Archive for the ‘Sampling’ Category
Friday, December 7th, 2007
Continuing what’s become something of a small business-themed week, today we look at how conducting DIY market research can be fraught with difficulty for the unacquainted and why it’s something that’s sometimes best left to the professionals.
In her article, Do-It-Yourself Market Research which recently appeared in Business Week, Karen E. Klein runs through some of the ways a start-up business can get to grips with their industry. Her suggestions include:
- Exhibiting at relevant conferences and disseminating surveys to attendees;
- Stationing one’s self on the pavement outside the such conferences, handing out surveys; or
- Hiring street teams to target particular areas;
Klein points out, however, that such techniques may end up looking like selling exercises and could even be misconstrued as harassment. Added to this, any business carrying out their own ad-hoc research would also have to be wary of the legal implications of doing so, as well as the issues involved in appropriate questionnaire design and sampling.
Given all this, and the fact that hiring staff or attending conferences could themselves be expensive ventures, the conclusion appears to be that there’s often no substitute for expertise where market research is concerned. As the article explains:
If you have the funds, an easier and more effective way to gather business data is to work with an established research company… caution [is advised] when conducting do-it-yourself studies. If a study is not well designed and executed, it can lead to poor business decisions, which are far more costly in the end.
Sound advice indeed – Especially because there are few worse things a business can do than not understand their market or their customers properly, as demonstrated by the picture below…
Posted in
Articles, Market Research, Sampling |
Comments:
None Yet »
Tuesday, February 20th, 2007
ARRANGING THE QUESTIONNAIRE LAYOUT. Questions need to be neatly formatted so that the questionnaire looks attractive and easy to complete.
Questions covering a specific subject, such as trends, suppliers or purchasers, should be grouped. Respondents should be taken gradually from simple to difficult (or delicate) questions in a logical order. Thus the first questions on the form are likely to ask easy to complete questions about purchasing behaviour. A simple routine question could then lead him into the body of the questionnaire.
By structuring the questionnaire from simple to the more difficult, the researcher is easing the respondent into his work. Once started, she will find a motivation to continue and finish. The embarrassing or controversial questions, left until the end, now stand a chance of completion.
Market research questionnaires are viewed by most as just another form and since there is no compulsion to respond there is a danger that they will end up in the waste paper basket unless the respondent has an incentive to reply. If the questionnaire is visually attractive it will help. It costs no more for the questionnaire to be printed on coloured paper. Not only does this look appealing, it helps stand out in the sea of white paper on most respondents’ desks. Cartoons can give visual relief and again add interest.
Every attempt should be made to lay the questionnaire out on one piece of paper. It looks better and avoids stapled pages. Ideally the questionnaire should be restricted to two sides of A4, i.e. A3 folded so that there are four pages in booklet form.
PILOT TESTING THE DRAFT QUESTIONNAIRE, The perfect questionnaire has yet to be written. When a draft has been prepared which appears acceptable to the researcher, it should be tried on at least two or three colleagues. Next the questionnaire should be tested in the field. Theoretically the test should be a small scale postal survey. However, the draft questionnaire is usually tested face to face amongst half a dozen target respondents who have the opportunity of saying what was in their minds as they completed each question.
Although the design of the questionnaire is the key element of a postal survey, the first thing seen by respondents is the covering letter. The aim of the cover letter is to persuade the recipient to give up time, exercise the mind and part with valuable information without any immediate or obvious reward.
The cover letter should attempt to bond the sender and respondent and so is likely to include liberal doses of the words you and we or I. The style of writing must be engaging so that the interest of the reader is held and yet it should also convey confidence that a reply will be put to good use.
The first paragraph, should explain the objectives of the research. No matter how vague, respondents need a hook or incentive for replying. This could be the promise of an improved service, increased efficiency, a more comprehensive range of products, etc. It should be explained that a reply is critical to the success of the survey and an assurance provided that the completion and response can be carried out quickly and easily. Finally, an offer of confidentiality should allay fears that there will be any future embarrassment or sales pressure.
Ninety per cent of the replies from the mail shot will be returned within two weeks of its dispatch. To increase the response a reminder could be mailed 14 days after the first shot.
For more market research articles like this visit our company website at www.b2binternational.com
Posted in
Articles, Industrial Research, International Market Research, Market Research, Quantitative Research, Sampling |
Comments:
None Yet »
Friday, March 31st, 2006
Today’s post is the third part of our ‘How Ideas Arise For Market Research’ series. The fourth and final part will be posted on Tuesday 4th April. Click here for more white papers.
The Information – what is required?
The information required may not have been presented in the most orderly fashion in the brief. A research sponsor knows what action will be taken if the outcome is positive and will have a view on the type of information that will help in that decision. All this will be shared in the brief. The researcher must now offer some order to the decision outcomes, the research objectives and any specific questions that may be asked. Examples of these three levels are as follows:
The researcher must work out what can reasonably be included in the project as an objective as well as what may have to be left out. As the researcher is thinking about the objectives, inevitably there will be consideration of the methods by which these will be achieved.
Consider the table above and think about what methods you would use for this range of outcomes, objectives and questions. Actually the researcher has a choice and could use a qualitative tool such as focus groups to get a reasonable fix on the answers. However, even with a number of focus groups it would still be a qualitative finding. You would have a good feel and understanding of all the answers to the questions but that is all it would be – a feel and understanding. If the research is being commissioned to make a decision on the launch of a new product, some quantification is required. Here the choices are two-fold; home placement tests or mall/hall tests (where the respondent is recruited from the shopping mall and brought to a nearby hall to experience the product). The arguments in favour of one approach rather than another or mixing different approaches will be made in the proposal under the “methods� section.
The accuracy – how accurate does it need to be?
When professional market researchers ask their clients how accurate any data should be, the answer is often such as “very accurate” or “as accurate as possible”. However, accuracy, at least where fieldwork is involved, has a price and as general rule, increases in accuracy not only cost more but disproportionately more .
Nor is a high level of accuracy always needed to meet the overall research objective. If a company is entering a new market, where common sense and observation tells us the market is huge in size, there may be little point in spending lots of money closely measuring its size. An approximation will do and the money saved may be better spent on some other information need. For example, a company that considers sales of $1 million per annum to be worthwhile might not care if the total market size was $100 million or $150 million (an accuracy of +/- 50%). If, however, in an advertising research study, the objective was to measure the impact of a campaign on brand awareness through comparing before and after campaign measures, the accuracy must be at least commensurate with the anticipated increase in awareness.
The required accuracy must, therefore, be linked to how the resulting data will be used – the nature of the decisions which the research will guide. Even if a precise definition of accuracy is not practically possible (this is often the case) some judgement should still be made on the reliability sought from the information. This may be as simple as a contrast between an attempt at measurement (quantitative research) compared to just description and explanation (qualitative research). Both approaches can contribute to effective marketing decisions but it is important that neither is used for the wrong application. Like information coverage, accuracy levels need to be considered before deciding on appropriate research methods.
The budget – how much have I got to play with?
What budget should be made available for the research project? The methodologically pure researcher would argue that the budget should be whatever is needed to meet the research objectives, provide the information required and to finance the methods needed to produce that information to the defined accuracy level. However, in practice, it is more a question of what funds are available or can be afforded for the project relative to other calls on business expenditure. Furthermore, even if cash is freely available, there are other considerations and especially the amount at risk in the decision which the research is to guide. If the decision entails capital expenditure of $20 million, a research budget of $50,000 may be well worth spending – if the research indicates that the planned expenditure is a poor investment, only the research cost will be lost rather than most or all of the $20 million investment. (In writing this we are reminded of the many chemical plants that have been built around the world in anticipation of growth in demand that has never materialised. We suspect that many were commissioned without any solid market research studies). However, if the business investment decision has low cost implications, the justification for carrying out the research will be less. Obviously there is no point in spending $10,000 on research to decide whether to invest in a project entailing only this level of expenditure.
The only qualification we would make to this bland statement is regarding research which contributes to a series of future decisions. We recently carried out a research project examining the effectiveness of a planned promotional campaign for an industrial gas manufacturer. The cost of the research was $30,000 and the campaign itself was only $200,000. However, the gas manufacturer runs many such campaigns across its divisions and the learning about what makes its advertising more effective sharpened all its campaigns and will do so for some years to come. The long time pay-back will be considerable.
Posted in
Articles, Market Research, Quantitative Research, Sampling |
Comments:
None Yet »
Thursday, March 30th, 2006
Almost all valuable market research is conducted using samples, which makes the area of sampling and statistics key to the market research function.
If business to business marketers experience twinges of inadequacy when compared to their counterparts in consumer fields, sampling is often the reason. Consumer marketing men trot out jargon about quotas, stratification, probabilities and the like which is mumbo jumbo to 99 per cent of industrial marketers.
Companies commissioning market research very often feel insecure with industrial samples as it is difficult (except with large homogeneous universes) to know the accuracy of the findings. We have devised different sampling tools that you can request from our main website at http://www.b2binternational.com/tools3.html. However we have now placed these tools online that is specific to our b2b blog
This area of the website allows you to determine the sample required in your particular market, work out the accuracy of response rate once you have carried out a market research survey and determine the growth rate of your market or company.
We now have some new content available to you via our blog. You may, or may not, have noticed the ‘useful interactive tools’ area that has appeared over the past few days.
If you would like more information on sampling in b2b markets then read the following article
http://www.b2binternational.com/article15.html
or Sampling & Statistics by Paul Hague and Paul Harris http://www.b2binternational.com/pub05.html
This area of the website offers you access to:
B2B Sample Size Calculator
This sample size calculator is to be used when there are less than 100,000 companies in a market. Make an estimate as to the number of companies in the market you want to survey (we call this the “population�).
Growth Calculator
What is the growth rate of the market you operate in?
What is the growth rate of the company you work for?
Use the compound growth rate calculator to determine the different growth rates.
Ready Reckoner
The sample size reckoner is used to find out how accurate a specific response rate is once you have carried out your survey (at 95% confidence limits).
Posted in
B2B News, Market Research, Sampling |
Comments:
None Yet »
|
