Do you know your staff?
Finding up to date useful information about people in your organization can be challenging – Who is the social policy officer in Bolivia? Who speaks French and has worked on conflict prevention in west Africa? Who contributed to the position paper on HIV and breastfeeding? Who knows.
We already have a plethora of different information systems about our people who are only partially connected. We have our HR system, our IT/Network directory, our telephone directory and we have various email lists and spreadsheets managed by individual offices and sectors.
Thanks to some great work by our IT colleagues we just launched a new staff profiles system in UNICEF at the beginning of the week that seeks to help with some of these questions. It is built using the Office365/SharePoint Infrastructure that the organization has already invested in, but linking to other systems as needed. I just wanted to share our experience in trying to come up with a useful profiles system, which we have called “Who’s Who in UNICEF” in order to address the challenges we currently face in knowing about our staff, in case it is of interest. Our new system was an attempt to pull together some of the different existing information sources to provide the best (but far from perfect) information that we could get about the people who work for us, in order to make it much easier for people to find out about one another from a single place.
Pulling this all together was tricky – but basically we built on the idea that useful information about any individual can come from three broad types of sources:
1. Official system information such as our HR systems and our internal directories. This is fragmented into several non-coordinated systems in our organization so we needed to map which is the “official” source for what information and then combine them together. This is already a step forward compared to what previously existed but official information is limited in that some of the more interesting things are not digitized or confidential where we work (such as performance evaluations, training records, past work experience) so can’t be shared in searchable profiles. Plus they often miss out key experiences and expertise – and as we found out through this exercise – official information about our people is often unreliable – wither out of date or just plan wrong.
2. What you say about yourself. A lot of people profiles in social media tools and enterprise social networks give you the opportunity to also enter information about yourself – what do you do, what are you working on, what are your interests, languages spoken, a flattering headshot etc. We merged these features with the “official information” giving staff the chance to be able to present themselves and their skills beyond what the organization officially knows about them. This makes the profiles more interesting and more useful because it makes them more “personal”. But self profiling also has its limitations – just because I say I have a particular skill or set of knowledge doesn’t mans that I actually do have it. And a superficial description while helpful doesn’t give the full nuance of exactly what your skill is.
3. The trail people leave in the system. One interesting way to figure out what people actually know about is the communities and groups they join, the information they share and the things they contribute to. Our profiles are starting to make use of the “social graph” i.e. what you work on, what you share and with whom you exchange information. We’ve incorporated some initial inputs on this such as highlighting documents people have worked on, shared documents, showing people’s activity feed and identifying who they work closely with through this information – but in the next few months we plan to do more, building on social graph tools that Microsoft have developed with Office 365. At this point I feel the potential for using this type of information to help recommend people and content based relevant to the person doing the search is very large – and also needs some experimentation to see how to get the best out of this capability.
A few lessons learned from this exercise:
1. Getting a common agreement about what should be included in a staff profile is challenging. Given the absence of good existing system the tendency is to want any new system to do everything and there is pressure to add a long list of potential attributes. Also some of the things that users want that could be extremely useful are just not supported by the underlying information that we have about staff. Having self completion fields can help with this, but if the profile has too many self-complete attributes then the chances are most of it will not get completed. Also there are some attributes that individual functions and offices would like that are not relevant to others so coming up a with a common list of useful core attributes is key.
2. There are a lot of data quality problems in our official systems of record about our staff that go unnoticed and therefore unaddressed (I expected this but perhaps not as many as we are currently finding). People care about how they are presented when they realize this is in a profile/search system so when they see this information is incorrect they want to change it – and I’ve had a deluge of e-mails requesting corrections. While frustrating this is important since it is providing an incentive for staff to care about what the organization knows about them and to get this fixed – and if we are lucky it will help put the spotlight on some of the challenges with our underlying staff information systems and create momentum to fix them.
3. Working with cloud service software like Office 365 has some great strengths, such as the ability to be able to search across content from multiple sources and to take advantage of the very promising work that has been done on using the “social graph”. It’s also challenging since the platform and interface is being frequently changed and updated and so there is a risk that any customizations are broken or obsolete without notice. In fact Microsoft are currently moving away from SharePoint profiles (on which our profile system is based) to Delve profiles, and we will soon have to recreate our profile system on Delve. However Delve profiles don’t yet allow the type of customization we have done in terms of grouping and ordering fields – but well we’re hoping that will change soon so we can migrate.
4. People don’t yet understand the “social graph” and can even be a bit scared of it. “How does the computer know what I want to see”, “Why can I see this document in my profile that should be confidential”. The social graph can at times be uncanny in knowing what you might be interested in. It may therefore show different results depending on who is looking at a profile. This needs careful explanation –and also reassurance that it’s only showing people what information they are authorized to view. It’s also a good reminder to check on the sharing/security setting on your documents – in particular a number of people put private documents in their ‘shared with everyone folder” probably assuming that if you didn’t know something was there you would never find it. The social graph it’s much easier to find things – which seems to actually a cause for concern for some :-0
5. The social graph only has what you put in it. We’re still only part way in migrating the intranet to SharePoint online and adoption of sharing tools like Yammer and One Drive is still fairly low. This means a fair amount of useful information is not being captured. The profiles have highlighted that a lot of staff are still not sharing much of what they do in any place where it would be easily accessible by others. We still have a lot of work to do to promote a culture of “working out loud”. Only when we have a higher level of participation in social collaboration tools and adoption of cloud storage (rather than people using the shared drive or worse their e-mail archives and hard drives as the main store for their important organizational work) will we really be able to get the benefits of the social graph.
I’m sure we will have a few more key lessons in the coming weeks and months about what works and what doesn’t and the key challenge of getting people to update and use the profiles. Will keep you posted!