How open data could help the City of Edmonton save at least $197,500 on 311-related expenses

While researching 311 for my previous post, I consulted the 2009-11 Capital Budget (PDF). The Capital Budget pays for both maintaining existing infrastructure and undertaking new projects (in contrast, the Operating Budget is a one-year budget for services and programs – you can find both here). In it, I discovered a Corporate Services (read: IT) project called “e-Business”, described as follows (on page 382 if you want to look it up):

The purpose of this funding is to put 311 statistics on our websites. First for internal staff then to public. The objective is to present 311 statistics in a manner similar to the what New York City does. Their 311 activity is presented by neighborhood or ward, in text form in a table as well as on a map. This project is being presented as a phase in of sophistication. This enhancement can be implemented based on what can be funded.

I went to check out New York’s site for 311 statistics, and it is indeed quite useful. If you type in an address and select a borough, you get a wide assortment of statistics, presented in tabular form or on a map. The bad news is that most of the tables and maps are in PDF form. The good news is that you can export all the data to CSV, which means you can map it yourself! Very cool.

The project description continues:

The e-Business program is driven by three primary factors. Our citizens are demonstrating increased use of the Internet and the City’s Web site, our population is growing, resulting in a projected need for increased services, and citizens and other stakeholders are demanding new and more extensive self-service access to government.

If you’re not thinking of open data and ChangeCamp by now, you should be! Here are some justifications for the project as outlined in the Capital Budget:

  • Program Managers would use this as an additional tool for monitoring service activity.
  • Residents and potential residents could find out what the issues are in specific neighbourhoods.
  • The Mayor and City Council could quickly determine what the major issues are, without having to make a request.
  • The project will help make the City more transparent to its public.

Sounds great to me!

So what’s the problem? Cost.

The justification section of the project description says: “At the City of Edmonton, e-Business is about business first – providing the services our citizens need and want at a justifiable cost.”

I wouldn’t call these costs justified, I’d call them outrageous:

Phase 1 Table display of 311 Statistics reports $27,000
Phase 2 Option to display of reported statistics by business area $27,000
Phase 3 Option to display of reported statistics by neighborhood $37,000
Phase 4 Option to display reported statistics on a map $133,500
Phase 5 Include statistics that are stored in other applications $974,000
  TOTAL: $1,198,500

This is why we need a policy on open data, so that projects like this one don’t get funded, wasting taxpayer dollars.

The first item consists of automating the conversion of reports generated by the 311 system to web content, presumably HTML. I think $27,000 sounds justified for that task. It’s potentially quite difficult and error-prone, depending on what the reports look like, what format they are in, etc. The rest of the items are ridiculously overpriced, however.

What the City should do instead

Spend the $27,000 to automate the conversion of the reports to CSV format. Then make those CSV files available for free to the public. I promise you the City would get the next three phases for free. If I had the data I’d gladly do it, and I’m sure there are others who would too. On top of that, the community is more likely to use standard/open tools and technologies (such as Google Maps) rather than proprietary, awkward ones (such as the City’s SVG maps). We’d probably get it done faster too. The City can then put money to more useful things, such as an open 311 API similar to what Washington, D.C. has.

I think Phase 5, which would include data from POSSE, CLASS, and other internal systems, is too big and broad currently (hence the very large cost). It could be approached in the same way though – spend a little bit of money to make the data available in an open format, and give it to the community to do the rest.

I should point out that the above total ($1,198,500) has no approved funding. Rather, it is the amount the project would cost if City Council approved the entire thing. They could choose to approve only one phase, a couple phases, or none. The reason I put $197,500 in the title is that only the first four phases seem reasonable to me given the information available, and I think the first one is a justified cost as-is.

If the City of Edmonton could save nearly $200,000 on this one project alone by embracing open data (not to mention the indirect benefits that will come along with that cost-savings), imagine what the benefits of embracing open data across the board would be!

Edmonton’s 311: six month status report

As you’ve probably heard by now, the City’s 311 service is not performing as well as expected. The Sun first reported the story last week, and the Journal followed up with an article yesterday. They key point mentioned in both is that wait times to connect to an operator are far longer than originally anticipated. As a result, an interactive voice response system is being considered for next year. I can’t imagine that will make callers any happier, even if it does make their calls slightly faster.

I personally think they should put more resources into 311 online. How many citizens even realize that they get online access to a lot of the information and services that 311 provides? The best way to reduce call times is to increase self-service options and quality so that citizens can bypass the phone altogether.

Fortunately, there are new self-service features being developed for release in October. The City will still need to communicate their existence effectively, however.

Background

The CRTC approved the use of the 311 phone number for municipal services back in November 2004, and Calgary became the first city to launch 311 on May 8, 2005. Here in Edmonton, City Council approved the service at its May 9, 2006 meeting. Edmonton became the first city in North America to use SAP’s CRM application to deliver 311.

The 311 service officially launched on December 16, 2008. Implementation was approved at a cost not to exceed $10 million (and it is on track to come in about $1.5 million under budget). Half of that amount came from an internal loan, which is to be repaid from operating savings (the other half came from a special dividend in 2005).

Six Months In

The report that went to the Executive Committee this week isn’t incredibly long at 7 pages, but it does have lots of information. Here are some graphics to help make it easier to understand the first six months of 311 operation in Edmonton.

Call lengths are one of the reasons everyone is complaining:

Wait times to get through to an operator are another concern:

The 311 system was supposed to help the City capture the estimated 160,000 missed calls each year, but so far it is on track to make things much worse:

The report contains information about the top ten services:

Transit inquiries make up a significant portion of all 311 calls, followed by Community Services inquiries. There’s a clear opportunity for transit to do more to reduce the number of calls going through 311. I find it odd that trip planning is such a common request actually, given that there’s a separate number for that (BusLink) not to mention the online trip planner and Google Maps.

Here’s the breakdown by department:

There are a few more graphs (without data values unfortunately) in the report, so take a look at those too. They show that the number of calls answered within 25 seconds is on the rise, and that the time it takes to get through to someone is declining.

Does this report suggest that 311 is “a disaster”? I don’t think so. All it shows is that there is work to do, and it sounds like the 311 team is on the case. Hopefully the departments they serve are as well.