Following a recent post I’d written on Agile, David Tebbutt (who’s on my blogroll and an occasional commenter) raised the following point:
One thing bothers me and that is that huge numbers of people, evangelists if you like, have enthusiasms for things like ’social computing.’ Their enthusiasm isn’t for particular products, more for improving the communication experience, but they have no clear destination in mind. Not one that makes any sense to the sponsors they’re seeking anyway (e.g. the board).
Around the same time, Dennis Howlett (like David, on my blogroll and an occasional commenter) had these things to say as a comment on Charlene Li’s blog:
You can’t realistically apply accounting type measures to blogs because they don’t fit. Instead you need a blend of measures that reflect the blog’s purpose so in other words you need to seek measures that cross departmental boundaries.
….
Early attempts at this are not going to be easy. How for instance do you define a measure for ‘idea generation?’ What are the value measures needed? Do we have the historical data that allows us to develop assumptions about value created from great ideas and even if we do, how well might that relate to today’s situation?
….
I absolutely don’t accept it’s not do-able. The problem as I see it is we have no historical framework of the kind we’d usually apply and HCM metrics are a relatively new and as yet not fully baked science. Which means we’re inevitably faced with a loss of precision. We can overcome that by applying degrees of confidence variables so we can say there is a range of possible ROIs.
Charlene’s written an interesting post on Calculating The ROI of Blogging, worth a read, and, if you can, do look at the comments as well.
I think the nature of the debate is indicative of the challenges we face when we seek to implement social software in enterprises. I have carefully kept away from the specific subject of the ROI of social software for some time now, choosing to concentrate instead on the organisation’s immune system responses. Now, with the happenstance of the comments and posts above, I feel it’s time to break my silence.
So here goes.
Nearly ten years ago, in an article headlined Strategy Under Uncertainty in Harvard Business Review, November/December 1997, the authors said a few things that greatly influenced how I think about ROI. Here’s the summary from archives.gov (my apologies, that’s the best I could find on the web, and my hardcopy is now packed away in storage):
It is important not to underestimate uncertainty when shaping strategy, according to the authors, who present a framework for determining the level of uncertainty surrounding strategic decisions and for tailoring strategy to that level of uncertainty. In practice, they have found that the uncertainty facing most strategic-decision makers falls into four broad levels: a clear-enough future, alternate futures, a range of futures, and true ambiguity wherein it is virtually impossible to predict. The authors introduce a basic vocabulary for talking about strategies: three strategic postures including shaping, adapting, or reserving the right to play; and three types of actions to implement strategy: big bets, options, and no-regrets moves. The choice of a strategic posture and accompanying actions can be highly dependent on the level of uncertainty facing the organization. This article offers scenarios illustrating the range of strategic challenges.
That’s the background. So here’s what I think:
1. ROI is not the only game in town, you have portfolio analysis and option pricing models as well
If we restrict ourselves to measuring investments by ROI alone, we run the risk of weakening our capacity to survive, much less thrive, in an age of strategic uncertainty. Big Bets are like early stage investments, you have to work out what percentage of your investment portfolio you want to expose to those risks and returns; that percentage could be zero, but you then give away the right to receive hockey-stick returns. Big Bets are measured like early-stage VC portfolios. Options are just options, the price you pay for a place at the table, and you decide which tables you need to sit at. Options need option pricing and suffer time decay. No-Regrets Moves, in contrast, are all about ROI. You have to do something, now all you’re working out is the best something. Build Or Buy. Which Build. Or Which Buy. And there’s probably no better tool than ROI to work this out.
2. You have to be able to value human capital before you can value social software. Social is about human.
We talk a lot about valuing human capital. I’m not planning to venture into the esoteric and nascent science of measuring it right now, this is neither the time nor the place for it. But there are some leading indicators that suggest we don’t value human capital. We send people on training courses, apparently paying to add to their innate talent, acquired institutional domain knowledge and relationships, and nurtured abilities . But we never seek to measure that value added, or try to “land” that value on a balance sheet. So, no surprise, when the going gets tough, the training budget bids goodbye. We “plan for attrition”. Wow. We work out replacement costs for assets but never really for humans, our most precious assets. Read Peter Drucker at random if you want to find out more. He knew more about management and knowledge than anyone else I have had the pleasure of reading.
3. The plumbing for social software is not expensive, that’s just a lie.
Even if I’m completely wrong about the first two points, this one stops the show. Somehow we are allowing ourselves to fight the social software fight on infrastructure grounds. Unit costs for implementing social software are trivial, almost negligible. It’s just a classic immune system response. Moore and Metcalfe and Gilder have seen to it by now, the cost of tin and wire is not what it used to be. And you have every flavour you want: FLOSS software, commercial adaptations of opensource, fully featured COTS applications. You can start small and cheap, and segue into large and cheap. It’s not about the plumbing. No point figuring out the ROI of angels dancing on the heads of pins.
4. Current accounting processes are not precise either, but at least they don’t lie. They just reflect one of many possible conventional representations of truths.
I have seen far too many sausage machines in my time. Cost centres till the cows come home. Allocation cycles built in Byzantium. Allocation keys that could be used to run the National Lottery. Inconsistent application of conventions and philosophies. Arguments about what a person is, what a man-year is, what software is, what maintenance is, what anything is. Roundings and smoothings and straightenings and what-have-you. It’s not anyone’s fault; I think that traditional accounting theory, as implemented in software, did not allow for the pace of change inherent in organisational structures and strategies. And rather than accept that and change how we view things, we continue to work on past-predicts-the-future models and 15th century practices, somehow finding ways to shoehorn today’s realities into them. Different ways. Different even within a single enterprise.
5. We will be able to place value on social software, but we’re not there yet. The destination is clear, though.
I have a passion for social software. [You’ve noticed? Great :-) ]. And while trying to do something about it, I’ve been accused of many things. Being a VC darling. A wannabe A-lister. An A-list groupie. A time-waster. A moron. A rootless academic. Many things.
So why do I do it? Because I believe we have a real problem to solve. A War for Talent. Because I believe we have a real solution to that problem. One that provides transparent communication and thereby creates trust. One that leverages the wisdom of crowds and thereby reduces process cycle time and improves customer satisfaction. One that embeds strategy in process and helps us understand our value drivers and levers and how to use them. One that reduces attrition and increases stickiness, be it staff or customer or partner. One that simplifies workflow. One that minimises rework by reducing the opportunity for misinterpretation or mishearing. One that enfranchises everyone. One that reduces the time to train. One that simplifies the process of education. One that helps people learn how to do their jobs, and get better at doing their jobs.
But all that is hot air, when I can’t prove it on a balance sheet. Because the tools to prove it aren’t here yet. In the meantime, we need to keep looking at proxies. Proxies such as customer satisfaction metrics. Retention rates. Process cycle time reductions. Time to market improvements. Error reduction and consequent rework minimisation. Avoidance of duplication and its related wastage. Consistency in global processes. The list is long.
The trouble is, these things are hard to measure as well. Every investment made in an organisation has sponsors seeking to attach credit to their decision, and I have seen too many “value turf wars”. Unseemly attempts to hoover up any improvement in anything.
Until I find a static organisation with a stable structure and zero money for investments, I probably won’t have a decent test case. If there was such an organisation, then we could have the Petri dish for social software, and watch all the proxies move in the right direction as if by magic. And nothing to point to as the reason. Except social software.
So I have to live in hope. Continue with my passion. Because I believe.
I think it was Howard Schneiderman at Monsanto who said, many years ago, something along these lines:
When you turn down a request for funding an R&D project, you are right 90% of the time. That’s a far higher rate of decision accuracy than you get anywhere else, so you do it.
And that’s fine. Except for the 10% of the time you’re wrong. When you’re wrong, you lose the company.Â
My apologies if I haven’t paraphrased him with precision, but the tone and essence is for sure what he said. Call it an acceptable conventional representation of the truth, if you will.
There is a destination. One that values human capital and relationships and institutional knowledge. And we will get there. So I will continue to track the conversations on the blogosphere looking for signposts that will make it easier to get to the destination.
I’m all for measurement, especially when it comes to some of the outlandish claims about direct causality in the marketing sphere, but my reaction here is why bother? Charlene’s post is interesting insofar as it breaks down cause and effect but you still have to create meaningful numbers for each of these partially subjective judgements. My take is that it seems that the cost of implementation and maintenance of social software is so small that it falls into the “no regrets” category you cite above. Unless I’m missing something, the outcome of social software would, at worst, be insignificant so where’s the downside? If there’s no real downside and the cost is small, why expend so many resources evaluating an ROI? The answer is probably that there is no genuine buy-in to the concept and that absence is what has to be overcome and I’m not sure the generation of a positive ROI will overcome that.
I am totally with you regarding the sever lack of social software to solve current business issues relating to communications. In many ways you are right about social software needing to be measured by ROI, however in some cases (such as distributed software development) it is essential to have the communication and knowledge sharing tools to get work done. In these sort of cases the measurement is not about ROI but more on project completion to deadlines and staff retention on the project.
At the moment for example you could use source control such as perfoce with a sharepoint portal for comms (which is clumsy and only e-mails updates) or you could do the same with blogs, and wiki’s and IM in the same environment. Both need maintainance and both have their pain points. But things are getting better and businesses are actually trying to implement these sort of things. Sometimes they are successful, sometimes they aren’t…
I think that some of the reasons why they aren’t successful is because they actually all give people more work to do. Communications fails when it becomes time consuming over just getting on with something. It also fails when there are social aspects getting in the way such as people seeming to be unapproachable. Social software can break down some barriers, but where it does it usually builds up another in time and cost category.
I don’t see this time consumption aspect of social softwares disappearing, and I do wonder if it will be accepted as the norm over time as a means to creating control over potential social barriers between employees.
John, the reason why I kept away from discussing ROI was precisely that; I wasn’t even sure that the cost of trying to measure the ROI could be justified against the cost of implementing the software.
The issue is time. Human time involved in the conversations and contributions that make social software worthwhile. Without that investment of time, there is no value in social software.
So you then move to how to make that time available. And focused. Then you find that the biggest reason why the time isn’t invested is that people don’t believe in the value. And often, perversely, they don’t believe in the value because it isn’t an expensive ROI-ed investment.
Catch-22.
I think that’s why a fundamentally emergent model needs top-level endorsement, even sponsorship. People will only put the time in when that happens.
This is not a blanket statement, there are numerous cases when the time investment is made because people believe it makes their jobs easier. But that’s as a result of education.
You have to get to the time. There appear to be two routes to getting it. Executive sponsorship. Education.
Am I making any sense?
Sarah, hope your cold is better. I think people don’t put the time in because they don’t understand how it makes their jobs easier. Whenever you try and collect information from people there is inertia, just look at the return rates for most surveys. Of for that matter voting levels at western elections.
the apathy is endemic.
So it can be with social software. Before we had all this, we had the same resistance to other forms of discipline and information sharing within IT. JCL management. Configuration management. Source code management. Even contract management.
My perception is that 87% of the expenditure on Y2K was driven by this kind of behaviour. I found it hard to believe that people didn’t know what contracts they had, what systems they had, what interfaces, what maintenance agreements, even what kit.
This is particularly true in environments where the firm operates as a collection of fiefdoms. High degrees of independence and autonomy, high levels of prima donna and not invented here.
You must want to collaborate before collaboration has meaning. Sadly this is often not the case,
We could fall through the floor on this one. ROI is far too simplistic unless one can examine the values involved in creating, evolving and sustaining relationships.
Assets are ashes without relationships. I see a chink in thinking as diverse as Baruch Lev and Jim Grunig to adopt a relationship value view (not to be confused with CRM – a management ploy at best).
The values involved are all intangible for the accountants and the phrase ‘goodwill’ should be offset with ‘badwill’ on the ballance sheet.
Then the ‘clear-enough relationship future, alternate relationship futures, a range of relationship futures, and true relationship ambiguity’ is significant.
If relationships are founded on a nexus of ‘personal values’, and in Social Software, the blogger, for example, is the nexus of values personified.
An organisation is not ‘the nexus of contracts’, it is the nexus of relationships – and that is where the real power an influence, not to mention asset lies.
Sorry to be late for the party – yesterday was the day of my monthly full-hard-drive backup! Having said that, I think the appropriate way to enter this conversation is on the topic of TIME. JP, you are definitely right that much of this argument pivots around the investment of time; but I think you are missing a significant perspective on the investment of time.
I would like to argue by analogy from an example I have encountered many times, which is the introduction of new technology (hardware or software) into a pre-college classroom setting. At the end of the day, there is only one question that matters to school administrators; and it is one of the few questions that clearly honors the interests of the teachers. The question is: “If I am a teacher, what is it that I am supposed to STOP doing to create the time for using this new technology properly?”
I hope you see the point. Whether we are talking about teachers in classrooms or just about any other work setting, the investment of time has to be modeled as a zero-sum game. When you are dealing with a work setting in which everyone is already stretching their temporal commitments beyond the limits of sanity, the sad truth is that questions of the investment of time cannot be addressed. The real Catch-22 is that there is no time to discuss how time should be invested!
The bottom line, then, has nothing to do with what a new technology brings to the table or how we can talk about that technology in terms of concepts like value. More important is that every work setting has a foundation of normative behavior, what Graham Button calls the “immutable work practices.” The REAL bottom line has to do with whether one can find a suitable way to “fit” a new technology into the context of that normative behavior. If the fit is there, then the investment of time becomes less of an issue, because the technology integrates with those immutable work practices that are already there. It is only when there is a poor “fit” that we have to worry about “change management” to accommodate the new technology; and then all the ugliness of zero-sum time investment stares us squarely in the face.
Let me now try to address this in terms of your “real problem to solve,” that “War for Talent.” The fact is that Talent needs “fit,” too. Otherwise, it is likely to be snuffed. Dizzy Gillespie got his first major professional break playing for Cab Calloway (I think); and, as anyone familiar with jazz can guess, that was a pretty lousy fit! Fortunately, Diz was able to build up both his chops and his social networking to a point where he could break with Calloway and cultivate an environment with a better fit. Jazz history is full of stories like this.
Unfortunately, the story of Diz the trumpeter is not that different from the story of Adolf Hitler, the aspiring painter. Like Diz, Hitler ultimately found a conducive environment. The problem was that it was not an environment for painting, and we all know what happened as a result. I offer this example to demonstrate that the “plumbing” argument may be dangerously naive. Yes, there are cheap ways to get into social software; but that is primarily because very little (if any) thought has been given about the consequences that can emerge from the use of social software. The bottom line is that, like any other social setting, social software needs some kind of framework for governance; and that is not yet in the plumbing. Furthermore, because the design questions have not yet received any serious consideration, we really have no idea what the expense will be if those questions ARE taken into account.
About a year ago I was discussing the role of social software in the operations of a small humanities-centered college here in San Francisco. I said that the most important thing about creating virtual spaces is that the be SAFE, just as there should be a presumption of safety in any classroom that discusses sensitive topics. This safety criterion is, I believe, a key element in the distinction between liberal and illiberal democracies that now concerns George Soros. Until it is duly recognized by the providers of social software, I think we all need to be very cautious about the adoption of their products.
Many of the arguments that dispense with the need for ROI represent a failure to join the dots between the potential for value implied by social software and the inherent value we already know exists in the difference between accounting balance sheets and market valuations. If we’re not prepared to make the connection then ROI will remain a mystery.
Nicely put, Dennis. Which is why I entered the fray, joining you guys in investigating this aspect. Hitherto I had kept away from the ROI issues.
Maybe what I am saying in the last post is that the connection you refer to, the connection that can bridge the gap between balance sheet and market, is about human capital. And, as you say in your comment on the Forrester blog, the measurement of human capital is a burgeoning discipline with a lot of learning to get through.
Stephen, I am still working through what you’ve said, finding some of it hard because of the implied Innovator’s Dilemma there. Your arguments do not seem to allow for radical change in any environment. But that’s unfair, I am still digesting what you wrote. More later.
JP
Yes you made sense and I’m not surprised (indeed I expected) that your view was that an ROI exercise was pointless (except for internal marketing puproses) if the cost of calculating it how ever vaguely exceeded the low cost of the social software. But I agree totally with Dennis as I indicated in my first comment – too many claims are made without an ROI proof because this allows causaliites to be suggested without testing.
JP, actually the Innovator’s Dilemma implication reinforces my point if you remember what Christensen proposed by way of a solution, which is that a large organization should create a “safe” niche within which radical change can be allowed to develop and receive a fair assessment. In other words, if the change is too radical to “fit” the existing organization (which is probably tautologically true if the change is radical!), create a NEW PLACE in the organization where it will fit. Hey, it’s not for nothing that the SHORTER OXFORD ENGLISH DICTIONARY now has an entry for “skunk works!” Traditionally, though, skunk works have been sandboxes where cool is the only currency. These days SOMEONE needs to think about consequences, too!
Hi,
You’re clearly a person who gets ‘it’. I’m organising a weeks worth of events related to the above in my organization. Lots of fun and challenging tracks. Would you be interested in coming along to talk about social software and its place in the enterprise? We’d love to have you come along…
All accounting methods and practices are the result of agreed-upon, shared and codified assumptions about social psychology.
There’s no reason i can see for why the interaction aafforded by interconnection will 9eventually) change work design massively, which will then need concommitant changes to compensation.
Likewise with many ‘bedrock’ assumptions about accounting in general, as interactivity, disintermediation and re-intermediation make value flows look and act differently .. such as when variable pricing into a range of niches becomes common-place.
probably be a while yet, though … not until it becomes abundantly clear that current GAAP has lost too much effectiveness to portray and report those flows adequately.
Hi JP,
I’m interested on your opinion of enterprise 2.0 and would love to introduce you to David Lavenda of Serendipity Technologies (http://www.srndpt.com/). If you’re interested in what David has to say about enterprise 2.0 or would be interested to learn more about Serendipity and David, please let me know and I’d be happy to introduce you.
Thanks!
Meghan