Like any city, Glendale, Arizona routinely procures for the maintenance and expansion of the city’s public infrastructure, particularly as part of the implementation of its Capital Improvement Plan. Each of the city’s contracts includes data points like when the contract was awarded, what services were included, how much those services would cost, and more. The city passed an open data policy in early 2017 and in the months afterward leaders in Glendale wondered: could the city get better bids on contract opportunities by making this data open?

To answer this question Glendale’s open contracting became the second pilot project of the Sunlight Foundation’s Open Cities’ Tactical Data Engagement work. The project combined the the Sunlight Open Cities team expertise on open approaches to government and government data with the city procurement expertise of Harvard Kennedy School’s Government Performance Lab (GPL), our partner organization on What Works Cities. It’s the first time our organizations have worked together so directly to support procurement improvements by making data more open and available.

We began the project by surveying staff at companies that had bid on or been awarded contracts with the city over the past several years. Sunlight and GPL reached out to 68 vendors and interviewed 16 to understand what information companies currently look for, and what information they would be interested in using if it was available from the city. The interviews yielded clear use cases centered around the information needs of different kinds of vendors during each stage of the procurement process: planning and pre-solicitation, solicitation, and contract and implementation.

Then we tested. Working with Glendale, we created a sample dataset of upcoming capital improvement procurement opportunities and asked vendors to tell us what was helpful, what they understood, what they didn’t understand, and how we could make it better. That testing phase ultimately resulted in our final recommendations for the data elements and data formats most helpful to vendors.

Following our experience working with Glendale on this project, we can offer the following recommendations to cities that are aiming to use open contracting data to improve procurement processes.

RECOMMENDATIONS

1. Cities should publish datasets of upcoming procurement opportunities. These datasets should focus on upcoming projects over the next 1-3 years and include data points such as project name, project description, expected timing, and estimated budget. The dataset should classify projects by industry and/or commodity code so that vendors can easily find projects relevant to their businesses. The dataset should also include information about the department leading the solicitation and contact details. When applicable, the dataset should also include include previous award information and a link to the past contract. In addition to releasing the dataset, cities should include a short summary describing the contents of the dataset, a data dictionary, and descriptive column labels to facilitate understanding. Although we specifically tested a dataset focused on capital improvement projects in our work with Glendale, feedback from the vendor community suggests releasing upcoming procurement opportunities is attractive to vendors from a variety of industries.
2. Similarly, cities should actively publish past contracts or information from past contracts. From the vendors surveyed in our project with Glendale, this was the most-commonly requested type of data. Vendors indicated that being able to see past contracts, even if they were selectively redacted, would save them time and provide relevant information for planning purposes.
3. Cities should carefully consider publishing contract performance data, i.e., audit reports that evaluate the performance of past contracts. Performance data was the second-most requested data type by vendors surveyed as part of this project. However, some vendors indicated that past performance data would not contribute to improving their own proposals. Moreover, some vendors expressed hesitancy about proactive releases of performance data, with some indicating their preference that performance information should be available only through public information requests.
4. Cities should use online vendor portals that are searchable by industry or service to make it easier for vendors to identify procurement opportunities that are relevant to their businesses. Such portals, along with the city’s website, are where many vendors indicated that they found bid opportunities. Also, email notification systems that notify vendors when new procurement opportunities are posted should include specialized notifications based on industry or service.
5. To lower barriers to submissions, cities should consider using a proposal submission template, where feasible.To accommodate, instances where a template would be detract from creative formatting that directly reflects the quality of the proposal/vendor, cities should offer flexibility in using the submission template.

To learn more about the open contracting data project in Glendale, Arizona, see the full Tactical Data Engagement case study.

Understanding how open data can improve contract bids is one way that cities can apply Tactical Data Engagement, our approach for putting open city data to use on cities’ toughest challenges. Our other TDE pilots have included putting data to use improving development services also in Glendale and creating more equitable and complete neighborhoods in Madison, Wisconsin. In March 2018, Sunlight’s Open Cities team began the next two pilots of TDE in Norfolk, Virginia and Austin, Texas. Sign up for the Open Cities mailing list to learn more about those projects as they progress.