Jim Sterne is an international speaker on electronic marketing and customer interaction. A consultant to Fortune 500 companies and entrepreneurs, Sterne focuses his twenty five years in sales and marketing on measuring the value of the Internet as a medium for creating and strengthening customer relationships. Sterne has written seven books on Internet advertising, marketing and customer service including, Social Media Metrics: How to Measure and Optimize Your Marketing Investment. Sterne is the producer of the international eMetrics Marketing Optimization Summits and is the co-founder and current Chairman of the Web Analytics Association. Sterne was named one of the 50 most influential people in digital marketing by Revolution, the United Kingdom’s premier interactive marketing magazine and one of the top 25 Hot Speakers by the National Speakers Association.
Key Points
Jim provides great perspective on how we should think about Social Media Analytics. The major takeaways from this discussion are:
- Social behavior is intrinsic to the web. Today it might be Facebook and Twitter, and tomorrow it might be something else. What matters is that the power to communicate exists.
- Reach, frequency and recall are basic metrics as always, but we also need to look at new metrics such as sentiment and viral behavior (such as shares and retweets).
- Relevant influence is another interesting metric. Oprah may have millions of followers, but may not be of much help if she tweets this analytics interview.
- As always, tying metrics back to business goals is critical. Retweets are not conversion goals. Sales / conversions, whatever form they may take, are.
- Before you do anything take a baseline, and know what types of metrics you are trying to influence.
- Once your baseline is in place, determine what your variables are. Do you want to try promotional messages at different points in the day? Through different social channels? Setup interesting tests so you can learn and improve.
- Beware the unexpected. One campaign may fail because it is a day filled with major news, and another may succeed because of a slow news day.
The Need for Social Media Analytics?
Eric Enge: Congratulations on the Social Media Metrics book. Why is social media so important?
Jim Sterne: Let’s go back to the Internet circa 1993 because, even then, it was obvious that social media was going to be important. It was a new, terrific, wonderful way to communicate. Most individuals agreed with this except for the skeptics who were concerned about their budgets and said “my customers aren’t on the Internet, it’s too early, and it’s not proven.”
Those of us who’ve played around with it know that social media is here to stay. Humans love to communicate so anything that makes communication easier, faster, and better wins. That is why the telegraph, telephone, fax machine, and email have all been popular tools.
Social Media is an easier way to communicate and, therefore, it will never go away.
I can text thousands of people at once through Twitter. I can communicate with my friends through text, and pictures, and video, and links to many things on Facebook. It’s an easier way to communicate and, therefore, it will never go away and companies need to figure out how to make the most of it.
Eric Enge: Social media has definitely had its skeptics. One of the things that changed is that Facebook, for example, has proven it can make money.
It doesn’t matter if it is Facebook, Twitter, MySpace or Friendster. What matters is the ability to communicate exists.
Jim Sterne: We are missing an important point which is that it doesn’t matter if it is Facebook, Twitter, MySpace, or Friendster. What matters is that the ability to communicate exists. Asking which platform you should be on is the same as asking which television network or magazine you should advertise in.
The question is not, “which is the better place?” because that’s a matter of test and measure. We can measure it, see whether the investment is paying off, and make incremental improvements. That’s why metrics matter.
What to measure in Social Media
Eric Enge: What are some of the unique challenges we have with measurement in the social media space?
Jim Sterne: The first obstacle is convincing the people upstairs that it’s a good idea. The next challenge is to understand that many of the old methods for measuring continue to be used.
Let’s start with advertising. If I spend millions of dollars to produce an ad and run it during the Super Bowl, how do I know it’s worthwhile? I will measure awareness with surveys. Have you heard of my company? Do you remember seeing this ad on the Super Bowl? How do you feel about my company compared to the competition?
I’ll ask those questions on the phone, in face-to-face interviews, in shopping malls, and on online pop-up surveys in order to benchmark the results. I will then run my ad campaign and follow it up with another survey to measure the difference. That’s classic advertising and, for the billions of dollars spent on advertising every day, that’s how it’s measured. In the end, we trust it will turn into sales, and sure enough it does.
That’s standard awareness. I keep track of my reach and my frequency and do surveys to determine whether it’s had an impact. I then measure response as a metric. If I Tweet something, or put something on Facebook, or build a viral video, does it get click through, page views, engagement, conversion, sales?
That is straightforward. However, we have some unique, challenging metrics that we haven’t seen before, and vendors are building tools to help us solve these challenges.
Let’s look at attitude and sentiment. We can measure how many mentions we get in the blogosphere, and on Twitter, and how many re-tweets there were. I can measure the quantity of visibility; however, quantity alone is not helpful unless I also tie in the attitude or the sentiment.
If my name is Anthony Weiner or British Petroleum, and everyone is talking about me, this is not necessarily good. I need to know if it is a positive, negative, or neutral conversation that’s occurring.
Next, I want to measure influence. For example, if Jim Sterne said “this is the best thing since sliced bread,” that has the potential to influence the thousands of people who follow me on Twitter. If Oprah tweets it that’s a different level of influence because there are millions of people who follow her on Twitter.
Eric Enge: Yes.
Jim Sterne: If I say a particular social media metrics tool is head and shoulders above everything else, chances are excellent I will influence more people than Oprah will. This is because the millions of people who follow Oprah most likely don’t care about social media metrics tools; however, everybody who follows me does care.
Eric Enge: That’s going beyond total reach and into measurements of the relevance of the audience.
Actual influence is a fuzzier metric to keep track of, but it is critically important.
Jim Sterne: Oprah has a lot more influence on the subject of books, not just reach, but actual influence. This is a fuzzier metric to keep track of, but it is critically important. When I see a huge spike in quantity of conversation and a huge spike in the positive polarity of that conversation, it’s great.
So, I have standard advertising metrics and my new social metrics. I tweeted, somebody re-tweeted, and somebody else re-tweeted that. Somebody blogged about it, and somebody else commented on that blog. Somebody created a video satire of my ad and put it on YouTube. Those are social interactions.
My business metrics measure response. Did they click through or did they buy something?
Tying Social Media Metrics back to the business goals
Eric Enge: How do we tie this back to whatever your organization’s goals might be – sales, leads, getting a story out, exposure, or brand mentions. Even though we have these new types of metrics to measure, we don’t want to simply generate a lot of social activity.
Jim Sterne: That’s called noise; we’ve generated a lot of noise.
Eric Enge: Yes, and we still want to drive real business results.
All metrics start with business goals; otherwise, there is no point.
Jim Sterne: Exactly. All metrics start with business goals; otherwise, there is no point. I can measure the number of Twitter followers I have. For example, 5,761 is that good, is it bad? It’s a number.
It doesn’t mean anything unless I turn it into a metric. You turn a number into a metric when you compare it to another number.
Is it more than I had yesterday? Is it a lot more, a little more, is it as many as I was hoping for? Am I hitting my goal compared to yesterday? Is it more than my competition? Is it fewer than I had before?
Now it starts to be a metric. So, I’ve got my measurements which include re-tweets, and I had umpteen re-tweets yesterday. Is that good? Compared to what?
I have to turn that into a metric, and those metrics have to be useful towards my key performance indicator. They tell me whether or not I am getting closer to or further away from, my business goals. These metrics are fascinating, but I must not get distracted doing things that aren’t useful for my business.
Measuring a Product Launch effectiveness
Eric Enge: Let’s look at a specific example. How does social media metrics play into a product launch?
Before you do anything, take a baseline.
Jim Sterne: Before you do anything, take a baseline. What is the background noise? What is the rate of increase for the attention and the mentions you are getting now? Is the direction of the polarity of the sentiment positive, negative, or neutral? You need to get these numbers in place to figure out what’s the starting point. What’s the benchmark.
From there, you will reach out to people whom you know are influential. In the good old days of PR they were the journalists and industry analysts. You would brief them in advance so when the story broke, they were ready. In this case, it’s the bloggers and the influential tweeters.
Now, you want to go to the journalists and the bloggers and say “next Thursday the story is going to break and we want to give you a heads up,” classic PR. Then it’s a matter of monitoring when the story breaks, and making sure you send that direct message to all of those influentials to say “here is something we know you’ll be interested in, click here, read about it, and please re-tweet.”
Now you should measure how much attention it got. If you launch a new product in the forest and there is nobody there to see it, does it make noise? You want to know does it cause people to talk a lot.
Apple recently announced their iCloud, which caused a lot of media attention, blogger attention, and lots of tweets; however, the sentiment is kind of iffy. Everybody is asking is this a good idea or a bad idea? Will it help? Is there a security problem? How might they have done their announcement differently and gotten better results?
Three weeks from now we’ll determine whether or not it went well. I’ll measure the quantity of attention, the amount of positive or negative it received, did that result in social actions, and did those result in business actions?
Eric Enge: A major reason for doing that is you need to see how successful your marketing plan was. You probably contacted certain people in advance and offered the opportunity to get advance info under embargo, or even given them the exclusive on the announcement.
Jim Sterne: Or we’ll send you a free version of it, or a sample of it.
Eric Enge: Right. Also, we had certain follow-up tactics we used. You can measure the results of your strategy by tracking number of mentions, re-tweets, increases in followers or friends, blog comments, and all the things we’ve talked about.
Then, based on what happened and watching the curve of engagement activity, you change some things, do something a little bit differently next time, and see how that curve looks. Did we get a bigger spike in the beginning? Did it have a longer tail of activity?
The metrics used can get sophisticated pretty quickly.
Jim Sterne: Exactly. The next fun part is the experimentation side. Once the launch is kicked off, I can start my standard ongoing promotional campaign and optimize it, tweak it, test it and try out different types of promotions. The metrics used can get sophisticated quickly.
For each blogpost, Facebook comment, or tweet I can use an unique URL so when people click on it I know that a particular tweet at 10 o’clock on Tuesday morning got x amount of traffic (Eric: for example bit.ly Enterprise can be used to do this). When everybody re-tweets it and copies your URL, your unique identifier, you can add those together to say “gee, that tweet got fanned out across the blogosphere, or the tweetosphere, and caused many click throughs which translated to page views for publishers, or donations for nonprofits, or qualified leads for B2B, or actual purchases for B2C.” I can tell which of my efforts is working.
Eric Enge: Right. There are also things like time of day. Was 10 a.m. Eastern Time the best time to tweet this, or launch this, or would 1 p.m. have been better? With tweets you would schedule a few throughout the day and then compare how each does and even track regional response, for example.
These are good tips to do with your product launch scenario. Of course, there is the process of measuring with the general building of social audience. I am not talking about a specific product launch but, rather, an ongoing effort in a social media environment to measure what’s taking place with your efforts to build your audience.
For example, you are working with certain kinds of channels, or certain blogs, and trying to measure the response you get in terms of growing your audience by working through that set of blogs. You compare that to the kind of results and growing your audience you get with your Facebook ad campaign.
Nothing happens in a vacuum, you could launch something, and it is suddenly gigantic because it was a slow news day, or it was terrible because it happened to occur at the same time a giant story broke.
Jim Sterne: The tricky part is you want to compare and contrast all of the above. By that I mean nothing happens in a vacuum. If you run an ad in a magazine, there are also radio, television, and banner ads, as well as search traffic, and social media. This all happens at the same time. You need to make sure you monitor the interaction between these things.
Our problem is we like to think we ran this ad and it was great. But what else was going on? We did a social media campaign, and the result was zero. Was it a terrible campaign?, No, it just happened to occur at the same time a giant story broke, and everybody was watching the news and nobody was paying attention to your product announcement. On the other hand, you could launch something, and it is suddenly gigantic because it was a slow news day. Congratulations!
Eric Enge: Thanks Jim!
