Thursday 23rd November 2017
A few weeks ago I attended an interesting seminar held by the ASC (the Association for Survey Computing). The topic was about the rise of automation in the market research industry. Four speakers who are well known in the industry gave their opinions about why they think that automation is either a threat or a great opportunity. Both sides provided convincing examples and arguments to support what they stand for. It stimulated my thoughts about automating market research.
Automating market research does not just happen now, it has been happening for quite a while. In fact a lot of tasks that we are doing on a daily basis are automated or semi-automated; quota management, SWB, automated charting etc. It does not only save time and increase productivity, it helps us have a quicker turnaround to deliver survey results to the client.
So if automation in market research is expected to accelerate in the next decade as the speaker who is in favour of automation suggested, then I am actually quite excited. If the task is repetitive, tedious and also takes a lot of time to do then what is not to like if it can be automated? We would be freed to spend more time on more bespoke parts of work – building customer relationships, better communication, project managing, understanding what the client needs.
Those who are in favour of automation argue that maybe the ultimate goal of automation is to take over our jobs. Should we be worried that automation will eliminate employment? There has already been an evident decline in the use of pen and paper surveys, face to face interviewers and data entry staff. The recent news about AlphaGo has demonstrated how smart and intelligent an AI can be superior to human. The use of SIRI and Alexa has gradually become part of our daily lives. In the market research industry, the demand of data scientists has increased sharply; even facial expression in videos can be coded automatically.
It looks like machines increasingly do our work for us. And yet I find an interesting fact that in the 45 years since the introduction of the automated teller machine, the number of human bank tellers employed has roughly doubled in the US from a quarter of a million to a half a million . Similarly in the recent MRS sector review by PwC, “the Business of Evidence 2016, concluded that the market has grown by almost £2 billion since the last sector review in 2012, with the number of full-time employees increasing by nearly 14,000 in that time.” . This poses a paradox if automation is doing our jobs, why do we see an increase in employment? There are actually two fundamental economic principles: one has to do with human intelligence, innovation and creativity. The other has to do with human’s desire for more or greed. The first one is called the O-ring principle which determines the type of work that we do, the second principle is the never-get-enough principle which determines how many jobs there actually are .
When parts of our daily routine tasks are automated we are freed to do more cognitively demanding job such as solving problems, building customer relationships, consulting, quality checking etc. Most jobs involve various complex tasks - automating some of those tasks does not make the other unnecessary and instead it actually makes the other ones more important . A project can be seen as a series of interlocking links in a chain. It is like the game show the Weakest Link . Every one of those links must hold for a project to succeed. If any one of them fails, a project or a service fails. Automation could increase the reliability in the chain however it also increases the value of improving any of the other links. The more automation or AI makes the other links become robust and reliable, the more important and essential the human link is. The whole project depends upon the tasks we humans do. As our tools improve, technology magnifies our control and increases the importance of our expertise . We would not call a project/product a success if we have a solid/reliable product with crappy customer service. This also explains why we see a decline in data entry staff and interviewers. Because if there are not many human demanding tasks left in the jobs then technology would take over our jobs .
The second principle is never-get-enough principle. A lot of services and products that we are using today barely existed a century ago including market research IT services. As automation frees our time and increases the possibility of what humanity can achieve . Innovations take place and we invent new products, services and ideas. This is why we see the roles of data scientists or data engineers emerge recently. They have the expertise to crunch the data and extract meaning from it . The expertise they have is still relatively new in our industry but it has demonstrated how much people desire new things.
Automation poses a challenge there is no doubt about that. Yes automation would eliminate some employment depending on the nature of the jobs. The past has shown us that we never stop innovating new things just because the productivity of the existing things has gone up. Perhaps if you understand more about your role you would/would not be so worried about automation.
Perhaps only time will tell the impact of automation and the effect on our lives and work.
1. PwC, MRS sector review, the Business of Evidence 2016, Market Research Society
2. D.Autor, Will automation take away all our jobs? Tedtalk
3. B. Rapolu, What’s a data scientist and how do I become one? The Guardian
4. D.Greenfield, Automation+jobs: Not a zero-sum equation, ProFoodWorld