There is much talk in marketing circles about how to measure the impact of social media. Some measurements are hard, such as actual campaign response and conversion rate measures. Others are a bit softer, such as measuring campaign reaction frequency and tone (e.g., positive, neutral, negative). While both are valid measures, I do think we have been missing a broader, yet critical component of our measures—the overall financial impact of word-of-mouth (WOM) spread.
It occurred to me today that I may already have a way to measure WOM influence and the impact of social media.
Years ago, I came up with a simple equation to measure what I called, the “Residual Value of a Customer.” In other words, this is a calculation to determine value of an average customers’ impact on your business relative to their individual influence on other customers. Keep in mind that this was before the internet and social media tools, so the sphere of influence of an individual customer was generally much less—maybe 7-10 people total. However, I think the logic still applies today.
The Residual Value of a Customer takes into account the annual sales to a customer, the expected tenure as a customer, and the estimated number of people influenced. For example, if “Customer A” spends $150 a year with a company and the average tenure is three years, then “Customer A’s” value to the organization is $450. However, if “Customer A” recommends the product/service to just one other customer who follows the same spending/tenure patterns (as the average), “Customer A” now has a residual value of $1,350.
Let’s take this thinking a step further. Recent research has suggested that the average Facebook user, for example, has 120 friends. The average user may interact meaningfully with between 10 and 20 Facebook Friends within a 30-day period. Using the calculations above, let’s say “Customer A” influences 20 friends within a 30-day period. “Customer A” now has a residual value of $27,000, as do each of those 20 friends who adhere to the average customer measures. In this first circle or ripple of influence, the residual value of these 21 customers is now more than one-half million dollars over the next three years, assuming the averages spending and purchasing life remains consistent.
These are significant numbers, and all brought about by one customer sharing experiences with a circle of friends.
I have used this model a number of times to demonstrate the power of WOM marketing programs to senior management. It is simple to understand, and proven using average customer sales and tenure numbers. In the majority of the cases, I’ve been successful in gaining support from senior management for at least testing WOM or now, social media, programs. I have also used the Residual Value of a Customer to demonstrate the opportunity cost for not engaging in WOM.
In the spirit of sharing, I’ve created an online version of the model for you to use here: Residual Value of a Customer Calculator. Feel free to use this model and share with others.
I’d appreciate your feedback.
2 thoughts on “How to measure social media”
Sorry Dave, but I have to disagree with you on several points:
1. The measurement of Social Media’s impact on an organization is not a question mark anymore. We know what to measure, where to measure it, and how to measure it. (And I’m not even talking about R.O.I. here, just impact.)
2. Averaging out the potential influence of an individual based on generalizations about usage statistics of specific channels = fuzzy math at best.
3. You can now pinpoint the actual influencer value of a customer during a certain timeframe. You don’t have to guess or estimate it.
4. Every customer’s influencer score or value is variable. It doesn’t stay constant over time. So in terms of measurement, it’s pretty time-consuming to focus on this when so many other (more stable) metrics could (and do) give you more valuable insight into your SM’s impact. (Unless you are bent on figuring out how WOM impacts your business.)
5. Since 90%+ of WOM happens offline, using Social Media to calculate WOM (or measuring WOM via Social Media channels) only gives you 10% of the data you are really looking for. That’s not super solid (yet).
That said, I think you’re on to something. We do need to refine WOM measurement and I think you’re on the right track: Methodologies and tools aren’t sophisticated yet in my book, and that’s a problem given the importance of WOM. Looking forward to reading more of your ideas. You aren’t there yet, but I respect the fact that you’re trying to do something with this. 😉
Olivier,
I’m glad you responded. It was your presentation at Social Fresh that caused me to dig this model out of the mothballs to stir the pot and start a conversation.
I will say I don’t disagree with too much of what you’ve said. I do think you are ahead of what I was attempting to articulate. I realize now that I probably should have more clearly established the intent of the post and model. So let me do this, now:
You know, early in my career I managed many large direct marketing programs. I lived to *prove* the impact and ROI a program, campaign or strategy with my measures and analysis. Yet, there were some things I just did not have the data, tools, or time to prove. However, I could note trends and continue to test my theories.
Later in my career, I learned the true value of approximation (your word: “fuzzy math”) in testing new ideas in marketing. I had a boss who was prone to say, “If it’s a good idea, it’s worth doing poorly.” He encouraged me, as I encourage folks today, to come up with a way to draw a line in the sand as a point of reference and then test to it. In the absence of reliable data, tools, or time, I think the Residual Value of a Customer model might just help draw that line. In fact, when I worked as a senior marketing executive, if someone gave me the results of that model and said, “We can’t quantify this yet, but here’s what we think could happen if we launched a WOM campaign,” I would have been likely discount it by 50%, then give the go-ahead to test if proper steps were taken to quantify the effort.
Many people reading this post may well have an understanding of what to measure, when and how with a WOM campaign. Others will not even know where to start. Aside from ROI, some may choose to measure things differently than what you or I would measure. Who is to say what is right or wrong? I think it depends on organizational goals and objectives.
I do agree that a customer’s influence score/value is a variable and does not stay constant over time. I also agree that with access to data, you can pinpoint actual value within a specific time frame. Yet, there are many, many, organizations that do not have, cannot access, or have the skills to analyze such data. If this happens to be*your* organization, where would you start? I think I would ballpark the numbers using something like the Residual Value of a Customer model, then test it out.
As for the 90% of WOM that happens offline, how would one approximate the impact of the work online? If the executive is worth his/her salt, the question will be asked. I would probably draw a line in the sand somewhere as my starting point, and then test it out.
Yes, I really should have qualified this post and model a little better. I *should* have said, “use this to draw a line in the sand,” or “use this when you don’t know where to start but your boss wants to know the potential impact before handing you the keys to the car.” I did not, and I deserve the lumps you gave me as a result of my lack of clarity.
For the record, I don’t think the methodologies and tools are *ever* going to be sophisticated enough deliver acceptable WOM measures—there are too many variables. I think we’re always going to have to do a little “fuzzy math.” 😉 I don’t think we can be afraid of approximation when trying “new things” in marketing. If we are afraid, we will never innovate. Social media/WOM is still a relatively “new thing” to measure for most organizations. Let’s not be afraid.
Now, I have started the conversation. Olivier has chimed in. How can we as a community define some acceptable baseline measurements? More heads are definitely advantageous in this conversation.