How I Address the Question of Enterprise 2.0 ROI
March 9, 2009 25 Comments
Establishing a solid ROI for enterprise social software is an ongoing discussion for the sector. It is generally a requirement for most technology decisions made by companies. At a high level, there are two sides to this argument:
- Measuring Enterprise 2.0 ROI is like trying to measure the ROI of email, it can’t be done
- Inability to measure ROI is a cop-out or evidence of the lack of value for social software
Martin Koser has a really thoughtful piece along these lines. It is something that I’ve met head-on in my work at Connectbeam. We have several large, blue chip enterprises with whom we’re engaging. And the need for some sort of ROI justification is a recurring request.
I want to share how I’m approaching this request. Before that, I want to describe my experience in providing ROI for enterprise software, of the non-social kind. It’s useful as a point of comparison.
The ROI of Re-engineering the Credit Approval Process
I worked for a company named eFinance from 2000 – 2005. eFinance provided a hosted application that let enterprises run automatic credit evaluations on their commercial customers. This is an activity in most large corporations badly in need of improvement. I had the privilege of designing the scorecards for Hertz Equipment Rental Corp., using basic statistical analysis of actual customer payments and defaults.
For purposes of understanding what’s important in the world of back-office credit, here’s all you need to know:
- Businesses have credit records with D&B and Experian, which enterprises access for credit applicants
- New credit applications are manually reviewed by a remote credit office
- Initial credit decisions = Approved, Further Review, Declined
- Further Reviews can be either Approved or Declined
Turns out, there are some inefficiencies in the process. Inefficiencies which enterprise credit software can solve. The table below shows them.
For the “R” part of ROI, the benefits are clear. The data costs were reduced with our credit system. Easy to apply the cost differential against the number of credit reports to arrive at a dollar savings. The credit reviews were another easy area for which to calculate the benefits. Each Further Review takes an average amount of time, which for a given volume means you needed N people. Reduce the percentage of Further Reviews, and fewer people are needed for a given volume of credit applications. Meaning headcount reductions.
The third benefit was faster response time to contractors seeking to rent equipment. While they didn’t have stats for lost business due to delays in responding to customers, feedback from the field was that this was an issue.Indeed, this third benefit was considered the most valuable.
From the eFinance work with Hertz, what characteristics are of relevance in considering Enterprise 2.0 ROI?
- Ability to measure ROI was directly related to the ability to measure the underlying activity
- The tangible dollar savings justified the project costs, while the intangible benefit of customer response time was the most exciting
- The software was applied to a very specific activity
With that in mind, let’s turn to the question of ROI for social software.
The ROI of Enterprise 2.0
The challenge with social software is that it addresses unpredictable, unmeasurable activities. And Enterprise 2.0 addresses a range of activities, not just a single process inside companies.
The graphic below is part of my ROI presentation for Connectbeam:
My X axis measures the predictability of the benefit. “Predictability” in this instance referring to the ability to know ahead of time how the benefits will manifest themselves. Reflect on this measure for the eFinance work for Hertz. Predictability was high for:
- Usage of cheaper D&B data reports with less data
- Reduction in FTE hours for processing Further Review applications
My Y axis measures the amount of value for the different benefits. “Value” defined in terms of revenue impact and dollar savings. In the eFinance example, the benefit of faster response time to customers, while not readily calculable based on existing data, was perceived to be a strong value proposition.
For Connectbeam, I put the benefits into three buckets:
- Time savings (see this IBM article for how much)
- Increased connections between need & knowledge (see this Connectbeam blog post for an example)
- Stronger more diverse employee social graphs (see my earlier post for an empirical study of this)
I plotted them as having increasing value, but decreasing predictability. I won’t go into detail on how I describe these buckets, but the links above touch on it.
Essentially, the time savings are real, but are the lowest return to the enterprise. I look at those as the easiest to predict, with defined dollar benefits. In the ROI presentation, I can show how these alone offer payback on Connectbeam.
It’s the higher-value benefits where the ROI story is harder to present. After considering my previous success in identifying and articulating an ROI story for non-social enterprise software,
The ability to have predictable ROI for software is directly correlated to the predictability of the underlying activity that uses the software
Think about that. With the credit software, there was a standard process with known unit volumes. Each step in the process could be measured in time. Frederick Winslow Taylor would have loved it for its predictability, standardization and amenity to quantification.
What underlying activities does social software address?
- Better decisions through improved access to relevant knowledge and content
- Tapping the little bits of knowledge employees have en masse to provide better direction
- More agile enterprise through improved connections and ambient awareness
All of those activities include elements of being unplanned, ad hoc, and creative. In short, they’re unpredictable and unmeasurable. The benefits also apply across a wide range of activities within the organization. Maybe Finance as a better handle on a new accounting issue that’s cropping up. Sales is up-to-speed on a customer’s hot buttons faster. R&D connects with the right field person to talk through a new innovation.
Like the feeling that faster response time to customers will result in higher sales and more satisfied customers in the eFinance example, the activities and problems addressed by Enterprise 2.0 are known to anyone inside a company. But no existing measures for the problems associated to these activities exist. Nor is there a good way to systematically measure their improvement.
That’s why anecdotes from the front line are so important. They show that improvements are happening with social software, even though you couldn’t pinpoint where at the start of the project.
Dennis Howlett is one my personal favorites out there in the world of enterprise software. He has an accounting background, so he’s pretty hard on soft, fuzzy feelings about the value of Enterprise 2.0. I did find it interesting that this was his perspective on where ROI comes from:
In my argument, breakthrough ROI comes from seeing these technology through the lens of collaboration, which in turn implies process and context. I am mindful that huge amounts of value continue to be locked up in supply chains. AMR quoted a number of $3 trillion in 2005. Has that materially changed? Simply being able to communicate across supply chains in a meaningful manner could do wonders to lubricate those rusty wheels.
Note what he’s saying there:
- Apply social software to a specific area (supply chain management)
- Lack of communication among various parties is causing enterprises to tie up too much cost, capital in the supply chain
- Even here, the benefit is one step removed from a hard, tangible ROI. Improved communication begets the benfit, although how it does so is on the intangible side of things.
To wrap it up, my approach is to push forward with the ways in which Enterprise 2.0 delivers ROI. We cannot escape this duty in the industry. But I also am working to set expectations for how predictable this ROI will be going in to a project. After all…
Software ROI is only as predictable as the activity for which it is used
See this post on FriendFeed: http://friendfeed.com/search?required=q&q=%22How+I+Address+the+Question+of+Enterprise+2.0+ROI%22