Market research data

Quantitative data is numerical information that can be measured and counted — such as sales figures or percentages — answering questions like 'how many?' and 'how much?'.

The mean is found by adding up all the values and dividing by the number of values, giving an overall average. The median is the middle value when the data is arranged in order. They differ because the mean is affected by extreme values (outliers), which can pull it up or down, whereas the median is not — so for skewed data the median often gives a fairer picture of the typical value.

Quantitative data gives numerical evidence — for example, that 70% of customers dislike a new design. This shows the scale of an issue and is easy to compare and put into charts, telling the firm that a problem exists. However, numbers alone do not explain why customers feel that way.

Qualitative data fills this gap. Opinions from interviews and focus groups might reveal that customers find the design 'old-fashioned'. This explains the reason behind the numbers, so the firm knows how to fix the problem — for example by modernising the design.

Using both together is therefore more powerful than either alone: the quantitative data shows that there is an issue and how big it is, while the qualitative data shows why and what to do about it — leading to better, more informed decisions.

The reliability of market research data depends on several factors, and it matters because acting on unreliable data can lead to a costly wrong decision.

The first factor is the sample. If the sample is too small, the results may be due to chance rather than a real pattern; if it is chosen in a biased way (e.g. convenience sampling), some groups are over- or under-represented. Either way, the data does not reflect the whole population, so conclusions drawn from it could be wrong — for example, the firm might target the wrong customers despite believing it has 'evidence'.

The second factor is how recent the data is. In a fast-changing market, data collected some time ago may no longer reflect current customer tastes, so even accurate past data can mislead a present decision.

The third factor is source credibility. Data from an independent or official source (such as government statistics) is generally more trustworthy than data from a self-interested source, such as a competitor's own marketing claims, which may be biased.

Therefore reliability depends on how the data was collected, not just the figures themselves. A precise-looking statistic from a small, biased or out-of-date sample is still unreliable. The most reliable data comes from a large, representative, current sample from a credible source — and a firm should weigh reliability most carefully when the decision at stake is major and costly.

There are reasons the firm can place some reliance on the data. It is based on the firm's own real sales records (internal data), which are accurate and directly relevant, and the data shows a clear correlation: sales and advertising spend rose together. If advertising genuinely drove demand, the data supports investing more in it, and using actual results is better than guesswork.

However, there are strong reasons for caution. The data shows correlation, not causation — two things rising together does not prove one caused the other. Sales could have risen for other reasons that happened at the same time: a seasonal peak, a rival raising its prices, or general economic growth. If the firm wrongly assumes advertising caused the rise, it could overspend on advertising when the real driver was the season, wasting money on a major decision. The reliability also depends on how long the period was: a few months is too little to confirm a lasting relationship, and one-off factors could distort it.

Judgement. How far the firm should rely on the data depends on whether it can rule out other causes and on how much is at stake. For a major decision, the firm should not rely on this correlation alone: it should investigate alternative explanations — for example, by comparing sales in a period without extra advertising, or controlling for the season — before committing. The data is a useful starting signal, but because it only shows correlation and covers a short period, it is not a sufficient basis for a major decision by itself. The firm should treat it as one piece of evidence among several, and gather more reliable proof of causation before acting.

Market research produces data that businesses use to make decisions, but the value of that data depends on how reliable it is and how it is interpreted. How far a firm can trust its research data depends on how it was collected, what type of data it is, and how it is analysed.

The case that the data can be trusted. Well-collected data is a sound basis for decisions. A large, representative sample chosen by a sound method (random or stratified) reflects the whole population, so its quantitative findings — percentages, averages, sales figures — give an accurate, comparable picture of the market. Combined with qualitative data explaining why customers feel as they do, the firm gets both the scale of an issue and its causes. Data from credible, independent sources that is up to date can reasonably be trusted as a guide to decisions, and is far better than relying on guesswork.

The case that trust should be limited. First, data is only as good as its sample: a small or biased sample (3.2.3) produces unreliable results, and a precise-looking figure does not guarantee accuracy. Second, data can be out of date in a fast-changing market, so past patterns may not hold. Third, interpretation can mislead — charts with misleading scales exaggerate differences, and assuming correlation proves causation can lead a firm to act on a false cause. Fourth, respondents may say one thing and do another, so stated views overstate real behaviour. For all these reasons, data should never be treated as certain.

Weighing it up (criterion). How far data can be trusted depends on the quality of the sample and source, how recent the data is, and how carefully it is interpreted. High-quality, current data from a credible source, interpreted with care, can be trusted a great deal; small, biased, dated or carelessly-read data can be trusted little.

Judgement. A business can trust its research data to a significant extent, but only when it is well-collected and carefully interpreted. The most defensible conclusion is that data should be trusted in proportion to its quality: a large, representative, current sample from a credible source — read with attention to scales, sample size and the correlation-causation trap — is a strong basis for decisions, while poor or carelessly-interpreted data should be treated with caution and supported by further evidence. So research data is a powerful but imperfect aid: it improves decisions when it is reliable and well-analysed, but it never removes the need for judgement, and over-trusting weak data is a serious risk.