Yesterday, we went over the need to create an activities timeline that basically plots every relevant action your company takes across all media. Press releases, product launches, blog posts, white papers, ad campaigns, Twitter engagement, etc. Today, we are going to look at creating outcome timelines. Same basic process, but actually easier based on what you choose to measure.
Here. Before I go any further, check out this video. It will hopefully be a good intro into some of the content of this post:
You can kind of see where this is going right? Once we have the activities and the outcomes timelines mapped out, we are going to overlay them. (That’s right: stack them on top of each other and look for commonalities.) That’s how the R.O.I. analysis will begin: By looking for obvious correlations between activities and outcomes, and then either proving or disproving causality.
So what types of outcomes should we map out along timelines? Well, that’s really up to you. Depending on your budget, capabilities and resources, you can create as many as you want. There is probably no limit to what you can measure and analyze. Every company is different. That said, here are two things to keep in mind:
1. Focus on measuring what matters. Ever heard the term “paralysis by analysis”? It starts when companies start getting bogged down with nonsensical measurement. When you measure everything, you end up measuring nothing. So my advice here is figure out what metrics actually matter to your business and start there. (A piece of advice shared by Avinash Kaushik in this post – Go straight to #5.)
Speaking of Avinash, let’s borrow some of his advice when it comes to figuring out what matters (what you should measure first):
“If you can’t get your management to identify goals for you, update your resume and apply for other jobs.
While you are waiting, focus really hard on only reporting metrics that will help 1. increase revenue 2. reduce cost and 3. increase satisfaction. Can’t go wrong with those three.”
Incidentally, although Avinash’s post on measurement focuses mostly on web analytics, it’s a great read and most of his lessons are adaptable to other types of metrics as well. Read the post, bookmark it, and revisit it often. (Hat tip to Christopher Penn for that link, by the way.)
So the lesson here is to focus first on metrics that truly matter, and then (assuming you want to assign additional resources to the measurement of secondary metrics) expanding your measurement repertoire.
2. Focus on what you can actually measure. Unless you have a quasi-unlimited budget, enormous resources and extremely sophisticated data management capabilities, you aren’t going to be able to capture and measure every bit of data that comes your way. Every company’s capabilities, expertise and toolkit is different. While some of you may be able to implement fairly impressive measurement practices over the next few weeks and months, most will find that much of the data you wish you could capture, plot and analyze will be beyond your reach. Don’t panic. Heck, don’t even sweat it. Measure what you can, make the most of it, and don’t worry too much about what you aren’t able to measure yet. Your ROI analysis will require a little more art than science, but that’s okay. As you will soon see, certain key metrics should give you a pretty good idea of the effectiveness of what you are doing.
Okay. The first step is to identify key metrics for your outcome timelines, right? For the sake of clarity, let’s divide them into two groups: Financial outcomes and Non-financial outcomes.
A. Financial Outcomes
Financial outcomes are among the most basic business metrics. Raw performance measurement. No interpretation or estimation required. The most obvious is of course sales performance. So go to your VP sales and get yourself some sales data. (Chances are that your company’s sales data is already plotted over time, so your sales $ timeline may only be an email or database query away.)
Along a timeline, your sales data will hopefully look a little something like this:
Simple enough, right? (Again, most of this data can be mapped out in Microsoft Excel, which ought to work well for most companies, but the platforms you use are entirely up to you.)
Another way to look at financial outcomes might be to look at the number of transactions along a timeline rather than just looking at sales dollars. (Remember the F.R.Y. methodology that started this whole series a few weeks ago? Frequency, Reach and Yield? Looking at deltas in the number of transactions might help you figure out where some of the changes in sales revenue are coming from. (Hint: If the number of transactions stays flat but your sales revenue increases, you know that you have increased your Yield – the “Y” in F.R.Y.) When it comes to analyzing the impact that your activities is having on your sales (the return or “R” in the R.O.I. equation), knowing whether revenue deltas are coming from Yield, Frequency or Reach is pretty important.
All this to say that tracking transaction numbers in addition to sales $ is a pretty good idea.
Also in keeping with the F.R.Y.methodology, consider tracking active (transacting) customer counts so you can see if your increase in sales is coming from existing customers or net new customers. (It would be nice if you could see how many new customers you have gained over the last six months, wouldn’t it?)
If you can, also try to track the frequency with which transacting customers do business with you. (Membership/Discount cards are pretty helpful with that last one.)
By the way, if your well paid measurement “expert” doesn’t understand this stuff, perhaps it’s time for a new expert. It’s okay if you don’t quite grasp this stuff, but if your expert is still stuck on Google Analytics and web conversions only, or “impressions” as ways to “estimate” the value of media and potential returns, chances are that you aren’t any closer to calculating the R.O.I. of any of your activities than before you cut them a check. Remember: Measure REAL performance. Not eyeballs, not media, not what may happen in some magical marketing-friendly future. Measure what’s real. Measure what you can prove.
B. Non-Financial Outcomes
Now that we’ve covered financial outcomes, let’s get into the realm of broader metrics. (This is more fun but requires a lot more work.) Web measurement people will probably feel more at home in this section, as will Social Media measurement purists. As discussed in parts 6 and 7 of this series, non-financial outcomes (or forms of non-financial impact) typically influence financial outcomes. (At least that’s the goal.) They can be any number of things, from visitors to your website and followers on Twitter to (social) mentions of your companyand foot traffic to your stores. The best way to describe this category is this: Non-financial outcomes are all of the things your customers do as a reaction to your activities except transact (buy something). Any activity that is not a transaction falls into this category.
Warning: There are tons of things you can measure in this category, so remember the two precepts we covered earlier: Measure what matters, and measure what you can. (Don’t worry about measuring everything. Give it some thought, test things out, experiment, and see what works for you.) That said, a few key metrics should immediately come to mind: Web traffic, clickthroughs, non-financial conversions, (social) mentions, multi-channel engagement, foot traffic (yes, in the real world), postive WOM/recommendations, negative WOM, blog activity, retweets, linkbacks, etc. Some are web-based while others are not. Don’t get bogged down with media categories: Get out of the traditional media measurement mentality and instead, learn to measure customer behavior. That’s where real insights and valuable metrics are. What people do is usually the best indication of how they feel about your brand/company/product. So learn to observe and measure behavior.
Again, whatever you choose to measure has to be be plotted along a timeline. We will be overlaying outcomes and activities next, so this is important. (We’re only skimming the surface at this point, but you should be able to see how this is starting to come together.)
Some of your non-financial outcome timelines might look a little something like this:
Note: Yes, I could be showing you screenshots of tools like Google Analytics, PeopleBrowsr, Techrigy and Radian6 reports, but I am trying to keep these posts as tool-neutral as I can. My own biases for certain tools shouldn’t get in the way of the principles conveyed in this post. (Just know that the number of tools out there is growing, so feel free to test them all out and see what works best for your needs, budgets and business culture.)
Okay. I could go on, but I’d better stop this post before your brains explode. Let this stuff sink in, digest it a little bit, and we’ll be back with more later: In Part 10 of this series, we will talk about the importance of deltas (changes) in patterns, as well as causality and correlation. (We’ll start putting all of these puzzle pieces together.) Think of it as detective work for measurement geeks. (Oh yeah. Exciting stuff to be sure.) Any of you starting to wish you hadn’t asked me to show you how to do this yet?
Let me know if this is making any sense yet. It doesn’t need to be 100% clear at this point, but as long as the fog is starting to lift a little, we’re all on the right track.
And remember: As I mentioned in the video, this series isn’t intended to be a detailed A-Z “how to” for every company. If I tried to be THAT specific, each post would be 30 pages long, and even I won’t put you through that. Think of this series as a methodology “how to”: Principles that you can then adapt to your own business measurement practice. These are guidelines. Bend them as much as you want to make them fit your needs. Once you understand these principles and feel comfortable with them, it’s up to you to either adapt them to your specific needs (based on your capabilities), or recruit the help of someone like me to help you build a measurement program that will actually work. Doing this right is pretty serious business, so feel free to experiment but also understand that the success of your company depends partly on the quality of the analysis that you extract from your measurement program. (Knowing what works and what doesn’t work – and why – is pretty vital stuff.)
One last thing: Feel free to share with me what non-financial metrics you think would be crucial to measure (whether you measure them already or not). I have listed a few here, but I would love to hear what you feel should be added to the A-list.
Thanks, everybody. 🙂
“Ever heard the term “paralysis by analysis”? It starts when companies start getting bogged down with nonsensical measurement. When you measure everything, you end up measuring nothing.”
I thought of this when I recently saw a post by a social media firm that detailed the 50(!) or so metrics they claim to track and measure associated with EVERY blog post.
Classic case of more being less.
50 metrics per blog post? Wow. That’s probably taking things a little far. After sorting through those 50, you might end up only using a dozen (if that many).
A few non-financial metrics for a content site:
– Average time on site
– Content viewed per unique
– Visitor frequency
For an application type site:
– Interactions per user
– Depth of interaction (i.e. if there are several steps that a user can do, how deep into the process do they go)
– Feature utilization (i.e., number of features used)
Just off the top of my head, the types of things that I’d care about.
Good stuff, Rathan. Yes. Those would be typical of web analytics. I like your second list too. I wonder if depth of interaction deals more with the site’s UX/UI than with pulling the user to an actual transaction. In some cases, yes. In some cases, no. Depends on what the site is for. 🙂
Good stuff.
That’s a good point. Sometimes the UX/UI doesn’t make it easy to access functionality/transactability on a site. Although you can evaluate whether spending your investment on a social media campaign to try to highlight those features is enough to get over the hurdle versus doing a UI overhaul (or at minimum get insight on what part of the process needs the overhaul).
Plenty of meat on these here bones. I’ll need to chew awhile before I can pick.
A couple of observations. First is the bridge between financial and non-financial measures. It’s probably pretty obvious to those following your blog that this bridge is a myth, but for those of us who live in deeply siloed organizations it may be a revelation. We commit serious resources to executing PR strategies and related outcomes like positive mentions and share of voice. Yet there is often no real effort made to track those inputs and outcomes against sales activity — deltas in revenue, transactions, market share. And at the end of the day, those are the only measures that count. Your timeline discussion really drove that point home for me.
The second observation is a question. I have no trouble digesting this analysis from the standpoint of measuring ongoing SM activities, but what if you’re building a case for adding a social media component to your existing strategy? How do you go about demonstrating ROI on a prospective basis when there’s relatively little historical data on the space?
Your prospective question is a very good one. I don’t like to bullshit clients, so my preference is to not estimate or assume anything ahead of time. I don’t even like to use existing “accepted” media estimations of value (eyeballs, impressions, etc.) I don’t discourage clients from using these types of predictive metrics, but I don’t encourage them either. I like to measure actualized impact, not divine potential impact based on media-buy scenarios.
I think it’s important for companies to a) not get sucked into fuzzy math from the start, and b) not base their hopes and expectations on numbers divined out of thin air. I much prefer to start with a clean slate with them and review their activities so far: 1. From their perspective what is working and what isn’t working? Where are they getting the biggest bang for their buck? 2. How do they know? (How are they measuring this? Is their method good? Is it bad? Is it incomplete? Etc.)
The next step is to ask them what they want to accomplish. Establish specific goals with them. “Sell more stuff” is not a goal, by the way. They have to be more specific. “Sell more of X” is better. “Sell 2,000 more of X this quarter” is getting close. So, establish goals with them. Not a number out of thin air, but goals that are specific and achievable.
What comes after that is essentially a proof of concept discussion. I am not going to try and predict the rate of their success. I can give them anecdotal evidence from companies like DELL (that generated $3M in new revenue from Twitter alone), but that doesn’t speak to their business. That is just an illustration of possibilities and a proof of concept by proxy. I always want to be very careful when discussing expectations because every company is completely different. What worked for DELL may not work for HP or Nike or Starbucks. So anyway: Proof of concept. That’s what we’re doing. The first 6-12 months are a pilot. A test. And one we expect to either kick ass, or I’m going home.
The idea is to get them to clearly understand what their goals and expectations are, and work backwards from each to create an action plan. Not just a strategy but an action plan. That’s the first half of the equation. What we’re going to do, how we’re going to do it, and who’s going to execute. The second half of that is to have a frank discussion about monitoring and measurement. These two may sound like one program, but they are in fact two very distinctive programs: Action and Measurement, for most companies, aren’t going to live in the same headcount. You need different people doing different things if you want to do this well.
This is starting to to turn into a bonafide blog post, so I’ll stop here and wrap it up with this: It’s a gamble, Bill. Investing in a social media program is as much of a gamble as investing in an advertising campaign, a microsite, a trade show, a sponsorship or a promotion. They’re all gambles. You don’t really know for a fact what’s going to work or to what extent. This is too. Promising a client something for the sake of telling them what they want to hear and getting the contract is not professional, in my opinion. I would much rather be honest with them and admit to them that, if done right, this will work, but I can’t tell you to what extent yet or how fast. That’s why we will be monitoring success and making adjustments constantly.
Being able to demonstrate that you are interested in results (not just talk), that you know what you’re talking about and that for perhaps the first time in their company’s history, they will see real performance feedback and ROI metrics should do the job. But asking a company to invest resources in any program is always a leap of faith. If anything though, what they will get from the measurement part of the equation (assuming you can do this) is a pretty clear understanding of where their marketing and bizdev dollars work and don’t work. Even if SM is a mild success (let’s say it pays for itself but doesn’t significantly increase revenue – hey, it could happen), they will also be able to see the real value of their PR and other marketing activities through their new measurement practice. That’s a nice bonus most companies only yet dream about.
Does that answer the question at all? (I know I just rambled a lot.)
“Investing in a social media program is as much of a gamble as investing in an advertising campaign, a microsite, a trade show, a sponsorship or a promotion.”
I think you’ve captured what is bugging me. Yes, investing in SM is a gamble. Part of the trick is convincing decisionmakers steeped in traditional media that SM is a gamble on par wth the others.
By the way, the more I get into this, the more I’m convinced that if I were making a play for a significant SM investment I’d get expert help. And this is coming from someone extremely confident in his ability to mess up on his own.
Getting help in this area is pretty crucial, I think. Unfortunately, it’s hard to tell the hacks from the real brains… which is why we have blogs. 😉
You’re pretty far along on the right track, Bill. 🙂
Truly excellent article, Olivier. Why don;t we both save time and just use each other’s articles? : ) You’re right, we do have a common perspective.
No kidding. 😀
Olivier, this is spectacular! You speak directly to the fundamental of business – to make money. What’s really interesting about all of this is that in some cases, Social Media doesn’t work. The ROI process determines that as well. It may also lead users to what will work.
The other point for people to keep in mind is that Social media is not necessarily all about selling stuff. It can be cost savings from a customer service perspective, competitive analytics to see how competitors are doing in the market verses how you are perceived, lead generation that provides sales opportunity and many others. Each has its own specific business process and ROI metric.
Thanks again for such terrific information, can’t wait to see more.
Yes. And you’re right: I haven’t talked much about cost savings in terms of ROI yet. 🙂
mind blowing comment oliver