CustomerThink Founder/CEO Bob Thompson interviews Pelin Thorogood about digital analytics; why marketers must solve data silo problems; and the role Big Data and data scientists play in analytics success.
- Why Do Marketers Still Struggle with Data Silos?
- Example: Optimizing Marketing Spend
- Has There Been Too Much Focus on Big Data?
- Are Data Scientists Needed for Analytics Adoption?
Interview recorded February 4, 2013. Transcript edited for clarity.
Bob Thompson:
Hello, this is Bob Thompson of CustomerThink, and welcome to another episode of my Inside Scoop interview series. This time my guest is Pelin Thorogood, CEO of Anametrix, one of the relatively new players in the hot space of digital analytics. Today we’re going to be talking about how Anametrix is positioned, and some of the challenges that marketers need to overcome to create real customer insight. Pelin, welcome to Inside Scoop!
Pelin Thorogood:
Thank you so much, Bob. Happy to be here.
Bob Thompson:
Now, Anametrix is a fairly young company. As I understand, it was founded in 2010. Please tell me the story of why it was founded.
Pelin Thorogood:
Of course. A lot of us at Anametrix have deep roots in analytics. The two founders of Anametrix in 2010 actually are associated with another big analytics provider from San Diego, WebSideStory. Blaise Barrelet, who is the co-founder and current Chairman of the company and Anders Olsson, who is a co-founder and Chief Technology Officer of this company were very much involved with WebSideStory, Blaise as actual founder and CEO and Anders as Chief Architect of WebSideStory.
So, we’ve had deep roots in real time cloud-based analytics for 10 to 15 years. We were the very first provider of real time cloud-based web analytics in the world at the time with WebSideStory, and we founded Anametrix because the world has changed quite a bit from those days, of course. There are now more channels for customers to engage with brands, not offline and web, but social, the increased force of mobile in our world. We wanted to provide a multi-channel analytic solution for marketers, which is really what Anametrix is based on, is providing marketing analytics to enable marketers to have better insights into these multi-channel consumers.
Bob Thompson:
How did you become the CEO?
Pelin Thorogood:
I was the Chief Marketing Officer of WebSideStory, the company that I just mentioned, and I’ve been in the space of analytics for a very long time, myself, both as a practitioner and also actually just by educational background, as well. After WebSideStory went public and we got acquired by Omniture, who, of course, then got acquired by Adobe, I have been consulting to other cloud-based companies for the last two years. Blaise tapped me early last year to come join Anametrix and really to drive marketing here. Knowing the team here, knowing the pain marketers are facing in this multi-channel world, I thought it was an excellent opportunity.
I joined as Chief Marketing Officer earlier in 2012. As the company got its financing in the October timeframe, we got our Series A from TVC Capital, and as we have some great customer traction and getting some notoriety in the market, as well, both Blaise and the board thought it was time to transition the leadership. They tapped me in the December timeframe to take over and really carry the flag as a data-driven marketer to lead where analytics can take marketing into becoming more predictable and more accountable.
Bob Thompson:
Can you give me a quick idea about how Anametrix differentiates when there are so many analytic solutions in the market? There’s obviously some very big companies like IBM and SAS, but also newer vendors like Cloud9, Birst and other SaaS solutions. Why do we need another cloud-based analytics solution?
Pelin Thorogood:
Sure, that’s a very valid question. Like I said earlier, we are a marketing analytics company created by marketers for marketers. It’s real-time analytics to enable marketers to have a much better understanding of how this multi-channel customer is engaging with their brand, and to give them insights that they can actually act on. Our solution is really a full stack. So it starts with being able to connect to any data source, whether that data source is from the cloud, whether those data sources are coming from internal sources, such as your financial systems or marginal information, whether it’s reference data, such as U.S. Census or market share or shopper information. So, we connect to any and all relevant data sources, leveraging our own proprietary technology.
Bob Thompson:
Is it a type of cloud-based data mart that you’re using to collect all these different data sources, so that you can then run your analysis?
Pelin Thorogood:
Regardless what type of data, whether it’s streaming social data or if it’s structured data coming from an internal data warehouse, we can connect to all data sources and then we map and align these data sources, so that we actually understand that customer journey across these data sources, have the context of that share of shopper, share of market that’s information coming from reference data, and perhaps most importantly, have those internal data sources, the financial margin and sales-type data sources so that we don’t just look at what’s happening, in terms of your typical reach engagement acquisition conversion metric, but what it means in terms of true ROI. So, we’re able to really bring the outside and the inside world together to give marketers a much, much better sense of return on marketing investment and how that’s driving both top line, in terms of revenue, as well as bottom line, in terms of actual profitability.
Bob Thompson:
This sounds a little bit like you’re the digital marketing equivalent of another company I’ve been following for a while called ClickFox. They do multi-channel customer experience analytics. You’re really looking at it from a marketing standpoint.
Why Do Marketers Still Struggle with Data Silos?
Bob Thompson:
We’ve had CRM solutions and marketing software suites. Why are marketers still struggling to get all their data together in one place so they can do something with it?
Pelin Thorogood:
I think there are a couple of different reasons for that.
First of all, part of it is the fact that the number of data sources and the amount of data has exploded as we’ve all experienced directly. Second of all, I think organizationally, people have been working really in silos. So, even when social came along, there were social marketing managers maybe trying to optimize that, there were people in e-mail marketing, there are people focused on advertising. But unfortunately, these data elements, regardless of their size, have been living in silos. And I don’t think anybody would disagree when I say if you’re trying to optimize a silo, if you’re trying to optimize a branch, what you’re really missing is the bigger picture, you’re not optimizing the entire solution. So, part of it is the fact that we still have a silo approach to the problem.
A second issue is that marketers, in their attempt to become more data-driven, have really focused on KPIs and a lot of metrics, which is a good thing, except oftentimes, the metrics they’ve focused on are top line metrics that may not necessarily give you a full sense of what’s happening. So, even the metrics, themselves, have been silo’d and they have been more like a thermometer in telling you the state of the system versus more like a barometer as in what it actually implies for the future. So, I think the issues have been around the fact that marketers are by DNA, more creative than data-driven because they have been living in a silo world, as well as the fact that some of the metrics they have focused on in their attempt to become more accountable and predictable have not necessarily been the right metrics and have been only giving them the state of the system.
Bob Thompson:
There was a Forester analyst recently on a webinar that Anametrix sponsored who said that creating customer insight to drive decision-making was still a top challenge by little over half of the people in the webinar poll. So, this multi-channel problem has been around for a long time. It looks like it’s not going away any time soon.
Example: Optimizing Marketing Spend
Bob Thompson:
Could you share an example, perhaps of one of your clients, where you’ve pulled together this data from multiple sources and created some insight?
Pelin Thorogood:
One of our clients, a top three U.S. automaker, like with many other companies, was looking at this information in silos and was collecting the data and analyzing it over time to get a sense of how their various advertising, social and other marketing elements were performing. After we deployed Anametrix with them, connecting to both their marketing mix data sources, as well as to their internal data sources that actually provided information around sales, revenue, as well as margin information and the ability to, of course, slice and dice and segment this information so you can look at it not just as an aggregate, but actually get to look at it by region, by brand over time, what they were able to understand was how their various product categories were performing in various markets in real time.
Before they used our solution, they were using their incentive marketing dollars to ensure that given they do not have this real time information, nor did they have this detailed segmented information, they were really spending it more widely, spending it, hoping that their spend was going to ensure they were meeting numbers. Once they had this very detailed information, they were able to pinpoint their incentive spend on brands in regions where they were seeing that real time trouble that signified that, perhaps, they were not going to make their numbers, being able to really focus that spend, and in the end, saving tens of millions of dollars of incentive spend that did not need to be spent because if a brand was doing well and if a region was doing fine, that means that the incentive spend, which only erodes margins, could actually be kept in place. So, that’s one example of saving money and becoming far more targeted with real time segmented information.
Bob Thompson:
You’re talking about optimizing marketing spend, correct?
Pelin Thorogood:
Correct.
Bob Thompson:
What type of analytics is this? Is this predicted analytics, is it real time? Is this just slicing and dicing, or are you trying to understand a relationship between spending and ROI?
Pelin Thorogood:
It’s both. What we do is we collect all this data from any source and then we bring it together in our cloud to enable real time ad hoc slicing and dicing, real time ad hoc segmentation. So, all of a sudden, clients have immediate access, real time access to information from a whole lot of different places, so they know what’s happening in their marketing mix immediately, and they have a much better sense of what is performing, what’s not performing. So, what, of course, then happens is people have a much better sense of how their marketing spend has been performing to date and they have this information real time, the next question they’re going to ask is, well, what can I do differently? What kind of future is awaiting me? And do I like that future? Can I actually change that future? So, which is where predicted analytics come into play.
I like to compare to the Maslovian term: If you are looking for food and shelter, you don’t look for necessarily big comfort. But once you’re actually taking care of the basics, you can actually move up the chain. It’s the same thing, if you’re able to look at all your data in one place, understand exactly what happened from a customer journey perspective across all these channels in real time, then you can ask the question, “Well, what does tomorrow look like?” And then you can ask the question, “Well, I don’t like this tomorrow. How can I improve upon this tomorrow by leveraging predictive analytics and doing some what if scenarios on that by understanding the levers of change.
Bob Thompson:
Does your solution help build those models or is there somebody at your clients’ site, a data scientist, let’s say, who is charged with building the models to make predictive analytics actually work?
Pelin Thorogood:
Our solution does take care of that problem. We, of course, have data scientists in house. But we have built in several forecasting and other predictive models into our solution. We have integrations into R and SPSS and other predictive solutions out there. So, we’re not recreating the world. We’re leveraging those very sophisticated models but from within our application to enable the clients to forecast the future, based on this multi-channel data aggregated in one place, so you do have access to big data from all these channels. And then play around with the “what if” analysis to see what if I spend an extra dollar with Google AdWords, what would happen? You could actually get a better sense of how to allocate your marketing mix to achieve the desired results.
Has There Been Too Much Focus on Big Data?
Bob Thompson:
There’s been a lot of talk about “big data” over the last year or so. We’re talking about social, or it could be web stream data, video, … the list goes on. You can see these stats about the massive amount of the so-called big data that’s out there. But when you look at what your clients are actually using, are you finding that they’re dipping into this big data to make these decisions or is it still looking at a lot of the relatively common but fragmented “small” data?
Pelin Thorogood:
One of my conversations with a well-known analyst from a big analyst group said it well. There’s all this talk about big data, and of course, we connect to the big data sources, but there is so much more interesting information to be gleaned from small data if you could only connect the dots. I am not by any means putting down big data. Like I said, we connect to streaming Twitter feeds and any kind of streaming data to actually get the information from there but the reality of the situation is there’s still a lot of small data silos that many companies are living in. So, connecting the dots between these small data points still provides a tremendous amount of value, and that overlay some of the big potential data on top of it. I think that’s where the value is. You can’t just jump into the big data if you haven’t solved the small data problems, which many problems have not.
Bob Thompson:
That’s very well put. I’m coming around to that conclusion, myself, after doing some other research here. It just seems like there’s tremendous opportunity that’s been missed in a lot of companies. Big data has the sex appeal, but you don’t necessarily need to go full bore on that to get value out of analytics.
Pelin Thorogood:
We can certainly go in and understand trending, millions and billions of lines of Twitter feeds to understand your sentiment analysis and that’s great. But if this company doesn’t know how yet to analyze their paid marketing spend or how it’s impacted by overall earned media activities, then perhaps going into a Twitter feed to understand sentiment may not be the first step that they should be taking.
Bob Thompson:
That’s a good point.
Are Data Scientists Needed for Analytics Adoption?
Bob Thompson:
Let’s turn to a final topic, and that is the leadership and organization support. It seems like one of the jobs of the so-called data scientists within a company ought to be to help the leadership figure out well where they should be looking, and not going overboard and exploring big data when there are other sources that need attention. Do you agree with that? And, could you expand on what this data scientist person should be doing and the role of top management, as well?
Pelin Thorogood:
It’s a very, very interesting question, and I kind of liken it to how we needed online marketers in the marketing department 10+ years ago because everybody was otherwise too creative and didn’t really know how to deal with the data. This is the same thing. It’s part of the evolution.
I think as the new world that we live in, which clearly whether if you’re in marketing or other departments, it’s all about data these days, we all, as individuals and as corporate citizens have to have two dimensions. We have to know our business and operations and our business objectives, so we need to be business people. But we also need to be data people because business is so much more based on real time data these days. So, I think data scientist is one of the situations where we have to put people who are very comfortable in the world of data with statistics and regression analysis and being able to really make sense out of data. But in many cases, they may not have enough business acumen to necessarily use it in the right places. So, what we really need is a balanced approach between a good understanding of business and a really good understanding of those data points to leverage it to drive our business.
Bob Thompson:
I’m hearing some commentary that the lack of data scientists may actually be a problem for the data analytics industry more broadly because if you don’t have these people, you can’t make sense of it. I wonder if that’s really true. Is the problem that the technology is not usable enough yet? It’s very complicated stuff. Do you see a time five, ten years from now when a normal business manager can use these more advanced analytical tools on their own? Without necessarily having to go get the Ph.D.’s who know how to do regression analysis and predictive models and so on.
Pelin Thorogood:
It has to, and for those companies who don’t actually achieve those goals, they’re not going to be in business. Technology has to be accessible. Technology is not for technologists anymore. The cloud has made technology available to everybody, so we’re not dealing with IT departments, we’re dealing with marketers, salespeople, people in HR – none of whom are about technology. They’re about solving their day-to-day business problems.
Between technology and the insanity of the explosion in data, I really look at it as a solution of having to have data scientists to come in to help these people, but really at the end of the day, technology and business processes have to align. Solutions have to become far more accessible, and I think on the other side, people have to become more comfortable with data and a different way of thinking so that everything gets aligned. I think data scientists will probably go onto the next challenge, whatever that may be. I really look at technology having to be more accessible to people and for people to be more embracing of that technology, leveraging better processes, better UI and overall better balance between data and business process.
Bob Thompson:
How long do you think it’s going to take for the industry to kind of remake itself? And for that matter, as you pointed out, for business people to become a little bit more data savvy? I agree with you. I think that’s where the things need to go. How long do you think it’s going to take to get there?
Pelin Thorogood:
I think which industry we’re in and which vertical we’re in, which type of job function we have is going to determine that answer. It’s a Geoffrey Moore adoption curve. I think there’s going to be certainly businesses where people who have been more open to using the cloud, people who have been more on the forefront of technology – let’s say retail, sports marketing, automotive, some of these businesses have been more ahead. Other business, perhaps financial services have not been as far ahead. I think it’s going to depend on which vertical we’re talking about, as well as the job functions within those companies, but I expect that we’re going to be in kind of that mainstream for a lot of businesses no later than the next five to ten years because once again, we must.
I feel like there’s a big acceleration. Cloud is being consumed in the businesses, and there’s certainly a big focus on big data and analytics and trying to make more sense out of it. There’s a lot of focus on this and I think technologies are retooling themselves to become easier to use, become more accessible so that that adoption can further accelerate.
Bob Thompson:
Pelin, thank you so much. It’s been very interesting talking with you, and I appreciate you sharing your views with me on Inside Scoop.
Pelin Thorogood:
Thank you, Bob, it’s been my pleasure. I look forward to staying in touch going forward.
