Follow these top tips for creating great survey research. For more info contact [email protected]
For more detailed resources for researchers click here
1. Clearly define the business objectives
Be clear about what you want the research to achieve for the business – beware of pinning too many badges to one piece of research. Great surveys are tightly focussed around one business strategy, but can be used in many ways by different departments. So you might have great client research which you communicate to the Board, feed back to your customer facing staff and include in your marketing materials.
2. Clearly define the research objectives
List the key pieces of information you will need to meet the business objectives. Ideally you want to read the minds of your clients! In reality you will have to piece together a version of the truth by asking about facts, figures, attitudes and preferences.
It’s all too easy to have research ‘hijacked’ by department heads from all over the business! Just one or two questions squeezed in as a favour to one can soon lead to a whole raft of questions which try to please everyone but ultimately achieve little.
3. Do you need the research?
When you have clearly defined your information needs, find out whether it already exists – look at previous surveys conducted by your organisation, look at market sector reports (The British Library is an excellent source) and interrogate your organisation’s database. Depending on what you find, you may need to repeat a survey conducted several years ago, design completely new research or find that you don’t need it.
4. Secure backing from the Board and a budget
Great research requires planning, support and a realistic budget. Present a compelling case on the business benefits of the research, involve the Board and seek their backing. Without top-level backing, research projects can flounder for lack of money and people to help conduct the research and implement the recommended actions after it’s done.
In addition to the Board, you will need support from relevant departments – HR for an employee study; Marketing, Communications, Sales and Operations for a customer satisfaction study and so on. Get them involved as early as possible.
5. Decide whether to outsource some or all of the research
Do you have the knowledge and resources to conduct all of the research? The key stages are: research and survey design, data collection/interviews, data processing, analysis & reporting, and communication/implementation.
You may be able to do some, all or none of this…getting a great research supplier involved is a good idea as they will have the infrastructure to easily collect and process the data. A research supplier will also know how to squeeze the most information out of the data.
6. Define who, what, where, when and how
These are the key questions for designing research: who will you need to survey and where will you get the list of clients? What will you survey them about? When will you survey them (for example, straight after a transaction or at regular intervals)? How will you survey them – by telephone, face to face, using paper, a handheld electronic device, mobile phone, online survey etc? There are so many options today thanks to technological advances.
The other ‘how’ is how you will survey – qualitative focus groups, one-to-one interviews, online qualitative discussions, quantitative (box-ticking) survey etc.
The best method(s) to use will be determined by the type of data you need, the types of customers you have, how they like to communicate; to whom you will be reporting back and your budget.
7. Know what you’re measuring and why
When designing your questionnaire, take a step back to understand what type of question you are asking, how you will collect the answer and analyse it.
The key types of data are Numerical, Categorical, Attitudinal and Hybrid.
1. Numerical: e.g. ‘how many cars do you personally own?’ You would collect an actual number. When reporting, you might want to report the average number of cars owned.
2. Categorical: e.g. ‘do you currently hold a valid UK drivers licence? Yes / No’.
3. Attitudinal: e.g. ‘how strongly do you agree or disagree with the statement “cars are the best form of transport?”’ Measuring attitudes can be a complex process as they are based upon personal values, beliefs and perceptions which we can’t directly measure. Using a balanced scale is a good idea – with equally spaced measures at either end of the ‘agree/disagree’ spectrum.
4. Hybrid: some data falls between these types. One example of hybrid data collection is using categories to more easily collect numerical information, e.g. ‘how many cars are in your household? None / 1-2 / 3-4 / 5 or more?’ When using categories for numbers, you need to be happy with an approximation.
8. Make the survey interesting
Professionals in certain sectors such as finance or ICT and board level directors are chronically over-surveyed. To stand out from the crowd you need to make your survey relevant, clear, intelligent and interesting.
Be ruthless – for each proposed question, if you can’t visualise a benefit to business objectives from asking it, scrap it or ask the question in a different way.
Draw on your own knowledge – talk to colleagues and use industry sources to help you to design questions and scales that are appropriate. There is no point in re-inventing the wheel and existing knowledge will point you in the right direction.
Step inside respondents’ heads – within a few seconds they need to grasp what you are asking of them. I’m not advocating a blanket ‘keep it simple’ approach, but ensure your question asks exactly what you mean it to ask; has clear and precise instructions and if necessary tells the respondent why you are asking them that question. If you were a respondent, what would your reaction to the question be?
9. Use quality controls
There are several market research quality schemes which ensure minimum quality control requirements in the research process. However, they do carry their own limitations if reduced to a ‘box-ticking’ exercise and they cannot tell you if your research design itself is sound.
Thankfully there are a few simple yet essential rules you can follow – or ensure your supplier follows. Refer to these tips, have a look at the Market Research Society’s qualifications and Code of Conduct and follow two simple yet essential rules:
- Keep a record of all major decisions, including all survey materials, agreements, contracts, things you did to the data after it was collected such as weighting, coding, removing outliers etc. This will help you if you need to repeat the survey at a later date.
- Keep respondent data secure (see the Data Protection Act)
10. Communicate recommendations and implement
Think about how you will communicate the findings before you start the research – it will influence your questionnaire. For great reports and presentations:
- Create clear and meaningful charts and graphics.
- Use simple but impactful messages.
- Use quotes from respondents (ask their permission first)
- Justify your conclusions and business recommendations using the data.
- Start a working group to implement the recommended business changes.