Automation: the good, the bad and the ugly

Editor’s note: Stephen Phillips is CEO of marketing research firm ZappiStore, London.

As claimed in a new report (registration required to view) I co-authored with Ray Poynter, this is the year of automation. It’s the most important thing for marketing researchers to get right in 2017. This year will reveal some MR companies coming out on top with growing success, while many others will fall behind. The report’s argument is that it’s important to embrace automation, while carefully considering what can and can’t be automated, because only the companies which are successful in embracing automation will prosper.

Corporate researchers, you must be wary of the good, the bad and the ugly among the four benefits of automation – cheaper, faster, scalable, quality – before you can take automation by the horns.

1. Cheaper research for industry growth.

As budgets are squeezed, it’s essential to find ways to do the same (or more) for less. Automation reduces costs, which can cause prices to fall and ultimately the market to grow. Looking at past examples we can see how the automobile industry grew with automation, supporting the birth of hundreds of millions of cars and the hiring of more employees. As clients, cheaper research has short-term benefits but looking at the long-term picture, cheaper is only an advantage of automation if it leads to growth of the marketing research industry.

2. Faster insights for same-day business decisions. 

Many corporate insight teams have been decimated. One person is expected to do the job of three. This is why corporate researchers are embracing automation; it delivers insights faster than ever. However, this is only a benefit if it speeds up the process of business. As corporate researchers, speed can move you into the decision-making process much faster. But automating for speed must only be used where faster actually helps, where research becomes a same-day result and speed changes the deliverable and the decision.

3. Scalable to keep up with the need for research.

The need for quality research projects is continuing to grow but the number of quality marketing researchers – both agency and client side – is staying static. People are the scarcest resource of all in a growing industry. Automation can assist in making the process scalable so researchers can continue to produce quality research which drives effective decision-making. Scalability is only a benefit to a corporate marketing researcher when you are looking to increase capacity and increase the impact of research and data within your organization. Many organizations are finding they can now complete eight or 10 times more research studies than they did pre-automation. Faster, cheaper projects can be used at many more points in the development cycle and fitted to the needs of multiple internal stakeholders.

4. Quality – is it better research?

The time that automation frees up means the researcher can focus on the parts of the research process that cannot be automated – design, analysis and storytelling. This means you can expect your agency to deliver more consultancy around your project, or you can spend more time thinking and analyzing your data. And finally, because of the reduction in cost, you can do more frequent studies to check the value and progress of your hypotheses and demonstrate the success of decisions. Insight can really sit at the heart of your organization. 

Avoiding the ugly

The most important thing for client-side researchers is to understand what should and should not be automated. There are some things we all want automated, such as cluster analysis and sample control mechanics. There are some areas within marketing research that are already distinctively improved when automated or have already caused a notable change. For example, online surveys have automated the interviewer, making them cheaper, faster and scalable (but in Poynter’s opinion maybe not quite as good).

So then, what can’t be automated? Most face-to-face qualitative research doesn’t benefit from automation. There’s also the work behind understanding, defining and explaining the research problem, which we are light years away from automating and which will always rest with the expert client. It’s also necessary to take note of the dirty word of automation – standardization. You cannot use automation without standardizing your output. Think long and hard about what you want as your standard deliverable and expect to stick to that to receive the ongoing benefit.

Ultimately, for something to be automated it must produce big cost and time savings for both the agency and the client. You should expect to automate the tasks where the benefits very clearly outweigh the costs.

You don’t need to be first

A final tip for corporate researchers is that unless you love technology and failure, avoid the bleeding edge. Embrace those MR companies that are taking risks but make sure you work with those that are proving ROI. Remember, Apple, Google and Facebook weren’t first. Make sure that your supplier has been running automation smoothly for at least 12 months and they can share multiple case studies of the benefits to other clients.

Move positively into the world of automated research, with alert senses and let it empower you to become more productive by allocating saved resources toward skilled intellect and creativity to produce better, faster research.