Social Media ROI Framework
Here is a simple, five-step framework for developing a social media ROI measurement program. Remember that not all social media initiatives will result in short-term ROI generation. It is also important to comprehend the results of programs that result in non-financial value or impact. (For a quick refresher on the difference between value and ROI read this). Holistic measurement programs should be designed to track and measure non-financial impact as well as ROI.
1. Set measurable objectives aligned with business outcomes
Failure to begin with measurable objectives is probably the most common impediment to proper social media measurement. A couple of award seasons ago I was a judge for a major PR campaign competition and was appalled by the low percentage of programs that actually contained measurable objectives – about 20% or so. Your objectives should be aligned with one or more desired business outcomes. Think through all the ways in which the social business effort will contribute toward driving the overall business forward. Make sure the alignment is obvious and understood by all involved in program approval.
2. Link to and understand the requisite business process
In order to demonstrate ROI in social media it is necessary to link the results seen in social media with the relevant business processes they are addressing. Social programs to date generally relate to one or more of the following business processes:
|
Business Process |
Description |
| CRM | Crowd-sourced help, info, recommendations, customer relationships |
| Research | Competitive intelligence, insights, voice-of-customer, trends, feedback, reputation assessment, influencers |
| Marketing & Sales | Ideation, product promotion, hyper-local marketing, lead generation and closure, fund raising, testing, brand attributes |
| Communication/PR/IR | Stakeholder communication (internal and external) |
| Innovation/Product Development | Crowd-sourced ideas, problem/opportunity identification, collaboration |
For example, in a B2B company, you might try to link social media efforts with the lead generation and closure process.
For a program aimed at employee engagement, you might link social media efforts to the employee recruitment and retention business process. For an e-Commerce company you might be able to directly link to the sales process through unique URLs or click-tracking technologies.
Understanding which business processes are impacted by social networks, and how, is fundamental to understanding ROI.
3. Select communications model, research approach and key metrics
In addition to understanding the business process impacted by social programs, it is important to have a communications model to assess non-financial impact. The accepted Outputs – Outtakes – Outcomes communication model is difficult and confusing for many to understand and apply. Here is an alternative communication model that is somewhat more intuitive and in tune with social media measurement.
- Exposure – To what degree have we created exposure to content and message?
- Engagement – Who is interacting/engaging with our content? How and where?
- Influence – The degree to which exposure and engagement have influenced perceptions and attitudes of the target audience.
- Action – As a result of the social media effort, what actions if any has the target taken?”
The metrics listed are a starting point. You will note these metrics come from all three “zones of measurement” – web analytics, digital content analysis and primary audience research.
Possible ROI research approaches include:
- Correlation modeling and econometrics
- Staff cost reduction and/or cost elimination tracking
- Direct linkage via unique URLs and click-tracking
- Exposed/Not-exposed primary audience research
- Integrated cross-platform research – web analytics, content analysis, click tracking, primary research.
4. Gather and analyze data
Gather longitudinal data for the requisite business process and communications model metrics. When attempting to show statistical correlations, the amount of data you include is important because it impacts the confidence level of the results. Also, what metrics you attempt to correlate and how is crucial. For example, we might try to correlate social media brand engagement and audience influence with metrics like likelihood to recommend to a friend, likelihood to seriously consider the product or likelihood to purchase the product in the next X months. Select the metrics that are most applicable to the business process you are attempting to drive.
5. Calculate ROI and report results
We started with measurable objectives, aligned them with requisite business processes, determined our research model and approach and have gathered the data. Now we can calculate ROI and describe the value of the social business initiative. Results for key program metrics should be captured on a dashboard that may be shared with all program stakeholders are a regular basis.
(Note: Part two of this post will cover specific research approaches to determining social media ROI in more detail)
metricsman.wordpress.com 
With all the tools and things we can measure it is a great time to find the great leaders of companies that are ready to engage and innovate their companies. Thanks for the information and the way you have formatted it.
Thanks for your comment, Steven. I’ll be on the look out for those great leaders!
As always, Don, excellent framework and post. The first step of setting measurable objectives seems to be a major issue in our profession. In fact, in my practice, that’s the starting place for strategic planning. You’d be shocked at how surprised some of my clients are by that.
There are many communicators who don’t want to attempt distilling the impact of their activities from the surrounding influences — this is particularly true when the sales cycle is lengthy. How do we know, for example, that our increased social media profile actually caused the sale? Southwest Airlines can tell because of the nature of their product, but will I buy a computer absent foreknowledge of the brand as a result of a Tweet? Unlikely, though maybe there are people who will.
That’s why considering the impact across the dimensions you suggest is so important. Can we definitively say that our activity caused the sale? No, but we can model out its relative impact through research.
We could use a few big research efforts that quantify the impact on purchase decisions of all of the various inputs — advertising, referrals from friends, personal experience, third-party objective endorsement, etc.
Good stuff!
Sean
@commammo
Hi Sean,
Thanks very much for your insightful comment. Measurable objectives always seem to be more of a challenge that they should be. In my undergrad class, I can assure you every student could write a measurable objective by the end of the class. They probably grew tired of my hammering the concept into their brains.
Understanding the impact and influences is a major challenge that may be addressed through modeling as you suggest. Last click attribution is a challenge – that is giving too much credit to the last click that resulted in a transaction and forgetting all the communications and marketing work that proceeded and perhaps enabled the ultimate sale. Modeling may be used in this context to perform value attribution of the sequential events leading to the last click.
-Don B
Great job of putting that framework into place. We’ve definitely written about similar topics i.e. the difference between impact and ROI. I don’t think it’s enough to set objectives aligned with business outcomes; you actually need to make sure you have the processes and infrastructure in place to help you actually calculate ROI. I’ve spoken with companies that say they want to increase sales, yet when I ask them questions such as “what’s the customer purchase cycle,” they have no idea. This means that they are setting objectives and aligning them with business outcomes but they then don’t have the starting data to begin with in order to see any effect.
Great job explaining this
Hi Jacob,
Thanks for stopping by and for your comment. Fully agree with your point – you have to understand the business process you’re impacting as well as the desired business outcomes. Understanding just business outcomes is necessary but not sufficient. Thanks again, Don B