A sample design that represents
Who has to be asked for the result to mean anything. It is the single decision that most determines whether the study is any use.
Opinion research, market studies, censuses and statistical processing. With the methodology stated, the margin published and the raw data handed over.
What it solves
It is the most accessible sample and the most biased: the people who already buy from you will tell you why they like what you do. The ones who did not choose you — who are the ones that explain why you are not growing — are not in that conversation, and a product or pricing decision made that way comes expensive.
What is included
Who has to be asked for the result to mean anything. It is the single decision that most determines whether the study is any use.
You can get almost any result by changing how you ask. That is why the questionnaire gets reviewed and handed over with the report.
Interviews verified and fieldwork supervised. Without that, part of a data collection can be fabricated and the analysis will not catch it.
Cross-tabulations, significance and segmentation, not just percentages. What matters is usually in the cross-tab, not in the total.
On top of the report. Anyone who hands over conclusions alone is asking you to trust them; with the data, anybody can check them.
Market studies, satisfaction, workplace climate, censuses and programme evaluation, each with the rigour it demands.
How it works
The business question first; the instrument afterwards.
Sample, questionnaire, data collection and quality control.
A report with interpretation, plus the raw data.
In detail
A thousand badly distributed responses are worth less than three hundred well designed ones. Size impresses; what matters is who it represents.
What gets asked first conditions what gets answered later. It is a measurable effect, and that is why the questionnaire is ordered deliberately.
If one group responds much less, the result tilts without anyone seeing it. It gets measured and corrected, or it gets declared.
A global 60 % can hide a 90 % in one segment and a 20 % in another. The decision is usually right there.
The deadline is part of the design: if it arrives after the decision, the rigour makes no difference. Better to narrow the scope and arrive on time.
With the raw data, their conclusions can be checked. Without it, you have to believe them.
Who it is for
And you need to know whether there is a market.
Of customers or of your own team.
And you need to evaluate a programme rigorously.
Common questions
It depends on who they represent, not on the number: a thousand badly distributed are worth less than three hundred well designed.
Yes, on top of the report. Anyone who hands over conclusions alone is asking you to trust them; with the raw data they can be checked.
With interview verification and supervision. Without quality control, part of a data collection can be fabricated and the analysis will not catch it.
Yes: censuses, programme evaluation and opinion studies, with the rigour and traceability it demands.
The deadline gets set at the start, because it is part of the design: a study that arrives after the decision is no use however rigorous it is.
Yes, and that is where the value is: the total usually hides what matters, which shows up when you cross-tabulate by segment.
And also
We will show you Surveys and Market Research running on a case much like yours, with no obligation.