Marketing research and the need for speed: interview with SSI CEO Chris Fanning
Marketing speaks with SSI CEO Chris Fanning about technology, speed and data management in this wide-ranging interview on the marketing research industry.
SSI CEO Chris Fanning’s background and training is as a software engineer. “Not your typical market research route,” he says. But, he also says, SSI is not a market research company.
“We’re a data management company,” he says. The board’s decision to hire him was an early recognition that technology was important and would increasingly become so. In the four year’s he’s been in charge, SSI has more than doubled in size and operates the world over. Next year will be the brand’s 40th anniversary.
Marketing spoke with Fanning about speed, technology, recruitment and data.
Marketing: Your company’s footprint is one of the largest in the field; how do you define what you offer?
Why do you buy certain goods and services? What might you do in certain situations? People want to do that quickly but accurately, with a high degree of confidence that the data is representative because you’re basing big decisions presumably on the results of those surveys or other data collection techniques that we use. If you think about Australia, you need to have a big presence in the people you have access to. We do that today in 100 countries. If somebody needs to find out something about Australia or the States or Japan or Indonesia or Italy, we can do it. It’s very unusual but we’ll do studies sometimes that involve 60 or 70 countries in one project. Mammoth world views of things, but we can do it. We’re actually a pretty unique person to go to for situations like that. That’s the product side.
On the people and service side, you want to have people on site who can sell and help people with solutions, Australian companies doing research in Australia. Australia does tend to be a fairly inward looking country. I remember when I was at BCG I worked with a number of Australian and New Zealanders who used to say that in Australia – just because of the distance – there’s not a lot of people who can be competitive here in any hard goods because of the transportation costs.
Maybe today with IT it’s a little different, and every major industry is a duopoly. Those are good rules of thumb and I find those to be still true as I visit and it’s just the way the country is structured and the way it works. It makes sense. We have people here who are selling and servicing local customers. We have people in Sydney, people in Melbourne, people in Auckland. We do that in a lot of countries around the world. Not 100 but we have people probably in 35 countries, physically. Usually in the biggest economies.
We work all over. We’re sneaky big. People don’t realise, we probably have 4200 people around the world. Some of that is we have sales and service people in local locations. Then we have teams of people who do the service side of the business and they’ll be distributed in India, the Philippines, Romania, Hungary, Bulgaria, lower cost, but really good talent parts of the world. Then we’ve got to sync them all up somehow.
M: How has the tech side of things changed in your time at SSI?
CF: When I came in, I inherited a good company, a great brand name. One of the ways I think about it is the very unglamorous but very necessary infrastructure. You have a global company and you have communities of people throughout the world, and then you have someone who needs to help a client. They might call up and say, ‘Here’s our need, it’s in these different countries, here are the specifics, can you guys do this or not?’
Within hours we can say yes or no. That’s because we have a system that allows anyone in the company to very quickly put in some information from the client. That’s really important because I think speed is the dimension that the research industry is, in general, not particularly good at.
If you’re a brand manager, how do you synthesise it and turn it into good results? Not everyone really has the time or the ability or the tools to do so. We’re increasingly able to provide access to people. If you’e a brand manager and you’re trying to just do some very simple, but very important work that requires a very quick time dimension, we actually can do that very effectively. What we don’t do is the, ‘let’s think about rebranding the company and let’s go talk to a lot of people and let’s have a three month study with a big report out at the end’. We don’t do that and we’re never going to do that. That’s what an agency would do.
I hear people talking about speed more and more. I know when we deal with clients, we win business because they’ll send us a request and within hours it’s, ‘we can do it’. By being more action oriented we allow them to get their work done faster, which should allow them to be more effective at their job. We’re trying to help people grow their business.
M: The data-as-a-service (DaaS) space is growing, and it may be distinct from what you do, but how do you look at it?
CF: I would think about in two ways.
The first way, I think, is around people who have a syndicated data product. Oftentimes, people that have those types of products are using us to do data collection. That’s okay, whether they share that with people, that’s up to them. I don’t think there’s anything evil if they don’t. I think most clients, even if they outsource or aggregate, some people are very open about it. Some people like to come off as the full service, we do it and if they’re using SSI to help, why do you care as long as we deliver the goods? That’s fine, we’re good either way. That’s one way it manifests itself.
The other way is for somebody who has their own data management platform (DMP), we can match data. For example, in the States there are companies that do credit ratings. There’s multiple companies that do that. There’s a lot of information on me and if I have a good or a bad credit rating. We don’t typically have that data on people but what we can do is we can take our panel and match it. We have to have an agreement and you have to have the technology to do it, but we can match it with, for example, for one of the credit bureaus. You’re not going to have a 100% hit rate of course, but like a Venn diagram there will be overlap. For those people who overlap, we can augment our panel data onto the existing data.
One of the benefits we have is, I think, owning the data asset long term. That will be very valuable because there’s a lot of different use cases you can come up with. It’s expensive, it’s hard, I think most people would agree. It takes a lot of people to do that well, and if you’re doing it in a lot of different countries, every country is very different. How you recruit people here, versus the US, versus Japan, is very different. You have to know the nuances of the different markets as well. It’s a high barrier to entry business.
M: On recruitment panels and data collection, how has technology changed that end?
CF: If you looked five years ago, 80% of how panels were recruited would have been through affiliate networks: ads on websites. People click, come in, they join your panel, take a survey. That’s how many people operated 80% of the time five years ago including us.
Today, that’s probably 20% of how we recruit. I’d say the world is still doing it the old way, because when we go into big clients and we show them how we’ve changed over time, they look like we did five years ago. Tthat’s because they’ve been doing it the same way for a long time and don’t have different people who have come in with different perspectives.
We’ve opened up a lot of different ways. People have loyalty programs that we recruit from. We use things like YouTube influencers. Bethany Mota is a young woman in the US who started off, I think, as a 16 year old who would give fashion tips and life tips to teenage girls. I have a teenage daughter. When I mentioned that name to my daughter she said, ‘Oh, my god, everyone knows Bethany Mota!’ She is a multi-million dollar brand on YouTube, who has fashion deals with US retailers now, and she’s quite a business sensation at the age of 20.
From there, you can get that demographic and interest group to sign up. How we do it, if you think like a classic marketer, you’re segmenting the market and you’re having much more specific targeted advertising and recruitment versus just broadband, in the old days when it was US three TV channels and you just ran ads on all three and you covered pretty much the country.
Now with the fragmentation of media you’ve got to be very targeted. We can do that because as we create these separate communities to come in, they’re a panel, and then what we can do is all these panels can get joined into one giant virtual panel. That’s because of our tech stack. Without getting too geeky, it’s like database joins. You could have one giant database with all the old and the young and people who love cars and foodies, and you just push them all into one panel. That may limit how appealing you are to any one particular group.
Or you can go after very specific groups with very specific rewards that resonate with that particular community and then you can somehow connect those all together. Which we do. Sometimes you’ll get somebody who is into food, who might be an IT decision maker who is Hispanic and who is a female. Those four might be the same person. If you think about what we do, we recruit all those different types. How we get you really doesn’t matter, as long as we get you into our panel. Once we get an IT decision maker to join our IT decision making panel, we find out over time, that they’re female, Hispanic and a foodie. We get all that information over time by profiling you. You just happened to come to us through some IT recruitment channel in the US that we opened up.
Then we get to know you more and more over time. What we find is, as we get to know to you more, there are two things we try to balance. There is the representation of the population that people come to us looking for, so obviously you’re looking for unbiased views of the world and you’re looking for people who are willing to share information so you can predict things very accurately.
Over time we also find that if you give your opinion on a subject that actually you care about, you tend to give more thoughtful answers and the quality of the data is richer. If you were to ask me about sports, I’m a huge sports fan, I really think about it and really give you an informed opinion, versus if you were to query me on opera which yeah, I’ve been, I enjoy, but I’m certainly an ill-informed individual on the topic.
By knowing people, and again our tech platform allows us to match, so we can actually say, ‘Okay, a client needs two more 20-year-old females that live in Melbourne, and we have 30 available. What is the survey about? Are there two people that might actually be a better opinion to get?’
That’s part of the technology value we bring. Part of the reason our quality of results are so high is because we’ve got a lot of smart software doing a lot of this. It’s very scale oriented, so if you think about it, as we get bigger, because we have more people who need to use us, we have demand. As our panels grow we have supply. If we match them quickly, almost like a stock market, the faster we match the quicker people get results, and the smarter we are about who does what and how engaged they are, and over time you keep the panellists active. That allows you to continue to grow and deeply profile people over time.
Then as that data asset gets bigger, you can start overlaying things, like the DMPs we talked about. I think that’s in the very nascent stages but I think five years from now that will be very common. It’s going to be more common, one you’re going to need the technology, the capability and the infrastructure, and then if you’ve got a match, would you rather match with a 20 million person data management company or a one million person data management company? It’s a numbers game at some point too.
M: How do you account for some of the biases that naturally come with surveys, like people who take surveys for the sake of taking surveys, or the fact that 100% of the population is not on the internet?
CF: On the former, when I first started there was an urban myth that there are 10 people in the world who do all the survey taking. I refer to them as professional survey takers. There are people out there who, I think, if you let them, would take surveys every day, multiple times a day. There’s not that many, but there are some.
We’re pretty good at avoiding the professional survey taker. We also have found, because we’ve bought companies, when you look at their panel and our panel and you look at the overlap, there’s always this fear it’s going to be very high. I forget the exact numbers, but when we bought a company last June in Australia called MyOpinions, the overlap was very low. Which is great because, when you put the two together, you get a much broader reach of the country.
Regarding people who don’t have access to survey vehicles, that’s always a challenge if people aren’t online. You have tablets and smartphones which pretty much everyone has over the age of five and what we’ve done is we’ve invested very heavily the last couple of years in mobile technology. We have panels that are smartphone oriented. There’s a little bit of a difference in how you present things to people on the phone versus if you’re home on a laptop or something.
We have all these different panels that we use for targeting purposes – same thing on mobile. We can actually combine the online and the mobile worlds into one big virtual panel. In the mobile world you’d tend to see a younger audience than online and a more male audience than online. That’s fine because when you’re combining and blending them together you get even more of a balanced, robust representation of the population.
You can also do some specific things with the mobile technology as well, like geolocational stuff. ‘Is this person at the Australian Open? Have they been there for more than five minutes?’ If they have been we’ll assume they’re there to see some matches and we might want to ask them some questions right then and there, or maybe later.
We also call people if we have studies with really tough-to-find people. Sometimes we’ll have to literally use our phone capability and dial people, which might be someone who doesn’t have internet access and doesn’t own a smartphone. But with that you’re talking about an increasingly smaller segment of the population.
M: So if someone is launching a product for hermits… Although the other end of it is people who are really high value, very time-poor decision-makers in a business. Getting the attention and especially getting action from those people is very difficult. My assumption would be they rarely take surveys. Is that correct?
CF: You’d be surprised. We’re all consumers. If I wanted to, I could give my view on a toothpaste survey. I also just happen to be the CEO of a tech company in the States of a certain size. I would be an example of someone we would profile over time, so you’d know I’m middle aged, caucasian, living where I do and married with three kids and all my personal statistics, which might be valuable if you’re doing car purchasing or home purchasing or things that are more consumer oriented.
Then, if you’ve also successfully profiled my professional life, I could also be targeted to do surveys regarding IT purchasing decisions or health benefits for employees for a company of our size.
I’d be a scarcer resource in the latter bucket, so we would typically offer a higher incentive to people. Sometimes the incentive could be a sanitised summary of some of the results to the participants. We tend to be very, very, very good at the low incidence work because we have such large panels.
That’s the beauty of a data company because it’s super scalable and the more data you have you can do stuff easier and do more difficult stuff. If you’ve got the technology to be able to know you can do it and then do the work and then deliver it to someone quickly while maintaining all the quality standards, you have a nice business model.
Image copyright: adiruch / 123RF Stock Photo