Draft Statement of GovTech Principles for Cities

// This would be a kind of "code of conduct" or "statement of values" cities would prescribe for potential govtech startup vendors.
// VERY DRAFT & UNOFFICAL

- *Citizen-first experience*: In the private sector, simple, beautiful, and easy to use technology has become commonplace -- and that’s happened because user-centric technologists have prioritized user needs and experiences. The city expects the same for the citizen experience: seamless, smart, and accessible.
- *For (All) the People*: The City doesn’t have the luxury of catering to one demographic; our technology must work for any affected resident, no matter ethnicity, access to technology, gender, etc. Government technology must have a bias towards all.
- *Built to scale*: The City prioritizes software-as-a-service technologies that not only can solve a problem locally but also scale nationally, and with that national reach continually deliver better and better product for citizens.
- *Default to open*: Any technology put in place will generate massive quantities of data, and the city has a long-standing commitment to transparency and openness--not only to ensure accountability but also to enable plug-and-play flexibility as we continue to pilot new tools. Good government software should default to open data.
- *Solves real problems, for real people*: The City is not interested in technology for technology’s sake; instead, we recognize software's *pivotal* role in governance and service delivery. Good government software should and must do the same, putting people, and their problem’s first.

First bullet: most cities serve more than just citizens or even residents (as the second bullet reiterates). For a statement of broad principles is there another more inclusive word that could work? Stakeholder?

Regarding "Citizen-first experience": there are plenty of solutions that aren't citizen-facing, such as fleet management, human resourcing, work orders, and procurement. And there are others that are equally citizen-facing but also employee facing, such as contracting, hiring, and case management. How do those fit it in to this picture - or will they be intentionally excluded?

Regarding "Default to open", consider "open by preference" instead. I think open by default is a great philosophical approach and important to embrace, but it also can't be taken literally when it comes to the technological end of things - particularly database queries.

This is a good start, thanks for doing this. If you could add some words to the effect that virtualizing the tech accelerators and incubators will help to bring more people into the startup queue, that would be great. 1M/1M run by Sramana Mitra is a good example of this, but there are others, not necessarily focused on govtech though. http://1m1m.sramanamitra.com/

Here is NASA's new OpenGov plan which includes several new initiatives to encourage small business innovation and plans for tech accelerators. I would love to see more federal, state and local OpenGov plans incorporating your best practices and some of NASA's innovation goals. https://open.nasa.gov/open-gov/

Good things here, Abhi! A few thoughts:

I'm not sure I agree with "Built-to-scale". SaaS encourages cities to outsource to closed solutions instead of bringing tech in-house, which would be more valuable for most cities. Instead, I'd strongly propose that cities learn to collaborate with other cities and organizations to build tools with their shared limited resources, and open source the results. E.g., DC's successful Free Law Innovation Fellow program.

In putting people first, the language should focus on the people not on the technology. This is a culture change and outreach problem, not a software one. How about something like "Work with the people – through meetings, town halls, and direct feedback sessions – to learn about their problems before beginning any project, and gather more feedback throughout the process in an iterative way."

Also, I would encourage you to not use "citizen", as you're excluding the many residents of cities that are not citizens - "people" is preferable under most circumstances.

This is a great start!

I find that the "Open and Protected Data Policy," created jointly by the cities of New York and Boston, a good policy platform for the governmental data aspect of your proposal. See:

• "Boston, New York City craft new open data policies" [ http://statescoop.com/boston-new-york-craft-new-open-data-policies/ ]
• "City of Boston: Open and Protected Data Policy" [ https://data.cityofboston.gov/City-Services/Open-and-Protected-Data-Policy/2rjs-rb6r ]
• "City Hall’s Technology Journey: Using Data to Improve the Lives of Citizens" [ http://www.bbhub.io/dotorg/sites/8/2016/05/GPL.bch-technology-v4-5.pdf ]

This is great! I'd like to propose a new bullet point:

Minimalist, modular software Otherwise known as the Unix philosophy. Wherever possible, large IT projects should be broken down into smaller component parts, with each piece capable of importing and exporting data using the most practical data standards.

Everyone that's been in the govtech space long enough has seen the failures of large IT projects. By pushing norms to break things down into smaller component parts, loosely joined, gov's will lower IT costs and increase the size of their vendor pool.

Also, a note on 'built to scale'. In my experience, putting the 'scale' ahead of your client's needs is a recipe for failure. For it to work for others, it has to work for at least one. Thinking about scale comes after you've built and deployed it at least twice. Many technologists and startups are blind to this.

Thank you for starting this, Abhi! Some thoughts:

• Let's watch out for the danger of saying essentially "software should be good". That is, every vendor will always say their software puts people first, enables plug-n-play flexibility, lowers costs, etc :-). I think Derek's point about minimal, modular software is good, partly because it can be an actual constraining guideline -- like, when a city procures a $500 million monolithic resource tracking system, they can be pretty sure they haven't met that guideline, cough cough.
• Here's a possible additional principle: "Default to open source. Whenever possible, prefer open source for both custom software and COTS software. Vendors should do development in the open, from the start, and the software should come with no restrictive copyright or patent encumbrances. Whenever possible, city IT staff should work collaboratively with vendors, using public collaboration tools. This promotes reusability and prevents lock-in."
• Another one, maybe related to / foldable into the above: "Maintenance is required, and is an opportunity. Any software procurement should include a long-term maintenance plan, which may include further development or enhancement. This phase is an opportunity to include new vendors, and procurements should be structured with that as a goal." (Well, I haven't worded that very well, and am not sure quite how to fit it in, but I'll leave it here as a point for discussion anyway.)
• Regarding the Default to open item: Specify not just open data, but data in parseable formats? Too many cities treat making opaque PDFs available as "open data".

A few quick thoughts:

• Bullet 1: I don't think the private sector has done a good job at this at all. There are a tiny number of companies on the west coast that do a decent job, but frankly, about 80% of private sector tech companies still have terrible UX. I also think the "private sector does it well" meme is not helpful in addition to not being particularly true in a broad sense.
• Bullet 5: I agree with the default to open, but WAY more important is that cities should insist on OWNING data. They should also OWN the relationship with the resident. Applications and companies that try to get between residents and the cities with their own brand are, in my mind, a problem.

Finally, I get the intent here, but I think these are just principles that cities should adopt, telling companies they have to adhere to them feels like it causes the judgement of this stuff to shift from the city to the vendor, which is a problem. This are rules for CITIES to follow, which they should then IMPOSE on vendors. We deal with a ton of cities that don't have a clue, and I shudder to think of how vendors could claim they at adhere to these and no one would verify the claim (I see this all the time around some of these best practices - the number of times I've seen vendors flat out lie is not trivial). So not sure this has anything to do with civic start ups. Think it has more to do with rules for cities.