Driving Open Government through Citizen Experience.

My group at Susan Turnbull’s Collaborative Expedition Workshop at the National Science Foundation this week focused on determining how crowdsourcing could improve government services as part of the Open Government changes which agencies might adopt.  First several concerns on using crowdsourcing were voiced:

• How do you prevent the crowdsourced solution from being manipulated or hijacked?
• Just because citizens prefer something, does that mean the government should be obligated to deliver it?
• How do you ensure that the audience involved in crowdsourcing knows something of the subject?
• How can we know which issues to prioritize for crowdsourcing efforts?
• What is the mechanism by which we select which issues to focus on?

The upcoming citizen participation efforts as part of the Open Government Directive were discussed as well as  the concern that those efforts may leave out some of the more bread and butter frustrations with more everyday citizen-government contacts.  Also there was a concern that those efforts might draw from “expert” groups more so than those who receive services or need to contact government agencies.  Open Government should not be restricted to people who comment on blogs to participate in solutions voting, it should also include the 100-200 million monthly citizen experiences when contacting the government.  In short, the citizen experience of government contact and government services needs to be brought into the discussion for Open Government to lead to more effective government.

Why include citizen contact data into Open Government Processes?

We discussed creating a collection of incoming data from citizen contact centers about what citizens were calling about in detail along with the more typical stats of resolution, call length, call backs, repeat calls.   Why is this important?  A key reason for the entire Open Government movement is to both make a more effective government from the citizen’s perspective and to increase trust in government by the citizen.  At a recent ACSI ( the most frequently used customer satisfaction index for federal agencies) event, it was pointed out that one of the best ways to increase trust and create positive word of mouth from citizen’s  is to prevent frustration (multiple calls) and resolve citizen’s complaints or concerns quickly.   So we proposed a mechanism to continually identify and prioritize citizen concerns and crowdsource better solutions to reduce the number of complaints and increase resolution of issues expressed by citizens at contact centers which includes safeguards to avoid spending government resources on non-starter issues (such as things which the government can’t do by law).

Aggregated Contact center data as a reliable source of citizen frustration in Open Government.

It’s a basic tenant of modern web design, systems thinking, and social media strategy to understand and be in touch with the needs and perspective of your audience.  If then we took depersonified but still context and content rich information from contacts centers and embedded some semantic intelligence into it (such as standardized coding for certain types of problems) from each contact center (which there are about 2000 for the federal government alone) and then aggregate them  into a federal wide feed..lets call it Citizen Trends. We could then we could have a better understanding government-wide of where the frustrations are with citizens in their dealing with the federal government from the citizen perspective.   (Incidentally I recently heard of a project called GovPulse which is in the planning stages and may address some of this in some cases. Yeah!)  Just aggregating citizen contact data alone (in ways which protects privacy),  has the potential to improve performance metrics of citizen contact.  But this data feed could also be made into a framework for crowdsourcing better ways to serve the public.

(Quick definitions…a  contact center usually refers to a point of contact an agency or department receives inquires from citizens through the phone, email, fax, automated voice service, or web submission forms.  Social media groups are usually not included as citizen contact centers, but I think they should be. Contact centers often have specific tasks and a specific set of rules they operate under. There are between 100-200 million citizen contact events monthly with the federal government.)

Using  Citizen Trends as a Framework for Crowdsourcing  Improvements to Government Services as part of an Open Government initiative:

So if we have aggregated data about what frustrates or bothers citizens who actively seek resolutions or services from the government,  then we can determine what the most frequent issues are which do not seem to be resolved right away.  The systemsthinking people such as John Seddon might call this dividing the data into demand value (contacts which provide value to the citizen in the most direct fashion) and demand fail (contacts which occur because some other part of the system failed and the citizen needed to find a workaround or correction).

OK fine. we got some more data to meet transparency goals… how does that become actionable?

This only goes beyond a transparency exercise, if there is a requirement for agencies to address the issues which bubble up to the top based on frequency and severity.  Now does this mean that any frequent or very severe issues will be solved or put into an endless cycle costing valuable time and resources?  No.  Not if its done right.  First agencies only need to be required to address types of or specific issues which recurr frequently or have severe impacts. They can group different issues into a category if they all have a similar answer as to why the negative citizen experiences exist.  If there is a clear budget restriction (don’t have the funds to do it that way) or a legal/policy restriction (its against the law) then the agency or department would simply post the reason why this issue exists.  In doing so, it would create a resource for congress and advocacy groups to understand when there clearly is a budget or legal issues which may be causing unplanned problems when providing citizen services.

OK better transparency for why the government does stuff, but how is this become a crowdsourcing framework?

This CitizenTrends centralized data becomes a crowdsourcing framework when we take the issues which seem to have budgetary limitations, have legal limitations or have no really good reason for existing and allow potential solutions and discussions for solutions to be posted online.  It may be that the 3 or 4 well meaning people who decided a solution would cost too much missed a cheaper way of doing it.   It may be that education or posting of information needs to occur and be more visible to reduce the recurrent frustration about a mistunderstood policy or regulation.  It may be that certain exceptions are warrented in a law of regulation to avoid unintended outcomes.  It may be that something was missed in the contact center knowledge base, training or simply an option to the IVR needs to be added.

To Sum up.. A Crowdsourcing Framework could be created to make both more effective and more Open Government…

Our group at the collaborative expedition workshop proposed to map citizen experience to process to policy to  law by allowing aggregate citizen information to drive the priorities on what problems in government solutions should be sought for but at the same time recognizing which issues are unavoidable because of law, policy or regulation  or seem unavoidable due to cost to rectify.  This information and any solutions suggested would be made publicly available so that citizens, advocacy groups and lawmakers could see what day to day issues were created when legislation was implemented.  To implement this, we would encode depersonalized information from the 2000 or so citizen contact centers or points of contact into a central feed whose data is coded to make it queryable in an intelligent fashion.  Agencies would be required  to address most frequent “demand fail” or frustrations of citizens by citing the reasons that the process causing the complaints exists the day it does by relating it to expert evaluations, budget constraints, policy or legal things.  They would also have to address whether the issue seems to point out a need for better contact center training or updating the knowledge base.   Unresolved issues would be allowed to discussed and potential solutions submitted. This would serve as framework for crowdsourcing solutions improvements to citizen services through a unique combination of  aggregated citizen experience, data, expert input from agencies,  and constraint information regarding legal and policy constraints.  This central repository could also serve to drive updates to call center training and knowledge bases.