We're all in the business of market research. Agencies exist to make money. Clients buy research to make (or save) money. Clearly this works, or there would be no agencies, and precious few clients. But 'how' does it work?

At this year's BIG Conference, I examined the research industry from a 'business' perspective. In turn, this provoked questions about the economics and business models underpinning it and led to three main observations.

We measure our industry as an 'industry' not a service'.

Market research statistics have traditionally focused on turnover: the amount charged by agencies to clients. It's how manufacturing and retail industries benchmark themselves - by sales. Yet practitioners in sectors that researchers usually compare themselves to - advertising, PR, brand consultancy, etc. - are more interested in measuring fee income.

This focus on the top line means we tend to measure ourselves in a way that lends itself to industrialisation and commoditisation. This is a global phenomenon. And to reinforce this, we have designated ourselves an 'industry'!

The P&L account of our industry looks production rather than service oriented.

Looking at profitability of research agencies, and the underlying business models that drive them, there is clearly a dominant economic model in the research 'market'. This is more analogous to production-oriented sectors than consultancy professions.

Similarly, it is clear that acquisition, economies of scale and cost management - rather than true revenue growth - enables the larger players to maintain profits.

Distinct segments in our market reflect different financial dynamics at play.

In any sector there are three basic ways to make money on the supply side. Firstly, sell lots of stuff very competitively and make production and delivery highly economically efficient: 'Volume'. Secondly, sell high value, high prestige stuff, and ensure it is appreciated accordingly: 'Value'. And thirdly, sell an idea; create demand for specialist new stuff that may not even exist yet: 'Vision'.

Readers will have an immediate view on where particular agencies or agency types might sit. Thinking beyond our sector, under (1) we would typically find manufacturing, retail, and production businesses; product and volume-oriented; whose business is cost control and efficiency-led. Under (2) and (3) we would find advisory businesses, with a less pyramid-like structure. So, if there are two different markets, where is qual heading?

My initial instinct was to assume that, as a 'sub sector', qual would naturally fall under the heading of the advisory business model rather than the production-oriented one. But this might not be the case, because:

MR in general is becoming more production oriented, likewise qual.

A common currency of unit cost per group discussion now exists. This key evaluative metric applies to most qualitative practice. Standard qualitative methods are becoming a - albeit high quality - commodity. As such, there are price and delivery time pressures.

MR, especially in the qualitative and innovations space, is blurring at the boundaries.

So, what is the definition of 'research'? Increasingly money is being spent on research activities that are not 'mainstream'; non-researchers are doing 'researchy-type things' as part of a wider offer. Researchers are doing other things too, of course, but more competitors exist than those we'd normally call research agencies.

Different business models are co-existing uncomfortably.

We established there is no 'single' model in our sector. We are structured as a hybrid of 'consulting' and 'manufacturing'. Buying patterns have traditionally reflected the latter: they are mainly driven by fixed price, units of product. In turn, models that are appropriate to the buying of large scale data collection are being applied inappropriately to various types of consulting.

In fairness, buyers' working patterns with agencies are changing too: witness increasing reference to 'data providers' and 'insight providers' in roster definitions.

So, should data collection be seen as a commodity? Yes. There is a measurable unit of production with associated quality controls, and that's all there is to it.

Should group discussions be seen as a commodity? Probably. The standard group discussion, without any bells and whistles, is now treated very much like a basic unit of production.

Will the MR sector as a whole become commoditised? Yes and no. There is an entire sub-sector of activities and advisory services unrelated to the data machine that should be priced on time/value, not cost, but currently it's insufficiently delineated.

It's being treated as one: partly because agencies are selling these activities and clients are buying them as such. Also, partly, it's because clients, and suppliers work to a fixed price overall project model. Is this really appropriate for current work practices?

Maybe it's time for a structural reorganisation within the MR industry, and a new narrative. Here, businesses would be segmented into 'makers' and 'advisers'. In MR, the 'makers' would include most of the mainstream and might well redefine their core business(es) around 'Volume'.

But at the 'Value' and 'Vision' end of our market, to avoid commoditisation, quallies can continue to fight their corner through relentless innovation. Luckily, this has always been one of their strengths. The best outcome is self renewal; a resurgence of new methods analogous to the original meteoric rise of 'qual' in the 1980s. Clients are saying that this is exactly what they'd like and expect to see. Most importantly, it will be very good business for everybody.