Welcome back to TakeFive with TweetReach, our ongoing interview series with influential members of the Twitter measurement universe. This week, we’re excited to speak with Jim Sterne, an early Internet Marketing adopter and a fantastic resource for breaking down big ideas in ways that make them easier to understand, which made us thrilled to get his point of view on social media marketing and more.
TweetReach: We like to start everyone out with one question, because there are so many different paths into social media: how did you get started using social media? Can you describe your first “ah-ha” moment?
Jim Sterne: The first “ah-ha” moment was in 1992 when a friend showed me a chat room for the first time. It was on CompuServe and it was what you’d expect– the equivalent of a private Twitter session. Maybe 20 or 30 people doing one-line entry, hitting return and their message pops up. There can be three or four conversations going on at the same time, usually very inane chatter, but broken up – not threaded into subject matter. My first thought was, “Why would anybody want to do this?” My friend said, “Well, you know, you can find out. We’re going to meet down by the beach tomorrow”.
I had no intention of going, but my wife and I happened to drive by and there was a huge banner that said CompuServe and about 200 people on the beach. The big “ah-ha” moment was that even in 1992, even when this technology was so awkward, even with the conversation so banal, it had such a huge following. There is value here that clearly I wasn’t seeing, but it was there. So that was the wakeup call.
There are still people today who say, “Oh, I don’t understand Twitter. Why would I care what you had for breakfast?” and you have to explain to them all the different ways it can be used. It was the same problem for me – I looked at it and didn’t get it- and then finally, I did.
What got me into being very social was the High Tech Marketers Discussion Lists, an email discussion list on a listserve. Kim Bayne was the hostess and we just talked about the marketing of technology. But then, in early 1993, the conversation suddenly turned to, “How do you build a website?”. Kim said, “Look, we’re here to talk about marketing technology products, not building websites. So if you want to take that conversation somewhere else, go ahead.” Glenn Fleishman said “Okay” and started the Internet Marketing Discussion List, which ran for three or four years, and I’m sure the archives are out there somewhere.
That’s where I truly understood the real value of asking a whole bunch of people- random people that you’ve never met- a question, and getting really good answers back. That, and being careful how you present yourself: personal branding.
That’s how I got started. Everything else: blogs, Twitter, Pinterest, etc., carried on from there.
TweetReach: Since 1994 you’ve been concentrating on how analytics can inform marketing decisions. How have you seen the role of analytics professionals evolve from a pure web analytics focus to starting to envelope and include social media metrics?
Jim Sterne: Web analytics was the first data set we could get our hands on: Log File Analysis. As analysts became more respected, they were asked to do more things. “Oh, while you’re measuring the website, we also want to know whether email is driving traffic to the website and how well search is going. And, oh, by the way– when we run an ad in the newspaper and we put a URL in there, does that drive traffic to the website?”
So the idea of being a web analyst was like being the webmaster who did everything, and then– no, one person can’t do all that. Web analytics is simply a deep, but narrow, datastream of “where did they come from/what did they look at/how long did they stay/did they convert/did they come back?” But when email, search, banner advertising, etc., came along, web analysts were responsible for measuring all of it.
When social media showed up, everybody turned to the web analysts and said, “You are measuring this, aren’t you?” And suddenly we needed a whole bunch of new tools. It was just yet another data stream. That’s never going to stop. Now we have to measure all the social media out there: Pinterest, Tumblr, Facebook, etc. Tomorrow we’re going to be measuring the Internet of Things. “How many of our 3D print designs have been downloaded and printed and in what colors?” We’re going to measure what your shoes say about your workout and whether that has an impact on whether you’re going to buy our product.
We’re the analysts and yes, go ahead, keep throwing new data at us– we can take it.
TweetReach: You recently wrote about Rod Bryan’s analogy of data mining and diamond mining, which made the point of finding the highest quality raw materials (data) and planning for the right setting (marketing/presentation/pitch). Getting this right will be a little different for each company, but do you have any recommendations on where to start looking, or what to definitely avoid?
Jim Sterne: The diamond analogy is really wonderful because it’s hard to get the diamonds and then you have to cut and polish them, and then you have to put them in a nice setting. Data just sits there by itself until you turn it into information. I’m six feet tall; that’s a data point. But in a room full of basketball players I’m short. Now it’s not just a measure; it’s a metric.
How do we make that useful? That depends on your goals. If my goal is to grow taller, and I’m still only six feet tall, then that tells me that I’m not succeeding at my goal. So that’s gone from a metric to a usable benchmark. It’s now knowledge. The real magic of doing analytics is when you evolve beyond knowledge to insight.
If everybody who comes to my website using this search term clicks all over the place and has trouble finding what they’re looking for, the insight is: maybe we should change our content on our website so we rank differently. Or maybe we should advertise to change what people are searching on. Or bid on different keywords.
That’s where the magic happens. It’s not “Hey let’s collect petabytes worth of data because we can”. It’s not “Let’s dig up this big data nugget; this giant diamond is going to make a huge ring!” That’s where the analogy falls apart. If the purpose is to find the giant, 25 carat diamond, then you’re going to dig differently and in a different place. If the goal is to find millions of diamond chips that you can then use in industrial purposes, you’re going to dig a different way in a different place.
The most important thing is not just to collect all the data in the world that you can, but to understand why you’re doing it and collect the data that is revealing.
TweetReach: You’ve also written about the power of predictive analytics. How do social media metrics play into the prediction of consumer behavior, and what might be some other insights we could glean from that data that you might not expect?
Jim Sterne: If you come to my website and all I know is that you searched for a specific keyword, well, that’s a lot more than I knew if you just typed in my URL directly. But if you stick around and then come back and sign up for my email newsletter- suddenly I have some personally identifiable information that I can connect to your Twitter feed and your Facebook timeline. I can start looking at a whole bunch of auxiliary data about you.
Now I can detect a pattern that suggests people who search on this term, who like that type of music, and enjoy bowling on the weekends typically have a higher propensity to click through on this offer than the other offer. Now this is not a 100% prediction, but the value of predictive analytics is when my guess is better than 50/50 — then I win. I can beat the odds. I’ll have made more sales. I’ll have made my customers happier. I’ll have turned more customers into advocates. Social media offers a new datastream that gives us a different type of information.
Essentially I’ve got these three different types of data that I’m looking at:
1. What do people do? What do they search for, what do they click on, how long do they stick around, how often do they come back?
2. What do they tell me, if I give them a survey? This is a customer satisfaction score. Are they happy, and what are they complaining about?
3. What do they say to each other? Which is pure branding. “I’m thinking about buying a bicycle, what do you feel about this model?” and then the community has a conversation.
That’s enormously valuable by itself as market research, but when I match it up with what you do on my website and what you tell me in the customer satisfaction survey, I’ve got these different types of data that I can use in aggregate to compare and contrast you to everybody else in my database to say, “Oh, well, you’re going to have a higher propensity to- for example – you’re more likely to click on two-for-the-price-of-one than you are for buy-one-get-one-free.” It’s the same offer, but it’s described differently and different people respond to them differently. If I can predict which type of person you are, then I have a better chance of making a sale. I do not need to try and to paint a picture of you– I really don’t want crawl into your head (it’s actually not valuable to come close to be creepy).
TweetReach: Your public speaking engagements are obviously tailored to each specific audience, but are there certain things you try to hit on when you talk? Measurement strategies that you think are important to analytics, for example?
Jim Sterne: Number one, know what you’re measuring for. If I collect all the data in the world about you and it doesn’t benefit you, I’ve wasted my time. If it benefits you by providing a service or making it easier for you to find what you want to buy, or discover that you don’t want my product– that’s a service. So have a goal in mind; that’s a critical piece.
The value is not in cranking out reports and dashboards; the value is looking at the data and saying “Hey, this is something interesting. I wonder…” And then start doing little tests: “I wonder if people who show up on my website on Tuesday respond differently than people who show up Wednesday morning. I wonder if people who came to me through this banner ad campaign are more likely to return product they purchased from me, and therefore I don’t want to advertise in that manner.”
It’s the insight, the creative part of the data that’s the most important.
Jim Sterne is an international consultant focused on measuring the value of the online marketing for creating and strengthening customer relationships since 1993. Sterne has written seven books on using the Internet for marketing, produces the eMetrics Summit and is co-founder and current Chairman of the Digital Analytics Association.