The differences between academic and commercial research

Larnaca April 2014

I am currently at an academic conference on mobile research in Cyprus, a WebDataNet event. I am a keynote speaker and my role is to share with the delegates the commercial market research picture.

I really enjoy mixing with the academic world, and I am intrigued and fascinated by the differences between the academic and commercial worlds.
This post looks at some of the key differences that I have noticed.

Timelines
More BlogsIn the academic world, timelines are usually longer than in market research. For example, an ethnographic project might be planned for 8 months, in the field for 4 months, and spend 12 months being analysed and written up. A commercial ‘ethnography’ might spend 4 weeks in design and set-up, the fieldwork might be wrapped up in 2 weeks, and the analysis and ‘write up’ conducted in 2 weeks.

In many ways the differences in the timelines result from differences in the motivation for doing a research project. Commercial market research is often conducted to answer a specific business question, which means the research has to be conducted within the timeline required by the business question – which is typically rapid. Academic research is typically conducted to advance the body of knowledge, which means there is often not a specific time constraint. However, there is a need to establish what is already known (the literature review) and a need to spend time creating a write up that embeds the new learning in the wider canon of knowledge.

The balance between preparation, action, analysing, and writing up

In the commercial world the answer is the point of the study; the method, providing it is acceptable, is less relevant.

In an academic study, the value of the specific answer is sometimes almost the least important feature of the project. For example, a commercial project looking at five possible ads for a new soft drink would seek to find the winner. An academic project would normally find that sort of result too specific (i.e. not an addition to the canon of knowledge). An academic project might be more interested in questions such as, what is the relationship between different formats of ad and the way they are evaluated, or the extent to which short-term and long-term effects can be identified. Indeed, in academic project the brands and the specific ads tested will often be obscured, because the study is about the method and the generalizable findings, not (usually) about which ad did best.

The definition of quality
Academic and market researchers have a hierarchy of types of validity but the hierarchy is not the same. Market researchers tend to value Criterion validity (does the measure correlate with or predict something of interest) as their ‘best’ measure.

By contrast, the academic world tends to prioritise Construct Validity, which relates to how well new findings relate to an accepted theory of how things work. This again probably relates to the specificity of the objectives. Market researchers need something that works well enough to solve a particular business problem. The academic is seeking to build knowledge and to connect that research to a wider framework.

The difference in samples
Most market research is conducted with a sample drawn from the target population and usually the sample is constructed to be similar to the target population in terms of simple variables such as age and gender – although it usually falls well short of being a random probability sample. By contrast, a large proportion of academic research appears to be conducted with convenience samples, often students.

The most common reason, for using convenience samples, is lack of resources. In some cases there is a belief that the phenomenon being researched is equally distributed across the population, such as preference for using left or right hand.

Access to the results
In commercial research the results are normally private to the client, unless they are for PR purposes. Traditionally, the results of academic research have been made available to the wider academic world. The future of access to academic research is subject to two contradictory trends. Firstly, commercially sponsored research is tending to be more secretive, because of the commercial interests. Secondly, Governments (who are often a major funder) are pushing the Open Data agenda, making research less secretive.

Which is better?
Academic research and market research differ in several ways, but that is mostly because they have different objectives. If you wanted to use a market research project for academic purposes you would need to add a literature review, add a comprehensive write-up, and be prepared to mount a robust defence of your method. If you wanted to use an academic project for a commercial project you would need to check the ethical clearance, check the timelines were going to be relevant, and check whether the study was likely to give an actionable result.

Other posts on this topic:

10 thoughts on “The differences between academic and commercial research

  1. I am a commercial market researcher with relatively little experience of academic research although I have recently worked on a joint project with students from the University of Bath.

    I am always quite surprised at the widespread use of convenience samples in academic research considering the amount of time and effort that seems to go into producing questionnaires in the first place. A weak sample makes a strong questionnaire worthless.

    Despite the time and effort put into designing questionnaires I am always disappointed at how unwieldy for respondents academic questionnaires tend to be. They are almost always long and complicated making them boring for respondents and difficult for them to complete.I recently completed an academic’s questionnaire sent out via Twitter. It was a short survey but it was made up of a 57 statement attitude battery and a few classification questions. After the 33rd statement it was quite hard to give the survey your full attention.

    I suppose if you’re using convenience samples you might not be too worried about annoying respondents!

  2. Great summary! As a commercially trained market researcher who now is on the academic side, I agree with (and experience) all of your findings. The long timelines were a frustration for me when I first started in academia!

  3. This is a good article and I agree with most everything said. Timelines are very different, I have worked on commercial projects with academic partners since the study was being released for public relations purposes. There was a belief that the academic partner would sprinkle “holy water” on the project and help get more press coverage. Inevitably, we were finished with the project in 10 or 12 weeks, but the academic monograph took about a year to complete.

    Other differences is usually the academic partner is more concerned about interviewing technique used, incentives, margin of error particularly for subgroups, randomness of sample, completion rates, interview dates’ and perhaps independently replicating the results, etc. (Good commercial research companies will also be concerned about these facts, and certainly CASRO and AAPOR’s codes of ethics and conduct will demand the same).

    Another difference is that commercial research companies will have a specific business problem or issue to answer and academic research is more interested in the broader meaning for society. Commercial research companies are also interested in making a profit, and academics are primarily interested in finding a breakthrough for new thinking and processing.

    Having said all of this, I do believe academic and commercial researchers can work together and make for stronger and more effective research projects. The key question is the client willing to pay for two different skill sets working together. Unfortunately, over the last six years many clients are not willing to pay for both a commercial and academic researcher working together.

  4. I spent 35 years on the commercial side, but also have academic research articles and have taught in a business school for 6 years.

    (1) One key difference is that commercial research designs are no more complicated than necessary to achieve the result, and re-use of past designs gives you access to potential comparison data AND avoids mistakes.

    With academic research you are trying to create new knowledge, and this often means coming up with a new theory / new methodology / new way of looking at things. As a result, academic research designs are often very complicated and difficult to replicate (both in terms of replicating, AND in terms of getting the same result the researcher did with a new set of data — a problem that plagues medical research as well).

    (2) Academic research attempts to answer A question. If the question being asks shifts in the course of research, that’s seen as OK since that just indicates the fish (new knowledge) are in a different part of the lake. With commercial research, there’s a much higher expectation that you will answer THE question (the one the client wants an answer to).

  5. Thanks Ray, a very important subject. Happy that you are addressing it.

    Jumping in, into the conversation – I have worn both hats: as a commercial qualitative researcher, and then as a research scholar in sociology last few years, for my PhD.

    While the long time frames, convenience samples point towards something that appears frustrating to us commercial researchers let me defend the academic world as experienced by me. Especially since all of us being in business focused Linkedin is a convenience sample of a diff kind 😉

    Taking ethnography as an example, it is hugely frustrating – and often pointless – the way ethno’s are conducted by us MR types. Not just the narrower time frames, but the way we descend on the hapless consumer. We call it ethnography – we as a team – marketing clients et al, land up at the consumer home with pre-conceived notions, seeking quick answers, and often not having the advantage of the wide body of knowledge that a true literature review can provide. One that helps us locate interesting patterns and seek out the right questions to ask of the study; the right jigsaw puzzle pieces to pick up from the vast universe out there.

    If any of you are on twitter you would know how it feels when someone you follow comes up with some comment that he or she thinks is smart and interesting and funny and new – while you know you have already discussed it earlier. The other person does not know cause he or she follows a narrow band of people. (As you can see, I find those who follow ’40’ but have 4000 followers pretty weird navel gazers)

    It is kind of like that. We in the commercial world live in a narrow world and are simply not aware of the VAST array of knowledge already existing – that can only sharpen our findings even in the business world.

    There is a lot to be said for meeting somewhere in between – not QADR nor ‘Till Kingdom Comes’ timeframes.

    Look forward to continuing this conversation,

    Piyul

  6. Hi Ray

    I am a commercial (or was) market researcher and have worked as a market research manager and consultant in the HE sector for 10 years

    For me the biggest difference is the “WHY” is the research being conducted in the first place and very often (not always) the primary reason for the academic research is producing a paper for publication whereas most of the research I did was to answer a particular question or problem. I do not mean this in a negative sense

    One area that commercial researchers could learn from is the large amount of desk research/literature reviews that the academic side put into their papers and use to the maximum existing knowledge whereas partly due to confidentiality and and partly due to laziness commercial researchers ignore what already exists (often even their own research).

    One thing that I found noticeable in my time in the HE sector was the willingness of the commercial researchers in the sector to co-operate and share learnings without compromising confidentiality or competition

  7. HI Ray — As I am sure you appreciate, generalizations about academic researchers and their methods are dangerous. There is broad variation across disciplines. Psychologists, for example, are (in)famous for their reliance on convenience samples of students. Political scientists or survey methoodologists (with whom I have worked a great), on the other hand, often have difficulty getting published unless they have used a probability sample. Academics who work with convenience samples often understand the shortcomings of their method in a way that commercial market researchers do not. They typically have a strong belief in replication and development of theory over time. And their work is subject to critical review that is completely missing in commercial research. The unfortunate thing, at least here is the US, is the general lack of appreciation from both groups of the difference in environment and purpose in the two spheres. Both tend to miss the opportunity to learn from one another.

  8. Hi Ray,

    Thank you for your presentation at WebDataNet meeting, for discussions, and for this summary of the differences between academic and commercial research. You maybe forgot to say about the different way of presenting the information at scientific and business conferences. I think you felt the difference 🙂

    Considering our discussion of the gamification, I would probably also add that academia is different in the attitude towards some innovations. Business-orineted research always aims to apply and find out an innovative approach (this is the core of the business), while academia is always far behind. After a while academia can accept innovations, but only if it improves data quality.

  9. Academic research is often more narrow in focus than commercial research. The aim of an academic study is normally to test a narrow construct – eg do 7-point scales work better than 5-point scales, do best-worst conjoint questions give better results than a simple choice, with some discussion of what do we mean by best – better predictive ability, more stable results, better spread. Commercial studies by there nature have to address commercial questions so there is more searching for answers than methods. My big problem with academic research can be the tendency to dress simple findings up with overly complex maths – using SEM when a contingency table or chi-squared test would be clearer, or where the sample data is actually quite weak, so making it look clever – but often obscuring something simple. Given how much commercial research is about how best to communicate results and insights, this is always a big difference.

    And like Gary, I also struggle with the low quality of questionnaires for instance big banks of scale questions with a forced response using words or phrases straight out of an academic journal. Sometimes it’s difficult to totally be sure about academic results because you can’t feel comfortable with how they did the research.

Leave a Reply

Your email address will not be published. Required fields are marked *