page contents Automotive Thought Leadership: Ralph Paglia: The Truth About Social Media ROI Measurement My title page contents

Tuesday, August 10, 2010

The Truth About Social Media ROI Measurement

Altitude Branding - 13 Truths about Social Media Measurement13 Truths About Social Media Measurement

M1. If you aren't measuring anything else, social media measurement isn't the problem.

Measurement is a discipline, and it needs to be business-wide. If you're going to ask about the ROI, value, or impact of social media and how to measure it, I'm going to ask how you're going about determining those things for other areas of your business, and ask you to translate or adapt some of those practices over to social initiatives.

If you're not measuring anything else, you'll have a learning curve. A steep one. It'll come complete with needing the right tools and platforms to collect data, the right people to analyze it, the buy in from management to spend the time doing all of this, and the commitment to use the measurement as a means to underscore your strategy. The social media data is available for the taking, so that's not the problem. The *real* issue is connecting the dots. See #4.

2. Measurement is not the goal.

The goal is to derive insights that teach you something of value, and then act on them. Measurement is a waystation, a path, but is not the goal in itself. You don't get a cookie for measuring.

You probably need to spend three times as much time and effort evaluating and acting on your data than you do collecting and formatting it. Why? Because the analysis is what yields direction, plans, action steps, you name it. You START with the data. You need to end up with a course of action, or the act of measuring (and all the time you spent doing it) is wasted.

3. Measuring activity isn't as important as measuring results.

Gathering fans on Facebook is an activity. How those fans chose to respond to your offer, sign up for your newsletter, or buy your product (or not) is a result.

Number of forum posts is an activity. How many of those forum posts converted new downloads of your latest ebook is a result. (Even better if you can take it a step further and show the ebook downloads that became leads).

Follow me? We're very caught up in trying to track all the stuff we're doing, and not spending enough time connecting dots between those actions and how they drive progress toward the goals we've set. Speaking of which…

4. Metrics are determined by goals.

Learn how to create measurable objectives and the metrics practically jump out at you. If you know where you're headed and have a clear definition of what it means to reach your goal, it becomes pretty apparent which signals (metrics) will tell you whether you're close, far, or right on target.

And you don't need 40 different metrics to underscore a hypothesis or progress toward a goal. Typically you just need a few. If your goal is to raise awareness for a cause, you can look at reach of mentions and messages, increased donations, or a surge in volunteer signups. Those go up, you can be reasonably certain that what you're doing is contributing to those things, and likely justify staying the course. Which leads me to a biggie…

5. Cause and Correlation are different.

Cause means that something you did drove someone to act. Directly, and usually singly. There's a clear line between initiative and result. (We could argue that nearly every causal relationship has external influences, but that discussion for a headier day).

Correlation is fuzzier, and where most folks get hung up with measurement. It's about a relationship between two things, usually an action and a result, but that relationship isn't exclusive of other factors.

We struggle with these two, because we're often trying to prove cause, when correlation can be just as valuable in terms of justifying our efforts. Think of correlation as "contributing to" or "influencing". So if you do an outreach campaign in social media and lead numbers through those channels increase, you can say that those two events are likely strongly correlated. (By contrast, if you do a campaign in social media and your offline event attendance increases, they might still be related, but likely more loosely).

Remember that today, we have any number of points where prospects and customers can be impacted by what we do. Proving cause can be tricky, because you can't trace every interaction someone has with your company.

But we strive for cause why? Because we want CREDIT. We want to be able to say that OUR effort is what moved the needle so we can justify time, budget, headcount. But the only way to truly prove cause to a major degree is to adjust ONLY one thing while leaving all other factors the same. We rarely if ever do that in business, because we're not conducting science experiments. We're simply trying to understand what helps and what hinders. Get comfortable with this phrase: reasonable degree of confidence.

6. Analysis is the hard part, not measurement.

The human brain factor is the complicated bit. Data is easy to collect, easy to smash together, easy to do math around. The REAL question is: what does all this MEAN to me and why? What does this tell me about the effect and impact of my actions?

That's the hard part because no tool in the world can do that for you. No case study will show you precisely the map you need to follow for YOUR business (though it might spark some ideas). No one person can hand you a turn-key set of metrics that will suddenly give you a lightbulb moment and show you the path ahead of you. Put the effort into goal setting on the front end and analysis on the back end, and let measurement be a process in between.

7. Standardization has limitations.

You might have some *types* of metrics that can be bucketed together – such as engagement or awareness metrics – but the unique ones that matter to your business aren't likely to be standardized anytime soon. That's a departure from the way we've always done it, but then again, some of our "standard" metrics haven't really gotten us very far (like ad equivalency) and others are standard in name, but not in how they're calculated (like customer satisfaction).

Instead of striving for metrics that are universally applicable, focus instead on the ones that consistently deliver valuable intelligence for your business. It doesn't matter what the guy down the street is measuring unless you're just looking for a little inspiration.

8. Reporting is not an outcome.

Related to #2, delivering the graph isn't the end of the road. It's what you outline as the next steps to either a) keep doing what you're doing or b) adjust something in order to try and change the results.

The report, in fact, is often the starting line. And reports full of data alone aren't very useful. The art in reporting isn't just packaging the information, but its in interpreting and translating that. When you give your boss the monthly report of PR impressions or lead volume, do they ever ask you what you attribute those numbers to, and what recommendations you would make based on that information? Have YOU ever thought about that? Why?

9. Measurement doesn't have to be complex to be effective.

You don't necessarily need convoluted indices to get you where you're headed, especially when you're starting. Sometimes, just a simple correlation between an awareness metric and a sales timeline can tell you whether there might be a positive relationship, and you can act on that. Think of it this way: pair one qualitative metric – like customer satisfaction – with one related quantitative one, like sales or call center costs or website hits. One metric alone rarely tells you anything valuable.

Are they both headed in the direction you want them to be? Over time, do you see them moving together, away from each other, or in unrelated ways? Do the strategies you have in place to move them both tie into one another?

CAN measurement be complex? Sure. Some really detailed measurement formulas can help you get super scientific and granular. But again, if you're getting mired in the process of measurement instead of the practice of deriving some intelligence from what you measure, you're doing it wrong. The average business simply needs a guide, not a dissertation.

10. Measurement is a constant evolution.

You set a goal. You back out a few metrics. Then you evaluate, and realize you haven't learned anything of value, or that you need more clarity, more specifics, a broader view, or whatever. That's okay. Look, business is an iterative process. It's part art, part science, and so is measurement. Who wrote the rule that said we had to have the perfect, bulletproof set of metrics before we start measuring?

If something doesn't get you the information you need, change what you're doing and try something else. If you're missing something, add it. Eventually, you'll settle into a few combinations of metrics that really illustrate to you those Almighty Actionable Insights.

We're way too caught up with being perfectionists about gathering and presenting information, and not nearly good enough with FRAMING the information in a way that gives us something to chew on.

11. Measurement is cultural as well as operational.

We're taught to fear failure, so if we track and measure failure, we don't want to share it. We manipulate numbers to show our work in its best light, instead of showing the hard truth in order to identify what we need to improve to be more effective. That's a *culture* problem, based in businesses where accountability is absolute, blame is personal, and failure is a dirty word. That's a conversation that can't be fixed with a PPT presentation.

12. Measurement is more than ROI.

Measuring ROI is something we can and should do. Track how much we spend (in time and capital), track how much we net in terms of return (usually $$). That's a smart move.

But we can't limit the discussion about measurement to ROI. We have to talk about qualitative metrics, like brand perception, customer satisfaction, advocacy. We have to talk about quantitative metrics that tie to things other than revenue, like reduced costs. We also have to understand the difference between justifying something from a "good use of time" perspective, and looking at a financial return as the way of determining success.

Ultimately, all roads lead to Rome. But so much of social media isn't about being the sales channel, but is about positively impacting the likelihood of sales through all other experiences.

13. "Social media isn't measurable" is an excuse.

Here's what people really mean:

  • I don't have the right tools in place to collect the data I need
  • When I have all the data, I don't know where to start
  • I don't know what data might relate to each other to analyze it well
  • I don't want to or am not empowered to spend time doing data collection and analysis as part of my job
  • I'm afraid of what measuring will actually tell me about the effectiveness of my work

The first one is a functional problem. The second and third ones are knowledge based, with no exact "right" answers, and require a bit of practice and applied effort, but they're solvable too. The last two are cultural, and are probably much more firmly rooted in the people rather than the process. That's a different discussion.

Above all, we have to stop blaming the medium for hindering the measurement process. It's not social media's fault at all. If anything, it's guilty of providing us too much information.

What we need to understand about our own measurement practices is whether we're equipped with the right tools and data, whether we're willing to spend the time evaluating that data and extracting the juicy bits, and whether we're functionally and culturally prepared for what it might show us, for better or for worse.

But make no mistake, folks, basic social media measurement isn't someone else's responsibility to sort out for us. And waiting for the manual is simply burning time and money.

Measurement is our job. It's our responsibility. And it's within our capabilities, without doubt. So let's get cracking.


image credit: BigTallGuy


 Comment by RickNo Gravatar Subscribed to comments via email
2010-05-27 07:39:25

Thanks – your posts typically have the most sense-making and work-meaningful points to make. Saving this one to work through later, along with your week long series a few back on measurement. That's my next big hurdle, to draw conclusions on cause and correlation as you and @thebrandbuilder/Olivier have been saying recently. No excuses for not getting there yet. :)

 Comment by Murdoch MarketingNo Gravatar
2010-05-27 08:09:19

What's the saying – a cobbler's son has no shoes? We're working on social media strategy in-house right now and have been struggling with how to effectively measure social media success. This post really helps put activity and results into perspective – thanks for your insight!

-Murdoch Marketing

 Comment by KristofNo Gravatar
2010-05-27 08:31:37

Hi Amber. Great summation. Thanks for putting this to rest – specially number 13.
Kristof´s last blog ..kristofcreative: B2B Marketers Gain Ground w/ Social – Perceived irrelevance still a barrier My ComLuv Profile

 Comment by bencurnettNo Gravatar
2010-05-27 08:39:43

I would encourage anyone struggling w making meaningful metrics to follow the first link in this post. It's a great overview of the measurement progression: "goals>objectives>strategies>tactics" for new readers of Amber's blog (that's me).

Thanks Amber
bencurnett´s last blog ..The Four Parts Of A Content Strategy For Small Business My ComLuv Profile

 Comment by JonathanNo Gravatar
2010-05-27 08:46:19

Excellent post with lots of good pointers (particularly Nos. 3 and 5). My only issue is with No. 6: People need to collect data with care because muddy data leads to useless (or, even worse, misleading) analysis. As a researcher, I've seen many sites use and abuse statistics (especially in the form of ill-advised online polls), and I find research design to be the most important step in the process.

 Comment by Beth RobesonNo Gravatar
2010-05-27 12:29:24

I agree with you Jonathan. A lot of companies drop the ball on this step and wind up blaming the end product instead of the process.

 Comment by Jeff MelloNo Gravatar
2010-05-27 09:04:45


GREAT post!

I do believe for many understanding what to measure when it comes to social media is still a very important discussion. A social media strategy is company specific but the following are the areas I believe will provide a good start when looking at building a social media ROI measurement system…

1) Sales figures from the last year
2) Website/Blog Traffic information
3) Customer acquisition models (Sales Funnel)
4) Conversion ratios
5) Monthly expense for each department affected by social media
6) Monetary value placed on a new customer
7) Monetary value placed on an existing customer
8) The Traditional Media metrics currently being used

To your point…how this data is measured and analyzed is where the aha moments exist. Thanks for providing great information and a platform for valuable discussions.
Jeff Mello´s last blog ..Press Release-Atlanta Spine Doctors My ComLuv Profile

 Comment by Keith BurtisNo Gravatar
2010-05-27 09:51:48

The common thread running through this is to make sure you have the talent necessary to analyze and create actions based on the data gleaned. In Avanash Kaushik's book Web Analytics 2.0 he calls it the 90/10 rule. He says that 10% of the total formula is in the collection while 90% is in the actual analysis and reporting on the data. In my experience this is 100% true.

Well written Amber. Keep it up.
Keith Burtis´s last blog ..Announcement ~ Joining Forces with New Marketing Labs! My ComLuv Profile

 Comment by Sean WilliamsNo Gravatar Subscribed to comments via email
2010-05-27 09:58:57

Amber – you've rocked it. These three sentences from #2 would have been plenty of great info for the uninitiated:

"The goal is to derive insights that teach you something of value, and then act on them. Measurement is a waystation, a path, but is not the goal in itself. You don't get a cookie for measuring."

The PR planning chestnut, ROPE (Research, Objective, Programming, Evaluation) still holds truth. We all too often skip the R and sometimes even the O. And what we have left isn't worth a bucket of…


 Comment by Don BartholomewNo Gravatar
2010-05-27 11:36:49

Really great post. One of those, "Damn, wish I had written that post", moments for me. So much good advice here. Couple of comments that build on your ideas:

- Too many people view measurement as a scorecard or report card and not a diagnostic tool to help us decide what's working, what's not working and what should we do differently going forward. This mentality leads to a lot of bad measurement thinking and decisions.

- With correlation analysis, the action-result relationship may in fact be exclusive of other factors, but we cannot prove that through correlation so we must assume extraneous factors exist and may have unknown influence.

- I go back and forth on the standardization issue. On one hand standards help with simplification and scale while on the other hand every company/brand has unique characteristics and circumstances that might be best served by unique metrics. Perhaps the middle ground will prove fertile – a longish list of metrics that are understood and have 'standard' definitions across the industry, and each company might use a unique combination of some of these metrics to best fit their unique objectives and contextual requirements.
- Regarding ROI, it is a financial metric and is always, not usually $$ – be it revenue, cost savings or cost avoidance. Fully agree that ROI should be a limited part of the social media measurement discussion. Most social business programs are attempting to create impact, not ROI in the short-term.

Thanks again for a very valuable post. -Don B @Donbart

 Comment by Amber KhanNo Gravatar
2010-05-27 11:45:23

I could not agree with this statement more, "But so much of social media isn't about being the sales channel, but is about positively impacting the likelihood of sales through all other experiences."

If companies enter into social media campaigns for the sole sake of generating sales, they will be very disappointed. I guarantee the very companies that have this primary goal will convey the wrong messaging to users on Facebook, Twitter, etc.

I feel the first goal of social media should be to gain trust with users and gradually get them to explore your services. You also want to interact with users and not just push your product in their faces.

Our content writers have had much success by taking this approach and have seen our clients social media efforts deliver results over time.

Great article!

 Comment by Bonnie HarrisNo Gravatar Subscribed to comments via email
2010-05-27 11:47:47

Measurements are important but first you have to determine clear goals so the point that measurable objectives lead clearly to metrics is so important. I find so many people jumping into the tactical social media activities,and they get discouraged because they can't measure progress on a real time basis. The old goal-strategy-tactic-measurement chain we learned in school still applies whether you're doing direct mail or social media.
Bonnie Harris´s last blog ..How to pitch a major social media influencer like Lee Odden My ComLuv Profile

 Comment by Beth RobesonNo Gravatar
2010-05-27 12:32:02


Great insights, thank you for sharing them. Haven't found anything as insightful on measurement since David Meerman Scott said, "What is the ROI of measuring ROI?
Beth Robeson´s last blog ..Opening Day Card: Unique and Powerful Way to Connect My ComLuv Profile

 Comment by Tim GreenhalghNo Gravatar Subscribed to comments via email
2010-05-27 12:47:21

Thanks Amber – first time in years that anyone has nailed what it means to measure. Too often, measurement is a tick in the box; comfort for time-pressed execs… evidence that *something* is working. Even with the profound changes that the maturing Semantic Web will bring (we hope :) , human analysis, inference and move to action remains at the centre of metrics. Goal and conversion-focused strategic thinking seems to me the only way to get a return, which forms a perfect circle, the endless cycle of measurement, analysis, inference, action… and back to measurement.

 Comment by Ryan McCormackNo Gravatar
2010-05-27 13:18:10


Another great post…Required reading for anyone thinking about social media in the context of their business.

For me, the strongest insight was your first. Many clients with whom I talk ask about the ROI of social media, but they're not measuring the ROI of their other communication tools which are taken as givens. Someone would look at you like you were crazy if you asked them, "What's the ROI of having a phone?" I look forward to the day when the equivalent question for social media is equally ridiculous.

This is not to say it's not important to think about ROI, but rather that the inability to define a *precise* ROI is not reason enough to not engage in some way, even if just to listen.

 Comment by John McTigueNo Gravatar
2010-05-27 13:39:33

Great post as usual Amber. One metric that is often more difficult to measure is the quality of an engagement or lead. The first problem is how to define "quality". Did a lead via your whole web site and convert on multiple forms, or did they just download your freebie and move on. I like the way HubSpot grades leads along these lines, because it gives you something more valuable than a statistic. Lead quality gives you priority and direction. The same can probably be said for other metrics. Too often we obsess on numbers. I would much rather have one high quality engagement a day that may lead to some business than a hundred mediocre ones that probably won't.

 Comment by olivier blanchardNo Gravatar
2010-05-27 14:54:15

olivier blanchard´s last blog ..Do you know who the mayor of your business is? My ComLuv Profile

 Comment by DaveMurrNo Gravatar
2010-05-27 14:55:43

Very helpful advice. Thank you for providing these insights. Curious, why do you think so many see ROI/metrics as the life raft of the social web? It seems like numbers are used as a security blanket and as an excuse for not taking calculated risks.
DaveMurr´s last blog ..Reflecting on the #DETChevySXSW experience My ComLuv Profile

 Comment by MikeNo Gravatar
2010-05-27 17:16:18

Great advice for the new business owner trying to improve their social web.

 Comment by Tyrone TellisNo Gravatar
2010-05-28 00:30:41

Hi Amber,

Great piece and thanks to Angela Giles for linking to it.

My only point of disagreement is no 12.

At the end of the day some say ROI and sales is the key figure.

When activities have an quantitative cost and a tangible cost in terms of money, to use intangible metrics like brand perception, advocacy is sure to give the finance people or even the CEO sleepless nights.

Intangibles must lead eventually to tangibles like footfalls, leads, sales etc to be impactful. Right?

 Comment by peoplestringNo Gravatar
2010-05-28 01:50:32

You have great insight into the intricacies of social media. Increasing web presence is every business goal. We should also remember that trust is a major element of success

 Comment by Matt HixsoNo Gravatar
2010-05-28 02:03:57

Boy that is a lot to say about metrics. I find them to be very helpful in gaining buy in from people when you are getting social media off the ground. This tends to be low on the priority list for most but they love the idea. If you can show they are not close to meeting the goals they all seem to love it can be a very motivating thing to have real data to show results. I agree that should only spur further action not be the final result.

 Comment by Jason BreedNo Gravatar
2010-05-28 08:31:11

Great job Amber as always!

While measuring social is so important, it's only the 1st step in the journey to measuring overall marketing effectiveness. For businesses, they need to tie together online, offline, mobility in order to make truly informed decisions otherwise it's simply 3 data points. Same goes for Service, Innovation, Research and everywhere else that social can be applied.

I spend many of my days in this world and that is the reason I will keep this post close at hand!

Rock on!

 Comment by Rob ClarkNo Gravatar Subscribed to comments via email
2010-05-28 10:02:46

Hi Amber,
While I agree with the overall thrust of measuring results not activity, I would hope that people still pay attention to the activity.

Observing the results in conjunction with the activity can give you excellent insight as to what drove those results. It may reveal opportunities for building upon or improving in future.

- Rob Clark

 Comment by Paola Finocchiaro @paolabrusselsNo Gravatar Subscribed to comments via email
2010-05-28 13:12:14

"Reporting is not an outcome."

I am going to quote you on that one !

 Comment by Leanne Hoagland-SmithNo Gravatar Subscribed to comments via email
2010-05-28 13:13:46

Your list is awesome and extends to all efforts and not just social media. How many times do we hear:

Training cannot be measured?
Customer loyalty cannot be measured?
Employee loyalty cannot be measured?
Productivity cannot be measured?
Customer acquisition cost cannot be measured?
Sales leads costs cannot be measured?
Marketing actions cannot be measured?

Thanks for a great list – will be referring to it in my blog tomorrow.

 Comment by LyndiNo Gravatar
2010-05-28 13:22:06

Great post! Look at all the conversations it started. Measurement needs to look at the whole picture. Analytics are not going to have all the data, Salesforce isn't going to have all the data, on and on and on. Objectives just like you said need to be clear and measurable. Everyone needs to understand from the beginning what those objectives are and agree to the how they are going to be measured. I love "Above all, we have to stop blaming the medium for hindering the measurement process. It's not social media's fault at all. If anything, it's guilty of providing us too much information."
Lyndi´s last blog ..The Evolving Marketing Culture My ComLuv Profile

 Comment by Martin SeibertNo Gravatar Subscribed to comments via email
2010-05-29 04:33:42

We are creating a social media metrics application at that will centralize all your web stats in one place. We only have twitter, facebook and so far. But I would love to get feedback from you guys about it, as soon as it gets live in the next weeks.

I fully agree with your statements and the service we create will not really tackle the problems you describe. A professor in our university once said: "A fool with a tool remains still a fool." And you are right, if you say, that there is no successful measurement without hypothesis and goals.

But we already implemented scientifically valid analysis to make sure to highlight statistically relevant elements. Thats more than nothing, is it? :-)

 Comment by Global PatriotNo Gravatar
2010-05-29 14:13:52

While some aspects of business are face-to-face, you most often don't have the opportunity to see your customers smile. To your point about "brand perception & customer satisfaction", sometimes ROI resides in having happy customers. They're the ones who tell their friends and bring you business that's a step or two away from your direct efforts.
Global Patriot´s last blog ..BP Oil Spill Quandary My ComLuv Profile

 Comment by Leanne Hoagland-SmithNo Gravatar Subscribed to comments via email
2010-05-30 08:31:34

If a business has a process in place to track those customers who come because of another person's smiling face, then ROI is measurable. Sometimes all you have to do is ask this question: How did you hear of us?

 Comment by Mic AdamNo Gravatar
2010-05-30 11:38:29

Great post. I just posted something along the same lines…


 Comment by bcandersonNo Gravatar
2010-06-01 07:59:18

This is a appropriate topic for a marketing medium in it's infancy. I especially appreciate numbers 1 & 12! Nice work.

 Comment by Kingsley TagboNo Gravatar
2010-06-01 11:42:51

Social media is certainly measurable, just like everything else on Earth is measurable. Measurable data gives provokes very clear questions and more complex questions as well. If someone can't see the value of measuring that data, then they will have to keep waiting until social media is so big that every Internet user is taking part in it or the next decade of Internet trending is upon us. The most popular websites around become that popular because they are valuable to millions of people in very similar ways and also in completely different ways for each of them.

 Comment by Adam GrayNo Gravatar
2010-06-02 13:18:55

Amber, as always a fantastic post.

The biggest problem that I find is that clients either expect social media to be like some magic wand to cure all of their mistakes over the last couple of years…and therefore they don't need to measure the success. Or they want the full Radian6 deep dive into their industry, and detailed analysis of who/what/where/when but assume that because it's social media this will all be free (or nearly free).

I wish that this kind of post was more broadcast (I'm doing my bit) so that clients accepted that these was the facts, rather than having unrealistic expectations.

Thanks for all your posts, I learn something from every one!


 Comment by Jacki SemerauNo Gravatar
2010-06-04 17:33:09

"Put the effort into goal setting on the front end and analysis on the back end, and let measurement be a process in between." Best point I've ever read on the subject!
Jacki Semerau´s last blog

[Sent from Ralph Paglia's iPhone]

Ralph Paglia
Director - Digital Marketing
ADP Dealer Services
cell: 505-301-6369 

No comments:

Post a Comment

Enter your comments or questions here...