Open Data: Challenges and Opportunities for Transit Agencies (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

49 The techniques used by the respondents to engage existing and potential data users and reusers are shown in Table 20. The most frequently used techniques are face-to-face events, followed by conferences and meetups. OPPORTUNITIES AND IMPACTS The impacts of providing open data were explored through the survey. This open-ended question yielded a variety of responses, generally in the categories of positive and nega- tive impacts, as follows: • Positive Impacts – Increased transparency (most frequently mentioned impact–mentioned by five respondents) – Labor reallocation – Increased return-on-investment (ROI) from existing web services – Improved market reach and time to market – Better and more accurate internal data – Better and more accessible transit information for tran- sit users, empowering customers in terms of choice and competition, and improving public rider percep- tion and awareness – Better visibility of provided public services – Easier use of transit system – Fosters innovation – Improved relationships/coordination with MPO, DOT, research institutions, and neighboring agencies – Reduction in the requests for data from other govern- ment and private (consultant) users – Allows the agency to realign its information services and delivery in a more cost-effective and customer- focused manner • Negative Impacts – Development effort and maintenance, and generally, staff time – Data quality n Time required to do quality analysis on data, result- ing in more pressure on staff n Increased awareness of data quality issues, requir- ing more resources to devote to data consistency – Requires cross-checks using disparate information sources – Obligated to provide up-to-date data – Additional workload when dealing with developers who may not fully understand transit data COSTS Survey respondents reported several types of costs associated with providing open data, as shown in Table 16. Just more than three-quarters of the respondents reported that staff time is required to update, fix, and maintain the data, and almost 70% reported that internal staff time is required to convert the data to an open format. The third most frequently reported type of cost is the staff time needed to validate and monitor the data for accuracy. Almost 90% (43 or 89.4%) of survey respondents indicated that they cannot quantify how much time is spent on any of these activities. This finding is consistent with the author’s experience in exploring the actual costs of providing open data—because much of the costs are internal, many agencies do not track staff costs, so they may not be aware of the exact cost of the activities associated with opening their data. Fur- ther, there are support costs associated with software that gen- erates open data. It is rare that agencies account for these costs when assessing the costs of open data. The amount of time being spent on these activities by the several agencies that do know their costs varies widely, as shown in Table 17. About 95% (46) of the respondents cannot identify the actual costs associated with any of these activities. Only one agency reported that the monthly cost associated with the web service for hosting data is $1,500. BENEFITS The survey asked about the benefits that the agency experi- enced as a result of providing open data. More than 75% of the respondents reported that it increased the awareness of their services. About 75% reported that it empowered their cus- tomers and encouraged innovation outside of the agency. The other benefits reported by respondents are shown in Table 18. DATA USER AND REUSER ENGAGEMENT More than two-thirds (33 or 69.6%) of the respondents stated that they engage or have a dialogue with existing and poten- tial data users and reusers. Table 19 shows the reasons for the engagement. chapter six SURVEY RESULTS: COSTS, BENEFITS, CHALLENGES, AND OPPORTUNITIES

50 Types of Costs Associated with Providing Open Data Number of Respondents Percent Staff time to update, fix, and maintain data as needed 38 76.0 Internal staff time to convert data to an open format 35 70.0 Staff time needed to validate and monitor the data for accuracy 28 56.0 Staff time to liaise with data users/developers 25 50.0 Web service for hosting data 23 46.0 Publicity/marketing 12 24.0 Consultant time to convert data to an open format 11 22.0 Other: Contract management Cost to develop prediction software or use prediction Software as a Service (SaaS) Everything above is already done for internal purposes and it is all automated Investigation project agreement with the Faculty of Computing Sciences Consultant time to build editing tool License Routing service Mentz No additional costs are incurred. From survey responses. TABLE 16 TYPES OF COSTS ASSOCIATED WITH OPEN DATA Activity Number of Respondents Range of Labor Hours per Month Internal staff time to convert data to an open format 4 3–40 Staff time needed to validate and monitor the data for accuracy 4 1–10 Staff time to update, fix and maintain data as needed 3 2–20 Publicity/marketing 3 0.1–2 Staff time to liaise with data users/developers 2 0.25–6 Consultant time to convert data to an open format 2 20 Web service for hosting data 1 1 From survey responses. TABLE 17 LABOR HOURS PER OPEN DATA ACTIVITY Benefits Number of Respondents Percent Increased awareness of our services 39 78.0 Empowered our customers 37 74.0 Encouraged innovation outside of the agency 37 74.0 Improved the perception of our agency (e.g., openness/transparency) 33 66.0 Provided opportunities for private businesses 24 48.0 Encouraged innovation internally 21 42.0 Improved our market reach 18 36.0 Become more efficient and effective as an agency 11 22.0 Increased our return-on-investment from existing web services 10 20.0 Experienced cost savings 5 10.0 Been able to reassign staff 3 6.0 From survey responses. TABLE 18 BENEFITS OF OPEN DATA

51 – More scrutiny because of increased visibility of data accuracy, including third-party users wanting zero downtime • Neutral Impacts – Thinking about data reuse versus public policies – Public awareness of what agencies are doing and how they are doing it. The impacts on the public sector (e.g., riders, community citizens) of providing open data also were explored. The fol- lowing were reported: • Creating and improving access to additional and higher quality public services, including more and free applications • Providing better, more accessible, and more timely public information and tools • Resulting in improved and sustainable mobility • Improving transparency and accountability • Improving customer service, customer satisfaction, and public perception/image of transit, including service reliability • Empowering the public • Making transit more competitive (reducing “costs” of trip related to customer uncertainty) and easier to use • Providing more visual information • Providing more innovative applications that government agencies may not be able to provide • Providing a better transit experience • Increasing ridership • Increasing competition for transit riders • Providing better regional coordination • Encouraging the development of third-party tools and applications Reasons for Engagement Number of Respondents Percent Obtain feedback on data anomalies and data quality issues 25 75.8 Find out more about how people want to use/reuse your data 21 63.6 Expose your data to a wider audience 21 63.6 Provide technical support 20 60.6 Announce updates, modifications, etc. 19 57.6 Find out more about the demand for our data 18 54.5 Suggesting features to improve the functionality of applications 17 51.5 Find out more about prospective users/reusers 15 45.5 Enable existing and prospective users/reusers to find out more about your data 14 42.4 Explain transit jargon and definitions 12 36.4 Solicit requests for future data 11 33.3 Enable prospective and existing users to meet each other 7 21.2 From survey responses. TABLE 19 REASONS FOR ENGAGING DATA USERS AND REUSERS Engagement Techniques Number ofRespondents Percent Face-to-face events 24 60.0 Conferences 18 45.0 Meetups 12 30.0 Hackathons 10 25.0 Application competitions 7 17.5 Unconferences/BarCamps (conferences with no set agenda—the agenda is set at the time of the conference by the participants) 5 10.3 Speed Geek events (participation process used to quickly view a number of presentations within a fixed period of time) 2 5.1 Other: Local Open Data Advocacy Group Email Various online industry and developer forums From survey responses. TABLE 20 DATA USER AND REUSER ENGAGEMENT TECHNIQUES

52 The impacts on the private sector (e.g., developers) were reported by survey respondents as follows: • Providing business/commercial and development oppor- tunities, including new and expanded companies that could create a new eco-system of private entrepreneurs • Enabling innovation and the creation of applications • Providing data to cover new needs • Decreasing the need for agency to develop apps on a multitude of differing platforms, which would be costly to do internally or to outsource • Providing more visual information • Providing a broader reach for customers • Adding value to existing services • Private sector interacting with transit more comfortably because they know more about transit • Adding data by large trip planning services (Google, Bing, HopStop) • Improving access with high potential for untapped growth areas that the public sector cannot fund or have access to • Creating interest in agency and desire to show off cod- ing ability Many of the impacts to the agencies, public sector, and private sector are repeated, proving that opening transit data has a significant value for all three groups. CHALLENGES The survey asked about the challenges associated with pro- viding open data and how the challenges were overcome. The survey responses include the following: • Resources and organizational issues – Limited dedicated resources (both time and staff) responsible for managing open data (including data conversion/cleaning and validation) – The process/philosophy is still not fully understood – Securing management support – Agency coordination challenges – The lack of technical know-how within the agency – Challenging internal parties who believe that we should be charging for the release of data – Helping internal groups see the benefit/value of par- ticipation and demonstrating how this can reduce manual publication load – Closer attention to change management – Internal fear that we should not do it because not all predictions would be accurate and we would be criti- cized for that • Data quality and timeliness – Ensure/preserve data quality, completeness and equity, and timely release of data – Necessary to clean the data – Interaction with regional systems and inconsisten- cies with what data are used in which of their systems – Improve/preserve rider perception of accuracy of arrival predictions, given operational impacts such as adverse weather, reroutes, construction – Ensure safety/security of files and information disseminated • Standards and formatting – Standards help overcome formatting issues – Better organization of the marketing of available information to public – Managing the evolution of internal data model – Data scalability • Marketing – Making the data uniform resource locators (URLs) known – Partnering with an organization (e.g., Mobility Lab) to publicize availability – Initially, resistance because of branding issues • Technical issues – Tracking users and developers – Process of making the data available when new sched- ules are released – Finding ways to represent some of the unique aspects of rural transit (such as deviations, or stops not always reached in the same order) in a standardized format – Ensuring the route changes are reported in a timely manner to the individual responsible for maintaining the data feed – The ability to get the data out to developers at the speed they want – Local development environment – Developers want a wide array of features – Slow data retrieval – Allowing direct access to data through agency firewall – How to provide large amounts of data in a timely manner LESSONS LEARNED The lessons learned noted by survey respondents cover four major areas as follows: • Data quality and accuracy are critical to the success of an open data program. Respondents mentioned the following: – Put quality checks in place when opening data; – Be as open as possible, but test the data before releas- ing it to developers; – Start small in terms of the amount of open data offered and then grow that when confident of data quality of new sources/data sets; – It is important to have good, clean data—things that you understand internally as a transit agency don’t always translate well to people who are less familiar with your operations; and – Data must be compatible with or identical among the different formats in which they are made available.

53 • Open data are not free. – If you do not have staff to support open data (plan- ning, engineering and maintenance, especially), do not implement such a program; – Providing open data requires a lot of technical under- standing when establishing it. Think of the costs that come with providing such data; – Use standards to make it easier to provide open data; and – Select a technology vendor that supports open data or require it in the contract with the vendor. • Recognize that opening data will create changes within and external to the agency. Respondents summarized this point as follows: – There is a shift that agencies have to get comfortable with—from providing solutions to providing data; – Open data will not solve every customer require- ment, and agencies will still have to stay in the game (e.g., SMS, accessible services); – Customers are smart—they can tell which third-party services are best, and they will not hold the agency responsible for third-party services that are poor; – It is important that agencies not interfere with the market to ensure that the benefits of competition can be realized; – Further, an open data program should be supported by a project champion; – Carefully assign staff roles and skills; – Having buy-in from coordinating agencies is crucial; – Considering open data is a fundamental part of the overall information system; and – Ensure that data reuse complies with public policies. • Engagement and developing relationships with devel- opers is key to success as well. Respondents mentioned the following in relation to engaging data users and reusers: – Early engagement with potential users is key. Find out what they want and how they want it. Try and track who is developing what, particularly to understand the successes and failures; – Respond quickly to opportunities. Developers work on much shorter schedules than planners; – Developers will know the latest mobile platforms and can utilize these with your data; – Make it as easy as possible for developers to access your data, and make the license understandable and intimidating; and – Developers will help you determine the quality of the data if you provide a forum for this type of feedback.