excellent way to find out how your customers feel about a new product, service, location, store policy or virtually anything that's important to your business?. Better cus

baldrigeplus adding value to excellence in your organisation

Exhibit

The mission of baldrigeplus.com is to add value to your investments in excellence

by publishing and supporting a set of analytical tools that make the Baldrige

system accessible, understandable and applicable to your organisation

COPYRIGHT

1999 Macpherson Publishing, PO Box 270, Alexandra, New Zealand

Telephone +64 03 448 9524, Fax +64 03 448 7986, Email macalex1@xtra.co.nz

This document is not to be reproduced without the prior consent of the copyright holder

MISSION

Customer satisfaction

measurement

Customer satisfaction measurement

A primer

Introduction

Baldrigeplus.com exhibits are single-topic publications designed to illustrate and

explain key Baldrige concepts Used with the associated workshops, worksheets, case studies and Baldrige-sourced material, exhibits are intended to add value to your performance excellence activities

Return to the Welcome page for links to the other resources

Do you know if your customers are satisfied?

Answer this questionnaire to see if there is anything you could be doing to know your customers better

1 Does your small business have a strong customer-retention program in place?

2 Do you receive feedback from customers on a regular basis?

3 Has your small business made changes or implemented ideas at the request of customers?

4 Do you call on customers just to see how your company’s products and

services are working?

5 Do you have an idea about what products or services your customers plan to purchase in the next year?

6 Do you send out surveys on a regular basis to solicit input from your

customers?

7 Are your customers happy and can you quantify their satisfaction level?

8 Do you know who your best and most loyal customers are?

Source - smallbiz.symantec.com

Your customers have the answers

Knowing your customers is key to growing your business Customer satisfaction is essential for long term success You want your customers to be happy with the

products and services you provide and you want them to stick with you as their needs grow and change

Your customers have the answers, but getting them to tell you what’s good about your business, and where you need improvement is the challenge Surveys are an

excellent way to find out how your customers feel about a new product, service, location, store policy or virtually anything that's important to your business A survey will tell you what your customers expect of you and your company, and clarify how well you are performing in their eyes

A customer satisfaction survey is one way to gather vital information There are a number of ways to get copies of the survey to your customers Copies can be included with orders, mailed directly to customers or done via a phone survey Not all your customers will return their surveys, but those who do will make it worth your while

If executed properly, you can achieve impressive results without spending a lot of money Customer service experts estimate it can cost between $3,000 - $5,000 to sample a representative segment of your customers

Better customer surveys

Surveys are an excellent way to find out what your customers think about your

products and/or services, telling you what your customers expect and what they are and aren't getting But the secret to any good survey is the questions! If you don't ask the right questions, you are just wasting your time and that of your customers

Here are ten tips for better customer questionnaires

1 Start with a clear objective

What are you trying to learn? Understanding why you are surveying your customers is an essential first step for better results. A strong survey has a clear

goal or focus (to find out how customers feel about a new location or store layout; get response to a new product or service; learn why once loyal

customers are now shopping elsewhere) Be clear about why you're surveying your customers, and you will get results that will help you make your business more effective

2 Be honest

Tell your customers why you are asking them questions and why their

feedback is important Don’t dress-up your survey as something else

3 Keep your survey short and simple

You are asking customers for their valuable time so don't waste it Keep

questions direct, simple and brief Long, wordy multi-part questions should be avoided The longer the survey, the less likely you are to get responses Keep the questions short Scaled questions (strongly agree, agree, slightly agree, don't care) or yes/no questions are the easiest to answer An example of a good question: How well do you feel your accountant meets your tax needs? Ideally, your survey should consist of 10-12 such questions, on one page

Consider using a response scale (called a Lickert scale) with an even number

of options, usually four or six, to force a positive or negative response:

m Strongly agree m Agree m Disagree m Strongly disagree

That way, your customers can’t sit on the fence by marking a ‘neither agree nor disagree’ option

4 Ask questions that are important to your customers, Make sure the questions

you ask directly affect the customer If you ask them questions they don't care about you are less likely to get useful information Customers aren't concerned with issues that do not pertain to them (hiring, promotion or store policies, outreach programs) Make sure that each question is important to your

customers' needs, not your internal management, and remember that

successful surveys are fun to fill out Don't waste your customers' time

5 Give them an incentive for responding

It might be as simple as a sentence at the top of the survey indicating that you are using the input to evaluate current policies/products and create new ones

− customers like to feel like they can have an impact on your company Ron

Zemke, author of Coaching Knock Your Socks Off Service suggests including

discount coupons for certain products or services with the survey, particularly

if you're using your survey to prospect for new customers

6 Encourage opinions

If you are seeking actionable data, and that’s surely the key purpose of your survey, leave space at the end for comments – or better still, add a comments line after every question, encouraging respondents to express their opinions Many service businesses, for example, have been successful with questions such as, "If this were your business, what would you do differently?" If

respondents feel strongly about the issue, pro or con, there's an excellent chance they'll answer it Look at IBM’s customer survey approach (see the exhibit Customer relationship management at IBM Rochester) If customers

have the option of marking your products or service unsatisfactory, or worse, always ask ‘what could we do to improve your satisfaction …’

7 Use a confidential self-mailer to generate higher response rates

The easier it is for people to respond, the higher the response rate will be Self-mailers are highly recommended because they require less paper and

postage Next best is including an addressed pre-stamped envelope ( It's presumptuous to expect respondents to use their own postage.) Also, stressing

the survey's confidentiality increases your chances of getting honest answers

8 Test your survey before mailing it

Andy Mosko, managing principal of Organizational Research Forum, Inc., Vernon Hills, IL, a company that specializes in designing customer surveys,

advises testing a survey before mailing it Try it out on a few good customers,

he says You'll be pleasantly surprised to discover it can be improved

9 Focus your surveys on your most important customers

These are the people whose opinions you value most If the lion's share of your business comes from only 8 or 10 customers (as may be the case with a small service business), design your survey with their needs in mind (and take

a look at the Solectron on customer relationship management in the

baldrigeplus.com case study collection and the related exhibits)

10 Preview your survey with postcard arousing curiosity

A brief attention-commanding announcement should be delivered a week before the survey is officially sent It's an opportunity to highlight the benefits (premium, discount or special money-saving coupon) of filling out the survey and sending it back promptly

Sources - Symantec’s small business resources,

at smallbiz.symantec.com, americanexpress.com

How Big Is Big Enough?

From an article by Tom McGoldrick, David Hyatt, and Lori Laflin, of American

Demographics/Marketing Tools, a unit of Cowles Business Media

When it comes to sampling, size matters

One of the most common sampling questions is, ‘How many people do I need to survey to have valid data?’ There's only one answer: ‘It depends,’ … on what you want to do with the data you collect

The goal of surveying is to find out how everyone who matters feels about a topic, but most of the time it's impractical to ask them all, so we ask a small number, a sample, and use the results to make assumptions about what all of them think We want to know how the population feels, but we are only asking a slice of the population

One of the first things to do before deciding on a sample is to think about the

characteristics of the population that will be important to you This ensures that when you're drawing your sample, you will represent those groups appropriately A sample

is only valid to the extent that it mirrors the population of interest

There are any number of ways to select a sample, most fall into one of two groups: probability and non-probability

1 Probability sampling is a type of sampling where every member of the

population has a known (usually random, and equal) probability of being sampled

2 Non-probability sampling is when certain members of the population are

chosen because of a judgment on the characteristics of the population and the needs of the survey Non-probability samples are created because the units appear representative or because they can be conveniently assembled; hard-to-find groups, surveys of specific groups, and surveys in pilot work Non-probability is not random because not all members have the same chance of being drawn for the survey Some have no chance at all

Types of probability sampling

Simple random sampling where each member of the population has the same chance

of being selected In order to achieve a true simple random sample, you will need to develop a list that includes all of the members of the population, then randomly pull names from the list The biggest advantage to a simple random sample is that it

provides an unbiased sample fairly easily The biggest downside is that it may not capture all elements of the population that are of interest

Stratified random sampling where the population has been classified into strata, or

subgroups, and then a random sample is pulled from each subgroup A common question is, what strata do you use? The answer is to develop strata based on

characteristics that should have an effect on the results, think about what groups you will want to compare when analyzing the data

Systematic sampling where names are taken systematically from a population using a

ratio of population to desired sample For example, if you had a random list of 3,000 customers and wanted to send surveys to 500 of them, you would select every 6th

person and send them a survey You need to be sure the population is not ordered to begin with, or that there is no naturally reoccurring pattern

Cluster sampling The final type of probability sampling, cluster sampling, uses

naturally occurring clusters such as business units, schools, hospitals, or city blocks, randomly selects a subset of them, and then samples everyone in the chosen cluster Cluster sampling is traditionally only used in large-scale surveys It is different from stratified sampling in that you start with naturally occurring clusters; you don't

develop them

Types of non-probability sampling

Convenience sampling A convenience sample is a group of people whom you can

conveniently locate and survey The sample may not generalize to the target

population

Snowball sampling requires you to identify members of the desired population who

are then asked to nominate other members, so that the sample snowballs in size For example, if you were interested in surveying a certain type of doctor, you might ask the known ones to nominate others, who would in turn be asked to nominate others, and so on

Quota sampling divides the population into subgroups, which are then sampled in

proportion to their presence in the population The key is estimating the percentage in each subgroup The difference between this and stratified sampling is that in quota sampling, you don't care how you got the people as long as the quota groups are filled appropriately

Sampling errors

No matter how good samples are, bias or error will slip in There are several types of error that can occur when you only sample part of the population:

1 Non-sampling errors − one of the biggest sources of error − consist of

definitions, or inclusion/exclusion criteria issues; and non-response issues

- The definitions you work with in describing your population of interest can greatly affect the accuracy of your sample

- Non-response error is an important consideration In survey research, it is necessary to assume that the people who respond are not too different from those who choose not to The question is whether that is a valid assumption or not Because we don't know, we usually have to qualify our interpretations

2 Sampling errors can enter into the sampling process as well Names may be

duplicated on a list, increasing the likelihood that someone will be selected, or the list may not be complete

Sample size

There are three ways to determine the sample size:

1 Copy what others do, or just doing what you've always done (the easiest)

2 Pick what is convenient, or what you can afford (not the best way but realistic)

3 Guess the level of accuracy you feel is right for the issues you are

investigating

Sampling tables will save you from the detailed calculations, but you will need to consider: your desired level of confidence, how much sampling error can be

tolerated, population size if the population is small, how varied the population is with respect to the characteristics of interest, and the smallest subgroup within the sample for which estimates are needed

Level of confidence refers to how likely the sample estimate will contain the value of

the population Recall that we are trying to estimate what the population is like by selecting a sample from it We need a confidence level because we assume there will be some error Common confidence levels are 95% and 99%, which means that

95 or 99 times out of 100, the sample we pull will be within the specified error range

of the population Sampling error refers to how accurate our estimates of the

population need to be

It is important to remember that if you are interested in comparing groups, you need to consider the number in each group, not the size of the total population

It is also very important to note that all discussions of sample sizes are talking about the number of surveys that are actually completed and returned In order to achieve the proper sample size, you'll have to consider your expected response rate and work backward, calculating the number of people to whom you will need to send your survey For example, if you expect to receive a 25% response to your survey, and you know that you need to receive 500 surveys to reach the appropriate levels of confidence, you will need to send out 2,000 surveys to your target population

How to…measure customer satisfaction

From an article by Entrepreneurial Edge Online, a service of The Edward Lowe Foundation

The costs of poor service

Businesses survive because they have customers that are willing to buy their product

or service However, many times businesses fail to “check in” with their customers to determine whether they are happy or not and what it will take to make or keep them happy

According to the US Consumer Affairs Department, it costs five times more to gain a new customer than to retain an existing one Other studies have reported that with just a 5% increase in customer retention a firm can raise its profitability by 25% and in some cases as much as 85% Similar studies also show the longer a company keeps a customer, the more money it will make What happens is that consumers spend slowly at first, but with succeeding years of good experiences, they will spend

increasingly more

The Profit Impact of Market Strategy (PIMS) data base shows that firms perceived as having better customer service can charge more for their products and services and still have higher market shares and returns on sales than their competitors

TARP, a management consulting research company, reports only 2-4% of dissatisfied customers ever complain to a business regarding a poor experience The others just leave and potentially do business with competitors Of customers that leave in a given year, 68% do because of supplier indifference or poor attitude A General Electric study found that word-of-mouth has a significant influence on consumer decisions, twice as much as advertising Negative word-of-mouth can be really dangerous since dissatisfied customers are usually more vocal than satisfied customers Depending on the industry and the nature of the bad experience, dissatisfied customers will

complain to 10 to 20 friends and acquaintances, which is three times more than those with good experiences Furthermore, this negative information is influential, and consumers generally place significant weight on it when making a decision

If that isn't reason enough, fierce competition is requiring more and more innovations

to differentiate firms from one another With technology available to virtually

everyone today, the traditional feature and cost advantages are no longer relevant Still, product and service quality provide an enormous opportunity to distinguish a firm from the rest

According to futurist and corporate advisor, Faith Popcorn, a new type of customer is emerging in the '90s They are ‘vigilante consumers’ — a new generation of super consumers that are smart, discriminating and vocal

They demand value for their money and expect the companies that they buy from to

be responsible and accountable When companies don't respond, vigilante consumers will make sure that they will tell anyone who will listen why they shouldn't do

business with those companies Satisfying these smarter consumers just makes good business sense

The customer service payoff

Customers are your best source of business information — whether it's to improve an existing product or service or whether you're planning to launch something new Customer data allows you open up the lines of communication; align resources; and make changes or launch products more quickly Talking to customers directly

increases the odds for achieving success; mistake-proofs decisions and provides a focus on what really matters When you routinely ask your customers for feedback and involve them in your business, they, in turn, become committed to the success of your business

Watch out for …

Even the best intentions are subject to problems along the way Temptations to avoid are:

1 Complacency — Customer surveying should be continuous Asking

occasionally is not enough Change is certain, and priorities shift The most successful companies are those that can detect and respond to customer changes quickly

2 Analysis paralysis — When you get your data, don't analyze it to death Many

corporations have departments full of statisticians to determine the reliability and validity of their data; but they never do anything with it In most cases the lessons will be obvious Get started immediately

3 Doing nothing with the feedback — Do nothing with their suggestions, and

your customers will lose interest Show them that you appreciate their input as well as communicate to them what has changed as a result of their input

4 Failing to listen to your experts — Employees are a valuable source of

customer information They deal with customers constantly and often have first hand knowledge on what the customers' hot buttons are Too often employers ignore this valuable resource

5 De-motivator for employees — Customer feedback should not be used to

punish employees Instead, use it to detect areas for improvement If customers discover that their input is used to discipline employees, they may stop

providing constructive feedback altogether.

Assessing satisfaction

Sources of customer data

Do you know what influences customer buying decisions, how important each

influencer is, or how they measure up to your competition in the areas most important

to your customers?

Your own organization — Uncover potential areas of customer discontent by

reviewing operational data Check on the status of backlogs or stock-outs If these are significant you may have customers that are not happy with your delivery cycle time

1 Review your ‘returns and allowances.’ If they are high, then your customers

are sending you a strong message that they were not happy with the product they purchased Look also at internal reject or yield rates If rejects are high or yields low, you can bet that some bad product is leaking out to your customer

2 Your employees are a valuable source of information on customer

satisfaction They interact with customers constantly and probably know a

great deal about your customers' likes and dislikes If you are a one-person organization, then you are the one dealing with customers You know what's going well, what needs fine-tuning, and what needs a major overhaul You should begin your search for customer data in-house Most likely you'll

uncover some things that you can fix immediately, which will make your customers happy and get you started on the right track

Customers — Start with a review of customer complaints and inquiries If you don't

have a systematic way of collecting these, you should develop one Both are good indicators of opportunity areas Don't limit yourself to just complaints and inquiries — only 2-4% of dissatisfied customers ever complain If you're only looking at

complaints, you're missing the other 96-98% who have problems with you

Surveys and focus groups are two popular methods for gathering information on customer needs Surveys are written assessments given to individual customers; focus groups are oral assessments administered to groups of customers Both must have clear and specific A broad questionnaire or focus group session provides you with a lot of information, but it's usually too general to do anything with Objectives must be clear and questions specific if they are to provide results that can be acted upon

1 Surveys are relatively easy and cheap to run, and can reach large numbers of customers, but the information can sometimes be limited; it’s a one-way exchange

2 Focus groups take more time and effort, tend to cost more, and don’t reach as many customers but they’re interactive and may produce clearer feedback

3 The best results are found when combinations of both techniques are used to identify customer requirements and expectations

The best customer data

More is not necessarily better, but getting the right kind of data is critical The key characteristics of good customer data are:

• Ongoing — Change is going to happen Your customers may change; their needs may change; the environment may change; and most certainly you will change As you improve, your customers expectations will likely rise, too In order to respond to these changing needs, you'll need to constantly assess your customers

• Specific — In order to make the kind of improvements your customers will appreciate, you'll need to have specific feedback While general inputs may give you an overall tone of the customer, you can only respond to specific feedback

• Timely — If you're working with old data, it may be obsolete and no longer relevant

• Focused — Organizations have limited resources While the problems can be overwhelming, you can realistically work on just a few If you try to do too much, you might not do anything well

• Weighted (according to importance) — This will help to narrow the list of opportunities to just the few on which you should concentrate You can rate the relative importance, but it will be much better if your customer does it

• Competitive comparison — You should always know where you stand in comparison to your competition If your customers are willing to provide you with that feedback, take it And you won't need a separate survey Look at the sample survey at the end of the module to find out how you can incorporate this information in the same survey

Surveying

The 5 steps of a successful survey are:

1 Decide on your objectives

2 Decide who to ask

3 Develop the survey

4 Analyze the results

5 Communicate the results

1 Decide on your objectives

What do you want to know from the survey? Be specific Your objectives will form the basis from which your survey questions will be developed Limit your objectives to just a few If you try to include too much, you will make the survey too long

(customers may not complete it), and you may uncover more than you can handle (you can't respond to it

2 Decide who to ask

First and foremost, know who your customers are and which are appropriate to

survey! As obvious as this sounds, it is not so obvious in practice If your market is large, you may have different segments of customers Or depending upon the industry you may have different levels of customers Nevertheless, your survey objective(s) will usually determine which customers you should survey Also, give some thought

to the amount of customers you want to survey Do you have a couple of key

accounts? Maybe you want to survey each of them If you have several customers, you may have to select a portion of those to survey Also, you may want to hear from different individuals at the same customer site Often, it is enlightening to get

feedback from individuals other than your direct contact They may have problems that you never hear about because it never gets to your contact and never is reported

to you.

3 Develop the survey

Develop the questions, which should relate to your objectives and be relevant to

what you want to measure

1 Ask concise questions Long questions are difficult to read and discourage

customers from completing it Eliminate unnecessary words Instead of: Did

the clerk provide you with personal service when you asked to be helped with

your selection? Ask: Was the clerk personable?

2 Avoid vague and ambiguous questions The customer should be able to

understand exactly what you are asking Instead of: Was the bank transaction

by the teller acceptable? Ask: Did the transaction occur quickly enough? (and)

Did the teller talk to you in a pleasant way?