Thinking about curation in the enterprise

[Esther Dyson, Clay Shirky, Marc Benioff and David Weinberger, people I consider as friends, have said and written things that have influenced this post considerably; the curated conversations between Umberto Eco and Jean-Claude Carriere, as recorded in This Is Not The End Of The Book, have also been very influential. Finally, Steven Rosenbaum’s excellent Curation Nation helped me bring all this together. Disclosure: I have co-invested with Esther in at least one company, and I now work for Marc Benioff.]

Content may be considered king, but distribution has always been the hand that rocked that particular cradle and ruled that throne.

And distribution had its kingmakers: those who weeded out the vast majority of possibles and probables and selected the content that satisfied their taste, then ordered and organised that content into a cohesive whole, then made that content accessible to the particular audience that mattered.


[For some time now, I’ve pondered about the distinction between filtering and curation, and my thoughts on the subject are as follows: curation is what happens when filtering is augmented by humans, sometimes expert, sometimes amateur, always passionate.]

Curation is not just to do with libraries and museums. As Steven Rosenbaum describes in his book, Esther Dyson curated a magnificent conference, PC Forum, for many years. And a great journal, Release 1.0; while the journal lives on, the conference, sadly, is no more.

Talking about conferences, I’m looking forward to Dreamforce next week, the annual Jamboree for those involved in cloud computing. Last year, at my first Dreamforce, a columnist from Barron’s described the event more as a political convention than a tech conference. I have some sympathy with that view, it feels like you’re part of a global movement. And that’s what Dreamforce represents, global transformation in the way enterprises drive, participate in and consume information technology.

This year’s line-up includes, amongst others, Eric Schmidt, Neelie Kroes, Vivek Kundra, Neil Young, Alanis Morrissette and Jay Leno. [You may still be able to get a free pass to the #df11 keynotes and Cloud Expo go to ]

Where was I? Oh yes, curation. The SOP of curation, to paraphrase Rosenbaum, is Selection Organisation and Presentation. Curation is about human beings adding their passion to the filtering process, in order to select what should be experienced, put the selections into some cohesive order and then to make those selections accessible to the relevant audience.

Curation is not something that is restricted to libraries and museums; people curate things all the time, all around us. Every DJ is a curator. Every reviewer is a curator.

In a networked, non-hierarchical world, every one of us is a curator. As I’ve mentioned before, I use twitter as a curation device. The people I follow have a good idea of what I’m interested in, and they make recommendations to me. Sometimes they do it via Direct Messages or DMs, sometimes through @messages others can see, sometimes by ReTweeting or RTing; one way or the other, they make sure I am made aware of the thing they want me to be aware of.

Personal social networks are powerful curation devices, a theme that has been tackled by Clay Shirky time and time again. David Weinberger adds a different twist to it in his writings, looking at the role of taxonomies and folksonomies in this context.

Which brings me on to Esther Dyson, whose writings about the future of search have been at the back of my mind all through my thinking about this. In Curation Nation, Esther quotes Bill Gates as saying (at a private dinner) “The future of search is verbs”. She then goes on to explain that “when people search…they are looking for action, not information….they want to find something in order to do something”. If you get the chance, you should read Esther’s writings on the future of search, just google it. In fact there may still be a YouTube video summarising her views.

This action-orientedness that Esther refers to is something that Marc Benioff is passionate about, a passion that shows in the way he thinks about enterprise software. Information flows in the enterprise should always enhance the ability of participants to do the right thing in the right place at the right time.

Which lets me segue neatly into the crux of this post, curation in the enterprise. Every enterprise has its own variant of curator, people who help decide who sees what, when, and in what shape. Information overload is everywhere, the Shirkyian filter failure is everywhere, and into the valley ride the usual six hundred, theirs not to reason why. So in order to understand how enterprise curation should take place, it’s worth looking at some extreme forms of enterprise curation as practises today. There appear to be four main forms:

The Signal Booster
The Spreadsheet Jock
The Soulmate
The Sidler

The Signal Booster obtains power by PowerPoint, thinks in bullet points, rarely knows more than what’s on the slide. Acts as a mediation layer between those that do and those that decide. Useful, but often not valuable; can’t take decisions, hasn’t got the depth of knowledge, runs the risk of exhibiting passive aggressive behaviours and operating weak vetos. A fence-sitter who works hard at joining the winning team when victory is in sight.

The Spreadsheet Jock believes there’s safety in numbers, that firms can be managed by algorithmic trading. Runs the risk of past-predicting-the-future, of being anchored, framed and constrained by the underlying assumptions, doesn’t always understand the assumptions, whether implicit or explicit.

The Soulmate is a crony of the powers-that-be, using that association to derive second-order power, and has an unusual effect: an inadvertent tendency to ensure that anything the CEO doesn’t want to hear doesn’t make it to the CEO. An unfortunate and unintended consequence of Soulmate speak, as they spend time interpreting the CEO’s desires and making those desires unintentional filters.

The Sidler is a rare beast, someone who can only thrive in the rarefied environment of “briefing” cultures. They are often seen alongside the CEO, whispering in their ears, advising and commenting on the status of things they aren’t involved in. Sidlers are chameleons, sometimes boosting signals, sometimes driving spreadsheets, sometimes being soul mates. But always sidling.

All these are extreme forms of enterprise curator, responsible for deciding what information is accessible, to whom, when, and in what shape.

And all these are fundamentally inefficient models of curation in the enterprise, since valuable information gets mutated, suppressed, discarded. Most of the time this is done without any malign intent, an outcome of weaknesses in the system and in the process.

If they were that inefficient, why did they exist in the first place? Probably because the structure of the firm, the culture of the people and the technology that supported them were all incapable of achieving much else. In large hierarchical organisations, some form of summarising and filtering takes place in all information flows, from top to bottom as well as from bottom to top.

But today it’s all different. The promise of the social enterprise is a remarkable promise. And this promise is based on a number of key characteristics:

1. The social enterprise is fundamentally based on networks of connected people, within the enterprise, beyond the enterprise into the supply chain and trading partners, and extending all the way to customers and their customers. And products and services.A web rather than a chain, the social enterprise is somewhere where everything and everyone is a node on the network.

2. This reduces the distance between the firm and its customers, between the designer and the consumer; it simplifies the links in the supply chain; more importantly, since it is built around the concept of two-way communication, customers are themselves designers; trading partners are themselves designers. Everyone’s not just a curator… Everyone’s not just a designer… Everyone tries out products and services, and provides active feedback.

3. This ability for two-way communication means that conversations take place without any loss of detail. The need for summarising is reduced, since the cost of hanging on to the detail is low. If you’re looking at a conversation, you can drill into the customer name or the order or the complaint or all of them. Summaries are now just forms of presentation, always with the ability to drill into the detail as required.

4. Drilling into the detail was historically complex for reasons other than just the cost of doing so, or for that matter the distance expressed as the number of levels in the organisation, the divisional silos, and so on. We had the added complexity of security systems that did not differentiate between systems of engagement and systems of record, and as a consequence didn’t know how to handle entitlement safely and securely. Good social enterprise implementations solve that elegantly.

5. It’s not enough to have access to the information, that still doesn’t solve the overload problem. So we need access to expertise. Real domain expertise, sometimes with letters after names, sometimes not. Peer-reviewed and peer-ranked expertise. Again something the social enterprise is designed to do. Find the people that are acknowledged rather than asserted experts, experts because of what they do rather than who they are.

6. Talking about doing, every customer is different, every product and service is different, every problem is different, every situation is different. This makes the traditional process mindset we problem in itself: we spend enormous amounts of time handling exceptions because the process is inadequate. Processes are just not repeatable enough, the volume of exceptions is too high. As we move from the Hit Culture to the Long Tail of problem solving, we need more and more experts, “long tail experts”.

7. Which is where the social enterprise comes into its own. Because everyone can be an expert. Linus’s Law (“given enough eyeballs, all bugs are shallow”) plays out very well in any community that is built to scale. So every staff member is empowered to solve problems. Every trading partner. Every customer.

8. Networked non hierarchical models. Involving everyone: staff, partners, customers. Two-way communications. Easy access to domain expertise. In an environment where aggregation takes place without any loss of accuracy or of the source data, where you can “follow an order or a complaint, safely, securely, efficiently, effectively”.

The social enterprise is about bringing the learning from consumerisation and community action and applying them to an environment populated by a new generation using new technologies, social, mobile, open and cloud.

Not surprisingly, I’ve now heard Marc Benioff speak about his vision a number of times. More recently, I’ve come to realise that the Social Enterprise is to traditional software what Skype is to traditional telephony, what Paypal and Square are to traditional payments. Quick and effective. Riding over the top of existing infrastructural investments. Focused on simplifying the customer experience, eradicating traditional frictions, reducing the distance between the customer and the firm. Engaging with the customer rather than with the back office.

In my next post on the subject I shall spend more time on precisely how these characteristics manifest themselves within and beyond the enterprise: Why social, mobile, open and cloud matter, individually as well as collectively.

In the meantime, comments welcome as always.

9 thoughts on “Thinking about curation in the enterprise”

  1. JP — Great piece. I think your characteristics of the social enterprise a right on the mark.

    Of course, the tools aren’t quite there yet. At HubSpot our networking tools (mainly Twitter, LinkedIn, FB, and a heavily used Confluence Wiki) do an excellent job curating flows of information internally, but not yet between our organization and our customer.

    Customers certainly find us through their networks, but we have yet to find the right mix of tools to serve customers through the network. I think we’ve run into two main problems:

    (1) In order to create the nodes of expertise you talk about in an enterprise setting, the network must be able to recognize the expertise, which means collecting, organizing and distributing large amounts of disparate data across the network. We haven’t found a system that can do this well. Salesforce feels like closest — but it’s not there yet, and, more importantly, it means a massive investment in moving all that data into one place, then customizing that place to fit your business. Lighter-weight, more modular solutions would be preferable, but they aren’t far enough along yet. If you go that route, you end up building your own solution.

    (2) Relationships w/ customers demand high fidelity networks, but the networks that work for curation today are relatively low fidelity. When our customer support reps browse tickets from customers, they need to see every ticket. In their base, most robust forms, our existing networks don’t support that fidelity — when I browse Twitter or FB I don’t expect to see all my friends posts.

    There’s lots of work being done to solve both these problems, but there’s a ways to go.

    Looking forward to the next post!

  2. What a fascinating post and one which I can identify with coming from the world of international broadcasting. That traditional form of shouting across the border is proving less and less effective. Those who want to influence realise they have to understand the information needs of the audience.

    Several conferences I attending last year talked about curation. In the broadcast world, the equivalent term is producer, someone who build a logical thread into a production whether its a performance, a debate, a conference or a radio program. That’s the reason that Esther’s conferences are so interesting. The context becomes more interesting when the producer/curation knows just how much production is needed. Too little and the train of thoughts are disjointed. Too much and you get something like a TV talk show that has clearly been rehearsed or even scripted.

    I’ve been experimenting with enterprise publishing models to replace the current international broadcasting model. With some issues, the conversation with and within the audience goes on far longer than the broadcast itself. Broadcasters seem to be struggling to understand that just launching an idea or topic on an audience without some kind of curated follow-up is a very expensive way to share an idea or concept. Al Jazeera English has made some superb documentaries on African immigration into Europe. But they just present the results and there is no curated conversation that follows on, so the hard work is quickly forgotten. The same goes for the voice of America which has been playing around with a multimedia site called about an important issue. But because there is no discussion with the makers a couple of dozen people have watched the videos. Its VOA broadcasting, not on shortwave, but onto a social media platform.

    I learned a lot from speaking with farmers in Northern Benin recently who were critical of their local farmers radio station. They said they appreciated the efforts made by the station to produce programmes for them. But they regretted the fact that not enough was done with them. As well as having curated information broadcast to them, they wanted to bring their ideas and suggestions to the stations for follow-up discussions. It taught me a lot about the long-tail of factual content. The radio stations announcers wanted software to help them perform, putting information into a continuous stream. The audience, while they appreciated being surprised by something new, also wanted to get a briefing. That required more of a wikipedia style service to extend the lifetime of the content.

    Social, mobile, public and the cloud are becoming so important in a world where access to information is impacting so much on our wellbeing, relations and mutual understanding.

    As ever, I look forward to further postings.

  3. May I throw a spanner in the works…the Social Enterprise: Non hierarchical curation involving everyone creating value requires designing an organisation culture focused on results.

    For example people managers (the corporals and sergeants of management) become obsolete. Instead employees set quarterly goals — stretch quotas like sales people — and complete engineering or customer service work where feedback can be measured by metics like adoption of new software features or customer satisfaction (née software sales).

    How many exec managers (or employees) are ready to give up having people managers with prodding sticks to discrete work around measurable metrics guided by domain experts?

    [How many org’s have the wriggle room from shareholders to stomach failures like MSFTs Kin / HP’s WebOS Tablet?]

  4. Rick, re your two points. Expertise can be asserted, earned, bestowed, and in each case the granting authority (even if it is self) must be clear. There are good lightweight ways of doing this already, and sharing the credentials across a network is not difficult either. This is where crowd sourcing and ramification meet. I have already made my feelings clear on the subject in my session at ReadWriteWeb earlier this year, when I spoke about “the lipstick of gamification on the pig of work”. Once mastery is correctly flagged and tagged, it can be searched for. There are already a number of firms offering Lightweight solutions using Salesforce. This is no longer a hard problem.

    Similarly, when it comes to the high versus low fidelity question for customer service, perhaps you need to take a look at what’s been happening with Service Cloud. I think you’re raising two different questions here, one to do with privacy and confidentiality, one to do with integrity of information. Both are solved today. Let me know if you need to know more.

  5. Jonathan, couldn’t agree more. I too remain fascinated by the conference – community – in perpetuity model. As you will see from my third post (still working on the second in the series, waiting for a few more comments before I close it) I am going to expand on some of the issues you raise.

  6. Clive, not sure I can agree. The problem with the prodding stick approach is that it does not work. Even if it did work it will not scale. Managers will have to adopt the new ways of working, or find themselves in defunct companies. Because the competition will not wait.

  7. JP, agreed. Prodding stick approach is no match competitors with a culture for managing “lumpy work”. Do we have a green light for designing systems for “Curation in the Enterprise” without a movement for educating organisations to manage work differently (Drucker 2.0 / Goldratt 2.0)?

  8. AMEN! I think you nailed it. I’m really surprised that I haven’t seen more activity surrounding your post. It’s always very exciting to me to see a company or organization that gets most of what you are talking about. Makes a big difference and people can feel it whether they understand it or not.

Let me know what you think

This site uses Akismet to reduce spam. Learn how your comment data is processed.