Usually, yes. Learn why the goal of market research is never perfection – and why that should come as a relief
Too many marketers want a Mary Poppins version of market research: “practically perfect in every way.” This is understandable. Research is expensive, it takes time to do it properly, stakeholders want certainty and, no matter how good your presentation skills are, defending a recommendation built on “our best approximation” can feel like an uphill battle.
But certainty in research very rarely exists. Extensive research in the Social Science & Medicine Journal (2019) found a gap between consumers’ hypothetical and actual willingness to pay, which was on average 3.2 times lower in real purchasing situations. What consumers say and what they do can be drastically different.
Another study focusing on the purchasing decisions regarding environmentally friendly products found that self-reported attitudes to green products did not transfer to actual purchasing behaviour. The pursuit of perfect data sets is impossible to achieve because it does not exist.
One of the most useful lessons covered very early on in the MiniMBA in Marketing is that research is there to reduce uncertainty, not remove it. Research does not and should not strive for perfection. In the search for flawless data, more often you end up with something else: no data, no decision, and no progress.
In the search for flawless data, more often you end up with something else: no data, no decision, and no progress
Marketers need to embrace probability. A common benchmark in survey work is 95% confidence with a 5% margin of error, which is a disciplined way of saying: we are confident enough to act, while accepting that some uncertainty remains. That is not falsification, nor is it an incorrect research methodology, it is quite simply how real decision-making works in a business.
The job of market research is not to make every datapoint fit neatly. The job is to build the strongest possible view from the evidence available, then make the call on whatever question or decision your research was designed to answer.
When digging deeper into data, remember that context matters too. The wording and even order of your questions can shape the answers you receive, with the ONS documenting that the order of questions in a crime survey for England and Wales changed the results by up to 8.9%.
The imperfections of market research do not make market research useless. They simply shift the challenge. The question is no longer whether research perfect. It becomes how to strengthen your evidence anyway. This is where synthetic data becomes interesting.
Synthetic respondents do not lie in the human, self-presentational sense. They do not forget what they did last week or try to impress the moderator, yet we shouldn’t take for granted that synthetic data is superior. The ESOMAR publication - Synthetic Data in Marketing Studies (2024) - Exploring the promise of generative AI and synthetic data - investigates whether AI-generated synthetic data can reliably supplement or boost underrepresented groups in market research, presenting results from over 7,000 tests using real-world datasets. The research shows that synthetic data can be valuable when used to augment real research, especially when there are clear validations and defined limits.
Is this the answer then? A healthy blend of both human and synthetic data? It could be and only time will tell, but what we do know for certain is that marketers need to get more comfortable with the uncomfortable truth that no research will ever be perfect.
To be clear, this is not a defence of bad research. It is not an excuse for weak samples or vague questioning. Good research still needs rigour: sensible design, robust sampling and a clear view of what it can and cannot tell you. But marketing rarely happens under perfect conditions. Budgets are finite, deadlines are real and some audiences are hard and expensive to reach.
We work in an imperfect field. When I meet these marketers that have been paralysed by the idea of getting perfect research, I always say the same thing to them...
In these cases, Mark Ritson implores marketers to go easy on themselves: “So many times on my journey in marketing, I meet really good marketers who have become constrained by the idea of perfection...
“Let me reassure you, there's no such thing as perfect research or pure scientific marketing. We work in an imperfect field and when I meet these marketers that have been paralysed by the idea of getting perfect research, I always say the same thing to them: The fundamental joy and challenge of market research is accepting the imperfections, but getting out there amongst the consumer and understanding the market” (Module 2: Market Research, the MiniMBA in Marketing).
Run the best research you can within the limitations you have, be honest about the gaps, and use it to make a better call than guesswork ever could.
Mark concludes, “It's never perfect. It doesn't have to be. But we have to do it and bring those insights back into the organisation.”
Stop trying to make everything fit. Stop treating untidy evidence as failed evidence. In market research, the standard is not perfection. It is useful, directional, decision-sharpening confidence. That may not sound as elegant as Mary Poppins, but in practice, it is the closest our discipline ever gets.
For more imperfect but effective marketing advice, read “Three marketing ‘tweaks’ to get the most out of a tight budget” – including how to approach market research when a quantitative study is off the table.
Or, read what Mark Ritson has to say about synthetic market research in “Synthetic data is as good as real – next comes synthetic strategy.”
Cover: Julia / Adobe Stock