How to Stop Drowning in Data

Nik Nanos cares what people think. In fact, he’s built his whole career on it. Nanos is the founder and chief data scientist of Nanos Research, a prolific Toronto-headquartered firm that gathers and analyzes public opinion on everything from consumer confidence to political sentiment. 

The company has earned a reputation for cool-headed clarity and useful context; it’s the pollster of record for The Globe and Mail, Bloomberg and CTV. Not bad for an outfit Nanos started in 1987 as a side hustle during his undergrad studies at Queen’s University. (He also holds an Executive MBA from Smith School of Business.) 

As someone whose work hinges on accurate and effective interpretation of information, Nanos knows his way around a data set. It’s a skill of ever-increasing economic importance, as business leaders of all stripes figure out what to do with unprecedented flows of facts and figures. 

On the heels of his recent appointment to the Order of Canada, Nanos spoke with Smith Business Insight contributor Deborah Aarts about the power of unique data, the value of outside perspective and the risks of chasing quantity over quality. 

So far, Nanos Research has survived 11 federal elections, three major recessions and the dawn of a new age of unrelenting digital transformation. To what do you attribute the company’s longevity? 

When we started the firm, we really thought about our brand. We defined who we were as a business, and what our values were, and where we wanted to be. And we looked at the competitive environment.

We decided we wanted to be “the bank” of market research. What did we mean by that? Well, you can always trust the bank with your money. When you’re the bank, people will wait for you to give the final word before something is done. We would not be as flashy in the media as others. We would stick to our knitting in terms of the quality of our work and our consistency over time. 

That was the first big decision that we made related to who we were, and that drives through the whole organization to this day. 

What does that brand promise look like in practice? 

We always believe that part of being “the bank” is to only speak to the numbers. 

One of the things that journalists did not like about me at the beginning, when I started doing media interviews, is that I would say something and then I’d stop. And they’d say, “We can keep going.” I’d say, “Well, I don’t have more data; this is the data that I have.” They’d ask: “Do you have an opinion on this?” And I’d say, “No.” I wasn’t very popular at the beginning because of that. 

The world has experienced a lot of technological, political and economic change in the past four decades. What has made the most notable difference in your work as a pollster? 

The biggest change is the velocity and pace of work. In the 1980s, 1990s, even at the start of the 2000s, projects would take weeks or months to do. Our work was much more labour intensive and our computers much slower. Clients didn’t have any expectations that the work would be done quickly. Today, they do. 

There is just so much information available now, and it comes faster than most humans can process. How do you think business leaders should be thinking about data? 

We believe that the future is about high-velocity data that is unique. Information now needs to be a feed, not necessarily a fixed moment in time. There’s a lot of data out there in the public domain, especially in the financial markets. But if your business has unique data that is your own — that is not out there for all to see — it can hold significant value.  

When you have the frame of mind that unique data is valuable, it changes the way you do your business, how you collect your data and how you add value to that data. The reality is that revenues can go up and down. But data is like money in the bank. 

What are the risks to businesses of having access to so much information? 

Companies can get wowed by big numbers. Just because a data set is large does not mean that it is reliable or accurate. It just means that it’s large. 

Often, when I’m brought in by a company, I’m told something like “We’ve looked at a million cases.” They have technologists or engineers running the numbers, or they’re running machine learning algorithms — which we do use — through supervised and unsupervised models. The companies run machine learning but don’t know what it means to have representative data or how to properly interpret the information. They’re not thinking about the questions that are being answered.

As a result of these methodology issues, they often can’t answer simple questions about their data, and they make decisions based on information that isn’t really reflective of reality. It can do real damage. 

So, does that mean executives need to dig up their old stats textbooks? 

One benefit of my career is that having done thousands of market studies, I feel like I’m a walking, talking repository of why businesses succeed and why they fail. Because I’ve seen it all. And I can tell you what I believe to be the best practices for the executives who, in my opinion, have been the most successful. 

All firms should now have in-house data capabilities. At the very least, they should have people who are savvy with data sets. Not everybody has a statistical background, but I always say, “We can teach the stats that people need to learn, but we can’t teach someone who’s afraid of math.” 

And I think all firms can get a lot of value from an external data adviser. Many times, people within an organization are too close to the information, or they know too much, or they know enough but don’t have context, or they think they have context [but] are missing things. It can be extremely helpful to have an outsider come in, even if only once a year, to say: “Here’s the value of the data that you have. Here’s how it should be structured and here’s how you can access it.” We have clients for whom we don’t generate any original research; we just give advice on their own data. Then their own team can run things after that.   

Data can sometimes show big shifts in public sentiment that may or may not require a company to act. When this happens, how do you recommend business leaders determine their response? 

When there’s an issue or change happening, executives tend to go right to “What is our position?” or “Should we have an ad campaign?” It can be easy to feel pressure to get involved, especially with social media. 

I think it’s better to start with the simplest of questions: Is this issue related to your core mission or not? If it’s related to your core mission — the quality of the product — then you have to respond, and you have to do so hard and fast. But if this is an issue of something happening that people are talking about, but it’s unrelated to your core mission, then it doesn’t necessarily require you to rush to react. You should not confuse what people take an interest in with what’s important to the success of a business. 

After all this time, what is it that still fires you up about market research? 

I have an inquisitive nature. I ask lots of questions. Many times, I know that I don’t have the answers, which is why I ask. 

I’m an expert at what I do, but I still learn something new every day, because at any one time, the firm is doing like 40 to 50 projects. So, I have a buffet of things to learn about. I can decide whether I want to learn about public transit use one day; the next day I can learn about medicine; the day after that, telecommunications. 

When you’re a lifelong learner — which I consider myself to be — to have a job where you learn something every day is just a great fit. It’s never boring.