Reviewing My Predictions for 2016, 2017 and 2018

Janus looking forward and backwardPosted by Ray Poynter, 2 January 2019


I will soon be publishing my predictions for the main themes for market research and insights for 2019, but perhaps it is useful (if painful in some instances) to look back at my previous predictions. In this post I re-visit my predictions for 2016, 2017 & 2018.

Predictions for 2016

There were 9 predications for 2016, published in ‘My Predictions for 2016’.

  1. Bigger Legal Problems for Facebook and Google. Conclusion, this was true in 2016 and 2017 and exploded in 2018 for Facebook. Expect more in the future. IMHO, the Facebook Cambridge Analytica scandal probably helped the GDPR legislation slide into implementation more smoothly. (Prediction score 10/10)
  2. Automation. Conclusion, 2016, 2017, 2018 were all powered and shaped by automation, this will continue for several more years. (Prediction score 10/10)
  3. Surveys will continue to suffer. Conclusion, as the ESOMAR GMR report makes clear, for the last few years surveys have become a slightly smaller share of the total market research pie, however it is also important to remember that surveys are still the biggest share of the MR methodology pies and that will continue to be the case for quite a few more years. (Prediction score 8/10)
  4. More Emotion, in the sense of recognising the role of non-rational, non-explicit, and System 1 types of responses. Conclusion, this certainly grew in 2016, 2017, 2018 and will continue in 2019. (Prediction score 7/10)
  5. Big Data Continues to Disappoint. Conclusion this was very true in 2016, 2017, 2018 and I predict in 2019. (Prediction score 10/10)
  6. Gender equality or suffer the consequences. Conclusion, the progress in 2016 and 2017 was measurable, but slight, but 2018 saw substantially more progress and this is now the status quo (at least in terms of what organisations say they do). (Prediction score 7/10)
  7. Tracking, the dam starts to crack. Conclusion, yes, but really small cracks. There are tracking studies that have updated themselves to new models, but in most cases companies have simply reduced the sample size and the reporting frequency and left the zombie that is tracking as the walking dead. (Prediction score 1/10)
  8. Social, Local & Mobile. Conclusion, Mobile has been the big story of the last few years, 2016 and onwards. Social has established a position in market research and insights and according to GRIT is nearly mainstream – but it has not expanded as much as I thought it would. Geolocation remains a tiny proportion of MR, a clear miss for the predictions. (Prediction score 5/10)
  9. Training & Competence Building. Conclusion, whilst there has been some growth in the demand for my training services (and for courses by the likes of AMSRS, ESOMAR, MRII, MRS, and ResearchRockstar) the report by Sue York and myself on training shows that most market researchers are not being trained. Prediction score 2/10)

Overall review of my 2016 predictions? There were a couple of duff predictions (about the decline of trackers and the rise of training) but the rest of the predictions held true for 2016 and beyond. The training and competence building miss is a good reminder that a strong growth from a small base produces a small result. If 1% of people are being trained properly and this expands by 25% a year, then after 3 years it will have reached 2% – i.e. it would still be tiny.

Predictions for 2017

There were five predictions for 2017, published in Five Market Research Trends for 2017.

  1. Automation. Prediction: it will continue to be a trend. Conclusion, yes it happened in 2017 and 2018, and it has been a key element in the ability of MR to sustain lower prices – this will continue to be a key trend for several more years (Prediction score 9/10).
  2. Insight Finding. Prediction: improvement in tools like E-Tabs, Infotools etc. Conclusion, yes the tools have got better (in 2017 and 2018), but the usage has not advanced as quickly as I thought/hoped. The tools will keep improving, but the use of them requires a change in thinking and training. (Prediction score, 5/10)
  3. Insights to Action. Predictions: more proprietary approaches for turning research into action, more client-side re-organisations. Conclusion, it happened but slower than I predicted, it is still happening, it will continue in 2019. (Prediction score 4/10)
  4. Implicit Measurement. Predictions: increased focus in non-conscious, a move away from direct questions, and the re-badging of existing techniques (such as conjoint) as implicit. Conclusion, whilst implicit is better understood it has not really increased its share of the MR business in a big way in 2017 or 2018. Research providers and buyers seem aware that direct questions are often flawed, but are sticking with them. Implicit will grow slowly until the paradigm of MR changes (which might not be soon). (Prediction score 3/10)
  5. AI, artificial intelligence was predicted to be everywhere in 2017, but mostly smoke not fire, but with a firm message to keep an eye on it. Conclusion, in 2017 it was mostly smoke without fire, but through 2018 the use of it has grown tremendously, as the report I co-authored with Rosie Ayoub shows. (Prediction score 9/10)

2017 overall? Two hits, Automation and AI, but with three weaker predictions. All of the weaker predictions were correct about more tools and understanding being available, but they overestimated the extent of widespread adoption. The key lesson was that buyers and providers kept going for low-effort existing options in preference to using newer and better options. This is a good reminder that most people do not want their business/way of life to be disrupted.

Predictions for 2018

The predictions for 2018 were in a different format, the predictions were informed by crowdsourcing and published on the GreenbBookBlog.

  1. What will be the hottest MR topic in 2018? Predictions: Automation, with a mention of fake polls and mergers and acquisitions. Conclusion, automation continued to drive change in 2018 (one interesting reflection of this is Mike Stevens Insights Platforms which lists 100s of MR platforms). The issue of fake or dodgy polls was a major issue in Canada (with implications for all of us), the German industry is becoming every more concerned about badly run polls, and the topic of fake news in general has grown in 2018. There have been some interesting mergers and acquisitions, but the biggest change must be the purchase by Ipsos of GfKs ad hoc research business, and the biggest acquisition was the purchase of Qualtrics by SAP for $8billion. (Prediction score 8/10)
  2. Which country will be the most interesting to follow in 2018? Predictions:the main prediction was China, with a backup prediction of USA. Conclusion, there are lots of interesting things happening in China in terms of insights, but few of them are big enough to have a global impact. Looking at the companies who are attracting the most attention at exhibitions the key locations are USA followed by the UK. (Prediction score 5/10)
  3. What will be the biggest MR disappointment in 2018? Predictions: ethical problems, fake polls, and any new technique that was slower or more expensive than an existing technique (e.g. neuro, VR, and AR). Conclusion, these mostly came true in 2018, in terms of ethics see Facebook/Cambidge Analytica and to a lesser extent Facebook/Crimson Hexagon, for poor polling and fake polls see the ESOMAR review of the Calgary polls – although this was late 2017 when it happened, and the GRIT report on emerging technologies shows that the better but more expensive and slower MR innovations have stalled.  (Prediction score 9/10)
  4. Which MR Company will be the hottest in 2018? Predictions, video-related companies such as Voxpopme, Living Lens, and Watch Me Think. Conclusion, video companies were generally the hottest in 2018, including these three along with others such as Big Sofa and Indeemo. (Prediction score 8/10)
  5. What is the biggest threat to MR in 2018? Prediction: lack of training for MR staff. Conclusion, it is clear the lack of training is widespread, whether or not it is the main threat has not been established, but it is clearly a major threat. (Prediction 7/10)

Generally, the 2018 predictions held up well, the main weakness was the prediction about China. This forecasting error was based on confusing rate of change with impact of change. The rate of change in China is fast, but it is a much smaller research market than the UK and tiny compared to the USA, so its impact is less than its rate of change might indicate.

Overall Conclusion

There are some misses, usually where I have assumed that common sense would have dictated action (for example that companies would dismantle the walking dead legions that are the tracking studies and that they would tackle the terrible lack of training). If tackling a problem requires doing something, then the something often does not get done.

However, in most cases the predictions have held up well, not just for the year they were predicted for, but beyond. Let’s hope the 2019 predictions do as well.

3 thoughts on “Reviewing My Predictions for 2016, 2017 and 2018

  1. Hi, thanks for taking the time to make these predictions and reflect back on them. Can you elaborate on the biggest threat to MR in 2018? What do you mean by the widespread ‘lack of training’? Are you referring to staff within established MR firms/client companies? Or that people who have no MR background are performing MR studies? Also, how did you make this determination? Thanks!

  2. Hi Alicia, there is more information in the original post, that was making the 2018 predictions, see it at https://greenbookblog.org/2018/01/17/what-will-happen-to-market-research-in-2018/. That post drew on a study we ran in 2017, see https://newmr.org/blog/market-research-knowledge-benchmarking-study-2017/. We also ran a study in 2018 which dived deeper into the problem https://newmr.org/blog/market-research-skills-and-training-study-2018-report/. The core issue is that too many researchers working in market research are not receiving the training they need if the industry is going to deliver a ‘value add’ to data. Two key areas are 1) lack or core MR skills and knowledge and 2) familiarity with the new tools and technologies.

Comments are closed.