Innovation Study Tour – Part 4: Making an Enterprise Social Network Work

The second part of the Hargraves Innovation Study Tour continued on 20 September right in the heart of the Sydney CBD at the Microsoft Store on Pitt Street.

L: Before opening on Pitt Street, but you can see the staff in the corner having their morning scrum. Downstairs is the shop floor, upstairs are the training and presentation rooms. R: The toughest job in the store: “Go stand in the window and play games.”

Scott Ward of Digital Infusions spoke again, and as mentioned previously it was a continuation of an information rich and fascinating discussion. I wrote about Scott’s take on the three E’s of explore, engage and exploit in social networking in organisations. But why bother? What’s the impact really? And, even if you decide to be all ‘digital’, how do you make it happen?

Why bother with social networking and all the new fangled digital stuff? Surely it’s all a bit of fad and we’ll be back to email chains cc’d to everyone and proper two-hour pre-planning meetings to organise the meetings before you know it. Right?

Capgemini Consulting undertook a two-year study of nearly 400 large businesses to quantify any links between digital maturity and financial performance. In good socio-economic research practice they developed a Four Quadrant Model™:

The four type of digital maturity. Capgemini Consulting.

The x-axis is a measure of how well structured the management and leadership capabilities are to drive digital transformation in an organisation. The y-axis is a measure of how much is being invested in digital technology. So you can start with the Beginners that have neither the management or investment. The Conservatives have good leadership and management structure, but haven’t invested. The Fashionistas are investing heavily, but without leadership or structure. And of course the ideal state is the Digirati that have both good governance and are investing strongly.

The report then demonstrates that firms classed as Digirati are 9% better at generating revenue from their assets, 26% more profitable, and have a 12% higher market valuation.  Of course different mandates and benchmarks make comparisons between large multi-national for-profit businesses and public-good government organisations fraught and it would be interesting to see a similar study focused on government. However, the study demonstrates that large complex organisations function better when they combine both transformational leadership and investment in digital technology. Why would a large complex government department or agency be different?

So, for the purposes of writing this blog, let’s say yes, let’s get into some enterprise social networking… Scott has developed a strategic framework for implementing and monitoring the impact of enterprise social networking in an organisation called BITIL™, which stands for Brokerage, Integration, Trust, Incentives and Leadership.

Scott Ward’s BITIL Framework for implementing an enterprise social network. Digital Infusions via Hargraves Institute.

Brokerage.

Provide the skills for using the enterprise social network and help make it part of everyday business. Scott gave examples where new social networks had simply provided space for people to post ‘Stuff My Kids Say’ or ‘Today’s Sandwich’. That may seem a bit goofy and not work-related, but it meant people were learning and practicing their posting skills – and also building the online connections that are the heart of an enterprise social network. More formally, some organisations ran events to encourage use of a new network such as idea challenges and ‘ask an expert’ online sessions. (Imagine Reddit-style Ask Me Anything sessions with the Executive…)

Integration.

Deliberately move business tasks into the social network. Put essential information on the social network. Make collaborative tools available to encourage crowd-sourcing ideas and solutions, running polls, etc. Build an experiment culture where staff can try things out in the social network and it is accepted that some things will work and some things won’t.

Trust.

Staff have to trust the network and have respect for each other. Consider the 90:9:1 rule… 90% of people will just observe, 9% will post a response, 1% will post original material. Someone posting always has to consider and respect the 90%. The classic, if somewhat stereotyped, consideration can be applied: “would my grandmother be embarrassed by this?”. Similarly people shouldn’t be slammed for asking awkward or confronting questions – remember the new travel expenses policy example – a well managed frustration can be a rich source of ideas and solutions. It would be easy for a government organisation to get all Governance Heavy™ about this, but that would defeat the purpose of the network as a connected ideas marketplace in the first place. The recommendation is for some simple guidelines which are more ‘do’ than ‘do not’ (e.g. “Do think about others” versus “Don’t post any of these long list of things…”). Discussion in the audience also pointed out the APS already has a lot of rules about behaviours like the Code of Conduct, so there really shouldn’t be a need to add more, just apply what we already have.

Incentives.

Encouraging adoption of new processes is always a challenge (and a whole topic into itself). Incentives can help, but they need to be well considered. Scott spoke about formal and informal incentives and used Gabe Zicherrman’s SAPS model (Status, Access, Power, Stuff) to describe the order that people want their rewards. Turns out that providing ‘stuff’ like pay bonuses or gifts is the least effective motivation (and can actually be negative, especially for creative work). In the case of encouraging use of a new enterprise social network, this can relate to the brokerage concept with online events like a photo competition to promote use and rewarding adopters with recognition.

Leadership.

In Scott’s opinion this is the most critical element of the BITIL model. He considers that the senior/executive leadership must have 100% compulsory involvement for the enterprise social network to work – people need to see their leaders involved. As Charlene Li put it:

“The problem was simple and obvious – because the top executives didn’t see collaboration and engagement as a good use of their time, employees quickly learned that they shouldn’t either.”

There are several other articles with similar themes (here and here for instance), but this isn’t to say that the Executive should be Twittering updates of every meal. Scott’s recommendation is:

Leaders don’t need to be interesting, but they must be interested.

A leader can start with the 9%: respond to ideas, be supportive, manage conflict constructively. Above all, have a plan, don’t assume that magic will simply happen when the new enterprise social network is switched on. One obvious question was raised about was about where does all the time come from to be active on a social network? Perhaps that concern comes from stereotypical portrayal of teenagers as being constantly on social media and that just wouldn’t be a good look for a serious professional. But participation doesn’t have to be completely time consuming. A regular contribution, even once a day, is enough to be seen as an active participant and demonstrate the vital interest that encourages others. Very importantly, Scott recommends, leaders must be authentic – they should write their own stuff. People are smart and easily recognise the spin in corporately generated messages.

And with brains already brimming with ideas we meandered to the big kahuna of innovation – Google…