Posted by: karisyd | April 1, 2012

Tool or infrastructure? (or: the question concerning social media ROI)

Do all Enterprise Tools have to solve single quantifiable problems?

James Dellow (@chieftech) asks this question in his latest post. This is an interesting question, and the answer is of course (as James implies): no, they don’t.

But the underlying problem when taken to the social media field is a bit more complicated. In my view it stems from a wide-spread lack of distinction between tool and infrastructure:

  1. Tools have one (or a few) specific purposes. Tools solve a problem. And solving the problem can be measured. Hence, tools are predictable.
  2. Infrastructures don’t have a well-defined purpose. Infrastructures are largely undefined, general purpose, but they open up a realm of new possibilities. They are unpredictable!

Most Enterprise Systems as we know them fall into the first category: ERP, CRM, ect. But social software are infrastructures! Social software opens up a space that needs to be filled by the users (see: Nutzungsoffenheit).

Most managers know and like tools; but infrastructures seem to be misunderstood:

  1. Tools solve a problem, tools help constrain and control the users in what they do. By giving a certain set of tools to the user (often those that have ways of using built into them: best practice!) an intended outcome can be achieved. Because of this, when introducing a new tool, we can seemingly predict the resulting changes (ROI!).
  2. Infrastructures on the other hand help open up new spaces for users to fill and define in practice. What they become and the kinds of context-specific purposes they acquire, is generally the result of an open-ended process of sense-making among the user group. This is why the same social media infrastructures result in such varied uses. But when we first see them, we don’t quite know what they are for.

And this means that if we don’t know what they are going to become, how can we expect to predict and measure in advance something like a ‘social media ROI’ in all seriousness?

Good, visionary managers know this. They understand that in order to see it happen, one has to let it go and let it evolve – in what I have described as a series of small safe-to-fail experiments. Success of such initiatives requires leadership, not management-as-constraint-and-control.

At the recent Yammer on Tour event in Sydney and Melbourne a number of those visionaries shared their success stories: Simon Townsend from Deloitte, Simon Terry from nab, Andrew Hedges from Westfield.

The question “how do you measure/predict ROI of a social media initiative?” came up a number of times.

The simple answer is: “you don’t” – at least you don’t predict. The more balance answer is: You don’t try to predict, but of course, you observe and measure what happens next. But we will have to rethink the way we measure the success of infrastructure-enabled change in organisations as opposed to more conventional tool-driven change. This is a question that is still largely up for grabs – something to work on…!


  1. I would like to add another view towards the point: “tool vs. medium” – This discussion is already quite old, and comes from the field of Computer-Supported Cooperative Work (CSCW) – see for example: Koch, Michael (2008): CSCW and Enterprise 2.0 – towards an integrated Perspektive. In: Proceedings Bled eConference. I myself like “medium” better than “infrastructure” because it better relates to communication.

    • Hm, I don’t quite agree. Granted, many of the issues that pop up now have been dealt with in CSCW before, such as Nutzungsoffenheit or awareness, but the are distinct differences in social media that have not been part of the phenomena CSCW typically dealt with.
      1) Scale and scope: the proliferation of social media, the bottom up nature of adoption, the wide-scale networking nature, the hosting online (cloud) and subsequently the social phenomena that arise from this have not been seen before and cannot be captured by established CSCW work, because CSCW takes a much more narrow user/group view (overall).
      2) Perspective: the way social technologies are developed and rolled out is very different. We are not talking about groupware space, we are talking very open, interconnectable, ubiquitous and accessible platforms that become part of people’s everyday practices. Again, the social phenomena (after all that’s what this is about) that flow from this, at least I did not find in CSCW.
      3) I don’t think that Medium quite grasps the phenomenon. Infrastructure is a much more comprehensive concept. It’s not new either, but I’d like to refer to the work by Susan Leigh Star, an American Sociologist. This helps in my view more for understanding these phenomena than the notion of medium in CSCW. Infrastructure does not refer to a technology per se, but a phenomena, but medium does.
      Much remains to be done, in my view…

  2. Great post, Kai. As an extra dimension to this I used to talk about the “infrastructure” part in terms of form versus function, meaning that people sometimes mistake infrastructure (or the medium) as a tool. In part, I think this creates an expectation that if firm x gets y from it, then that’s the only thing it is good for (tool view). The other challenge with social infrastructure isn’t that we can’t measure ROI, its just difficult to measure without using non-linear metrics (e.g. SNA) and potentially expensive to measure due to the micro-benefits involved (see Clay Shirky on what he calls “Coase’s Floor”).

    • James,
      I completely agree on the first point. Regarding ROI my point is that predicting ROI in any sensible form before embarking on a social initiative seems impossible under the conditions. In later stages, it should be relatively easy to argue for ROI by finding success stories. It should even be possible to measure in a more structured way the ROI of a social initiative. How well we will be able to devise more general or formal instruments that are not case specific remains to be seen.

      • Yes, on the point about predicting outcome and value. This is why I would classify them as design problems – so I do think you can shape outcome and build business cases based on current opportunity costs, but its hard to have absolute certainty around emergent systems (plus you might miss something of value that appears).

  3. Great post Kai, I just stumbled upon my feed reader.

    One remark: I guess social software (infrastructures) will probable by standard in the enterprises, like email is today. A comparison:
    – Anybody may be free to use email according to his/her own practice and processes (nobody asks for ROI of email)
    – However, managers or others may propose certain ways of how to use email in certain processes (who will send which information to whom at what time, …)

    Currently, I see the same phenomenon to happen with social software in the enterprise, too.
    – Anybody may be almost free to use it for whatever,
    – but still managers (or others) want to instantiate certain kind of practices / modes of use.

    Best from Graz, AT

    • Hi Alex,
      Thanks for the comment. Agree completely. Social software will blend into the background just like eMail. But just like eMail we will see quite different (emergent) practices across organisations. Much more different then with purpose-built tools.

  4. Excellent post Kai. Your points resonates with my hypothesis that measurement of social ROI can only be effectively calculated based on how we are currently calculating natural capital. Pavan Sukhdev’s study on the economics of Ecosystems and biodiversity talks about measuring the value of ecosystem services. We can talk about measuring social capital starting from the same context. Once the social ecosytems are adequately developed, what services would they provide to the various systems ( of engagement, records, transactions)? What would be their value if they stop providing such services. These are inchoate thoughts. I am expanding these thoughts in my blog post. Would love to hear your thoughts. You have a wonderful blog! Thanks to Peter Williams, who recommended your blog in our internal yammer groups!

  5. Kai,
    Please ignore my previous illegible comment written in spurts. Here are few of my thoughts as I began to chew the thoughts in your post.

    While the metaphor of infrastructure lends itself to the openness of the systems, I believe it still limits our deeper understanding of E 2.0 systems with its rigid, imposing structures. Infrastructure limits its scope to the base platform, ignoring other components which interact inside and outside the infrastructure. The metaphor of architecture,on the other hand, I believe, is quite inclusive as it includes the entire of gamut of interactive components which help us define, what we collectively call ‘architecture’. Infrastructure is either deployed or decommissioned.

    Architecture, on the other hand, grows organically, based on live-feedback loops, emerging from the interactions of spaces with the objects. Spaces, which play a major role in determining the patterns of interactions in these E 2.0 systems, lends itself to the architecture metaphor. Social spaces have been defined through warren and plaza architectural designs. ( See for detailed description on warren and plaza designs.

    It’s quite hard to imagine infrastructure blending into the background. I find it easy to imagine architecture that way. In fact, one can call it architecture only if it does. Of course, these are semantic differences. However, as a practitioner(rather than consultant), I feel they are giving us insights into the mental models we bring when we approach these technologies. Given the traditional baggage carried by the word architecture, I find the word infrastructure quite enchanting!

    2) On Social Media ROI – When green economists decided to confront the issues of climate change, they were quite concerned with the apparent economic invisibility of nature. When business executives agreed that Social Media can no longer be ignored, they were quite concerned with its economic invisibility. If natural capital could be measured by ecosystem services rendered by nature, can’t social capital be measured by the ecosystem services provided to the organization by infrastructures ( or architectures, if you agree) ?

    While I absolutely agree with your point that predicting ROI in any sensible form before embarking on a social initiative is quite futile, I think the only way it could be done in that case is through measuring the ecosystem services that are offered by a fully grown, mature social enterprise architecture. Although I see several difficulties in taking this measurement approach, I am quite curious to see through this.

  6. […] this case, especially more broadly though, it served to as a great example of how some technology is becoming infrastructure in our […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: