Team Pittsburgh Midyear Report

The Issues

Keeping track of all the different procurement procedures is anxiety-inducing for city staff.

For city staff who need to buy something not already on contract, navigating the ways the City adopts new contracts or buys off-contract is not an easy task. There are currently ten different procedures the city uses and determining which one is right for a particular good or service can be stressful. Is what you are buying more than $2,000 but less than $30,000? Is it an emergency purchase? Will you be making multiple purchases over time? Is it a service? If you are an office manager for a City department, obtaining a steady supply of paper requires navigating resolutions and memos just to determine what procedure you should follow. Suddenly, buying a ream of paper becomes a burdensome task.

Lack of well-documented, easy to find instructions often leads to wasted effort on the wrong process and extended back-and-forth.

Current procedures are also difficult to find. Staff typically ask around until they have found someone who contracted a similar good or service. Drawing from informal institutional knowledge can be a great thing, but sometimes leads city staff to pursue the wrong procedure.

Learnings from User Research

My biggest gripe is nothing is written down in terms of process.

We hear from a lot of city staff who get in trouble trying to get what they need. In one conversation we learned about a time when the City’s Public Art Manager needed to hire contractors to help maintain art installations. He set about contacting 8+ firms by phone and email, tracking proposals, and documenting his process with a summary sheet for each response. When he sent his documentation to the Office of Management and Budget (OMB) for approval, he was disappointed when his award selection was initially denied because he had not followed the correct process. While he was supposed to send a list of specifications to their office and let OMB do the vendor selection for him, he instead did all the legwork himself only to have his selection rejected. Further, the procedure he was supposed to follow left him out of the selection process so he could not contribute his expertise on the subject. As he put it, "how the hell do they know what special things I need? How can they be the catch-all experts in all fields?"

Writing the software RFP for 311 without any guidance or templates to draw from was a long, anxiety-ridden process.

Many in the City feel like they are left in the dark by procurement. Unknowns in the procurement process cause anxiety, as one project manager expressed when telling us their experience looking for software development help for the Pittsburgh 311 application. To put out a Request for Proposals (RFP), select a vendor, and move to a finalized contract, they needed to obtain many documents (project budget, Scope of Work, confirmation of competitive bid process, etc.). Whenever they were missing one component the entire process had to be put on hold. The extended back and forth led to frustration and significant delays, which more than just a major time sink can be taxing emotionally.

If City staff worked with us like we were a private law firm, things would go much more smoothly. Would you show up to your lawyer’s office, charging you by the hour, with nothing?

The law department also feels frustration surrounding documentation, albeit on the other side of the process. During the contracting phase, law is tasked with crafting an agreement that the City can use to make sure work gets done, but departments often show up unprepared, without many of the documents necessary to write a contract. Without proper documents or having done something incorrectly, law often cannot construct a strong and enforceable contract.

Metrics & Goals

We want to reduce anxiety and staff time creating contracting and bid documents. Confusion surrounding procurement procedures not only wastes a lot of time, but can strain professional relationships in the City.

To measure our impact, we will look to decrease the amount of time it takes to move a new contract from selecting a vendor to final contract. The contracting phase is where procedural mistakes hold things up and the time spent here is a good measure of the efficacy of the process as a whole.

Prototyping

We wanted to know where we could help surrounding documentation. Do city staff need help with templates, information about the larger process, or a combination of both?

We built a click-through site of City of Pittsburgh procurement procedures during our residency month in February.

While we originally used this as a way to onboard ourselves, we realized this could be a valuable tool for the larger City organization. City staff could come to this site to help them find the correct procedure for what they were looking to purchase. It also enabled us to link to other products we were developing, such as Scout.

Feedback on the tool was positive. Departments wanted more content, but overall they liked the interface guiding them through the different ways to procure goods and services. The City law department was especially supportive, seeing the tool as a way to clear up confusion.

Another idea we had was surrounding the act of document creation itself. What if the law department and OMB could create and adapt document templates that then became easy-to-fill-out forms? We built a prototype of this tool, which we called Template Maker.

With Template Maker, our city partners in the law department could create a central database of internal procurement documents. Standardized document templates would become forms, which could remove a lot of the confusion departments face while drafting Requests for Proposals, as well as other bid documents, resulting in better documents and less back and forth during the final contracting process.

Overall response to Template Maker was positive. The idea of a making templates that then made forms was a bit abstract, but the concept typically settled in during testing. OMB as well as the law department agreed that templates were central to a good procurement experience, which allowed for fruitful conversations.

Pivoting

After returning from our first trip back to Pittsburgh in April, we decided to stop actively developing Template Maker. After many long discussions, we came to the conclusion that the amount of work required to make Template Maker a successful and stable product was too high.

There were two main challenges that fueled our decision to move away from continuing development on Template Maker. The first was that Template Maker would need to be as easy or easier than using a Microsoft Word or Google Docs template. This would be a challenge from both an engineering and a user experience perspective, especially given the limited scope of time in the fellowship. Secondly, Template Maker also only addresses creating bid documents, which is just one part of the lengthy and involved procurement process. Our user research indicated to us that confusing existed throughout this process, not just in the document creation. By moving our focus off of documents, we can focus on the entire process.

Product

Focusing on the larger process, we are building out our click-through tool as the place to go when you want to adopt a contract or buy something off-contract in the City of Pittsburgh. We are calling the tool Atlas, as it is at once a guide through procurement procedures as well as a central repository for the processes that support procurement as a whole.

Core feature list:

• An easy-to-navigate click-through interface
• Final landing page at the end of each flow detailing the appropriate contracting or purchasing procedure
• A comprehensive checklist of required documents
• A document handoff mechanism allowing departments to initiate contracts with the law department

Next Steps

With the basic form of Atlas already in place, the major focus of our next efforts are in maintaining and creating content. In the months ahead we aim to:

• Pilot with 2-3 users in two City departments
• Work with law to create document checklists for all ten procurement processes
• Continue to fine-tune the mechanism used by departments to submit documents to law

The Issues

It is hard for city staff to find information about existing contracts.

Inventory management is a problem across the city. Contracts themselves are locked up on shared drives, and information about those contracts beyond the bare minimum is even more cumbersome to find and use. Various workarounds are in use in City Hall, from printed spreadsheets to thick paper packets. Knowing what is available is difficult and time-consuming, and leads to purchases of similar goods from more expensive retailers such as Staples.

The contract renewal process is lengthy, difficult, and non-transparent.

The responsibility for managing 300+ contracts through a 13-stage, five department renewal process falls on two purchasing agents. This renewal can take four months or longer to complete. There are many opportunities for a contract to be held up along the way, from the City’s Equal Opportunity Review Commission to Allegheny County’s Purchasing Department. Using departments who need information about these contracts often resort to calling the purchasing agents directly to ask about status.

Learnings from User Research

We have binders full of contracts.

One theme that we heard repeatedly throughout our research was that no there was no systematic way of looking up contract availability. Shadow systems, often kept on paper in binders, exist in every department. While some information about available contracts exists on a shared drive, it is incomplete information. Changes often do not propagate through these shadow systems, leading to problems down the line when contracts expire or receive new information from the City Controller’s office. This can lead to situations where purchasers go to order a product only to find that the contract is no longer valid, or payment is rejected by the City Controller’s office due to incorrect or incomplete information.

We spend thousands of dollars in staff time research the right forty dollar part.

A second theme that we heard from several using departments is that the current purchasing model discounts the value of staff time. The fire department in particular noted that they spent thousands of dollars doing research to buy $30 or $40 parts. While this is not necessarily considered when thinking about costs of procurement, it was a striking story to us. If we could successfully reduce the time it takes to do this research, we would be freeing up staff to use their time much more effectively. Particularly onerous is the process to look up information about contracts that are part of the State’s COSTARS joint purchasing system. The system allows vendors to register with the state and then sell directly to the City’s departments. City staff like the program because COSTARS-approved vendors are considered to have successfully completed the competitive bidding process. Actually taking advantage of it is another story altogether, however. Speaking about the COSTARS program, one City staffer said that “it’s the last thing I want to do. Sometimes it takes me weeks to get to vendor invoices.” Much staff time is spent navigating these difficult-to-use web resources that surround programs that people are otherwise excited about.

What I need is a multifaceted alarm clock, so nothing falls through the cracks.

It is easy for contracts to slip through the cracks during the renewal process, which can lead to major complications down the road. One story we heard was that the fire department was not contacted about renewing a contract they used to purchase gloves. As a result, the opportunity that was sent to the business community had the wrong type of gloves. The fire department is now forced to buy gloves separately as a result. The work puts a lot of strain on the two purchasing agents: it is difficult to keep track of which departments need to be reminded about what at different points along the process.

Metrics & Goals

Our primary goal in this area are twofold. We want to reduce the time it takes to look up information about a contract. We also want to help purchasing agents manage and update contract information.

Prototyping

Our prototyping around the first area began during build week in January. The problem of finding information about contracts came up in our very first conversation. We took the information from the original shared drive spreadsheet and added basic search functionality.

We put this in front of users as early as February, and received all sorts of valuable feedback. The first overall piece of feedback was excitement — people were happy to be able to more easily search the information on the shared drive. The second piece of feedback was that people wanted to be able to extend this functionality to more data. They wanted information about the previously mentioned COSTARS contracts, line item level information, and the ability to subscribe to contract updates.

We have also been prototyping different approaches to the problem of contract management. We have spent a lot of time talking to the purchasing agents and understanding how they use their spreadsheets to manage contracts right now.

One thing that has become clear through user testing is that the purchasing agents use their spreadsheets in very different ways, likely because a spreadsheet is not necessarily the best tool for the job. We have not landed on a perfect solution to the problem yet, but our initial prototypes have been met with good levels of enthusiasm.

Products

We took user feedback from our time back in Pittsburgh and created a new version of the search tool. We have renamed it Scout, because its role is to help you search and explore the contract landscape.

Core feature list:

• Light search-based interface into contracts, companies, line items, financial (City Controller) identifiers, and more
• Star contracts to find them more easily
• Follow contracts and get email notifications about updates
• View contracts others in various departments have starred/followed

Scout now incorporates all of the features that users asked about. However, it is still entirely ready for full adoption. We are still working on improving the search experience, and allowing the system to intake more data about different types of contracts. We are also going to experiment with ways to allow users to take notes on contracts and then index and search those notes.

We are starting to work on building out a tool we are calling Conductor. Conductor will play a central role in the procurement suite that we are putting together.

Core feature list:

• Manage soon-to-expire contracts by keeping track of where they are in the process
• Send email reminders and log actions taken on each step from inside the app
• Directly publish information about the contract to Beacon
• Update contract metadata

Not only will conductor serve as the central contract management hub, it will also connect with other tools, notably Beacon. We want to empower the purchasing agents to easily share upcoming opportunities with the business community. We want conductor to be a central location of information, updated by the purchasing agents and enabling other civil servants in Pittsburgh in their goals around purchasing.

Next Steps

Our next steps are mainly in product development and user testing:

• Continue to iterate and improve Scout’s search functionality
• Implement sustainability plan for incorporating new contract information
• Finish and test web prototypes for Conductor