Author: Nadine Touma Nadine Touma is a consultant in Insights, Strategy & Innovation based in Dubai. In her blog, Let’s talk business, she shares her 20 years of experience in the corporate world and 5 as an independent consultant with real-life stories, simple advice, and a no-nonsense approach that characterizes her work Original Post
This series of posts on the topic “Data are not insights” aim at debunking some myths around the mysterious and very much talked about the field of data, among which, the simple fact that, no, data analysis is not that complicated and certainly not reserved to a few enlightened ones. They also aim at giving simple and practical advice on how to avoid typical mistakes and reach the Holy Graal of powerful insights.
I’ve addressed in Part 1the importance of stating your source, allowing it to be questioned and cross-checked at any point in time. Obvious, I know, and yet few people do it. Allow me to point out here that, no, Wikipedia is not a source. It is an easy way to understand certain things, concepts, acronyms but cannot be featured as the basis of your demonstration. IMF, UN, CIA (yes, THE CIA… they publish excellent country reports), governmental bodies, respectable consulting and research companies are sources, Wikipedia isn’t. Nor are Siri or Alexa…
Apart from the data coming directly from your company (sales, profitability, traffic in-store, number of users, views etc.), everything else you will be called to use at one point in time in your business plan, financial forecast, or marketing plan will have a sizable share of estimation.
Estimations, by nature, are calculations that are based on assumptions, that not only can be debated and questioned but can change over time.
No estimate, absolutely none, is 100% accurate, far from it.
I once attended a presentation from a colleague who was tasked with sizing a certain category and establishing market shares of big players and major brands. The presentation included none of his assumptions, none whatsoever, and he dramatically concluded: “I can be struck by lightning now if my estimates are wrong”. Ok, so first of all, this is not an Egyptian remake of a Greek tragedy, so please refrain from your enthusiasm. Second, you cannot, by definition be 100% right and, third, the whole point of this exercise is not to be “right” but to provide a market basis to benchmark the performance of our own brands against the competition. This basis will need to be refined over time as we gather more qualitative comments and testimonials. Bottom line is, it will never be “100% right” for the simple reason that we do not know the exact figures of our competitors and the best we can do is get closer in our estimations with a lot of hard work and a bit of common sense.
Some sources are more accurate than others. They just are and will always be.
The size and global presence of the agency that constitutes your source is, usually, directly proportionate to the quality of the estimate they produce. Simply because they have large teams to work on it, greater access to, say, companies figures and key stakeholders, making their share of actuals vs estimates bigger and their assumptions closer to reality.
I say “usually” though because I have faced several counterexamples in my career that, to be honest, boggle the mind. I have seen globally renowned agencies have local estimates and forecasts so poor that you don’t know if you have to cry or laugh. I have seen global agencies, reference in their sector, produce consumer questionnaires so weak that the output ended up being automatically inaccurate. Very inaccurate (more on consumer-specific data in part 3). And worst of all, I have seen globally renowned agencies have ethical standards so low locally that it was common knowledge in the industry that their data was fabricated (gasp). Such agencies ended up starring at the top of my black list of sources to never, ever use, even under torture.
Beware the Middle East! Whenever “Information is power” and competitors’ figures are not readily available, estimates are, at best, indicative.
In defense of a lot of these agencies, research, and consulting companies, the GCC and the Middle East at large is a tough region to analyze. The reason is all the financial figures of big groups that are readily available in the West are not here, or at least not for many categories. The proportion of actuals vs estimates is much lower compared to the West leading to numbers that can be downright funny. I once attended a presentation by a major research company (respectable by all means) that wanted our business by making us subscribe to their reports. The only problem was… they had near to no actuals of our specific field while the position of the group I was working in gave me access to 30% to 50% of actual market figures. Needless to say, our estimate was de facto more accurate.
Newcomers to the Middle East need to understand that competitors don’t like to showcase or share their figures. The standard practice in the West whereby all competitors provide their figures to a third party and the understanding that this knowledge of the category is beneficial to all competitors is not common practice here nor commonly understood.
“Information is power” and as such nobody likes to share with their friendly neighbor. So third party data have always to be taken with a pinch of salt (sometimes with several handfuls of it) because the common practice is then to take global figures (because they are available) and project onto the Middle East according to the proportion, growth, trend worldwide. Even if you are not an expert of the region, you can see how that can lead to a lot of inaccurate conclusions. A lot of brands perform much better or much worse in our sunny part of the world, market dynamics are different, consumers are different making those linear projections almost unusable.
I recently worked with a report of a source previously unknown to me and asked a very simple question: “what is the share of actuals vs estimates in your figures and projections?”. The answer was so obscure and convoluted that I automatically deduced that they had very little local input and most probably were using the method described above. Sometimes you have no choice but to use this because you need a reference point and this is better than nothing. The best you can do is to cross-check with other sources (that will most probably use a similar methodology) to refine your understanding. As general rule categories in FMCG or that have a share in FMCG have a better chance to be close to reality because it is one of the most transparent fields (big players like Carrefour or Géant disclose their sales by brand on a regular basis).
Beware governmental figures, the methodology of data gathering can change overnight!
The example I will share with you will lack the relevant details because I seem to be affected by an early Alzheimer and forget things the moment they are not useful to me anymore… However, you should get the gist of it. Back in 2010 (I think), Dubai tourism figures witnessed a significant increase 2009 to 2010, to reach 10M in 2010, 15M in 2015, and 20M in 2020 (now these figures are 100% exact, I remember them because… well it was really easy to remember). How? What? Why? It actually took us a lot of investigation to realize that Dubai Tourism had changed its method of calculation switching from a number of hotel guests to a number of incoming tourists at the airport. By all means that new method made more sense as there is a sizable portion of tourists that are visiting family and therefore living with their relatives, not in hotels, however with no explanation, wrong conclusions could have been drawn as the increase was major (forgive me for forgetting the actual figure).
Conclusion: Beware the sources that are obscure and vague about their methodology, ask them questions about how they came up with their wonderful figure.
And if you have to use them, for lack of a better reference point, do so cautiously.
A little side note… Use the same caution when ordering a report. Don’t trust blindly the extensive table of contents and ask questions about what you will actually receive. No later than last month, I was about to order a “GCC report” that seemed to be also giving specific figures about the individual markets. Wrong! If I had bought it, knowing that I only needed the UAE at that point, I would have been served with overall regional estimates that are de facto less accurate than local estimates (estimate of an estimate) and would have wasted $1,000 for something unusable.
My examples in Part 1 and Part 2 focused on macro-economic figures and industry data, allowing me to move on to consumer data in Part 3… My favorite subject!
Photo credit Alina Grubnyak on Unsplash
Author: Nadine Touma