Early lessons from the micro-purchase experiment | Hacker News Reader

Early lessons from the micro-purchase experiment

18f.gsa.gov

93 pointszeckalpha41 comments

Instead, the learning point here maybe is that, if you spend the time to spec something out properly, and make it into a nice little package, and say its for the good of the government and community, maybe you can get some altrustic programmers to help out, if you give them sufficient credit.

Maybe instead you should think about how to better create a list of timeboxed /scopedtasks that any programmer can come and submit a PR for, and how to better reward those people with fame or other perks

- Leaderboard of contributors, # of prs, dollars saved taxpayers?

- Let users link to their contributions / put in resume?

- personal letter from the mayor once you submit 3 successful PRs that are accepted?

I'm not sure if your comment is sarcastic or not, but it actually wouldn't be a bad idea. Working on a well defined, self-contained task in an open source project can be very rewarding in itself.

not sarcastic at all.

gamification of altruism

Has a dirty feeling to it, but I don't see why it can't work =)

Encouragement to work in the US Digital Service is altruism along similar lines.

HN is always full of news about startups that burn millions of dollars in order to, perhaps, eventually, have enough of a product and enough exposure to become a profitable business.

How is what the winner of this bid did any different from what all these startups do every day? At least they staked their own money to make a name for themselves.

Isn't the term for this a "loss leader"?

They say that functional (and by all accounts fantastic) code was delivered for the winning $1 bid, and then they claim that the experiment was a success. But was it really? Is a $1 winning bid in the spirit of this? Couldn't someone who just wants to pad their resume submit $1 bids for everything, thereby "winning" 100% of the time and depriving of actual income these legitimate businesses who might count on these contracts for survival?

I'm not making a judgment either way, but allowing a $1 bid seems... strange. Are there other considerations that go into the "success" of the platform, or success merely defined by the delivery of functional code for the price of the winning bid?

Based on the parameters of the experiment, it was a success. Also, we learned things! The next experiment will look a little different. Iterate, improve.

*I work for 18F

I'm the author of the blog post, but the claim of success is based on a number of factors other than the $1.

As we note re the $1: > "This unexpected development will no doubt force us to rethink some of our assumptions about the reverse-auction model."

Also, the success is borne out of the data: > "Given these statistics, we think it’s fair to say that there is a market for open-source micro-purchasing, though we obviously will need to spend some time and effort refining our methods.

The success is in validation of the idea, not the actual result in this case. That said, the result is new, functional open-source software for the American public. So that's still a pretty good outcome.

If you don't mind me asking, do you have any plans to prevent extremely small bids like this in the future?

I hope they don't. In fact, I hope they encourage these kinds of small bids from qualified individuals.

100% test case coverage for $1 is amazing and to the benefit of all taxpayers.

I think they're great too. I'm just curious what 18f thinks about them.

What happens when the company that won tries to get on schedule 70, and are negotiated to contract to government for some fraction of a dollar?

It sounds preposterous, but government has done more preposterous things in this sector.

Are you considering allowing negative bids in future, as suggested by MichaelGG above?

From the post, it really seems like this was an experiment - and they seemed to have learned a fair bit from it. I'm not sure I'd take this single $1 result and say that'll be typical in the future.

I agree that $1 can be strange, and some can perceive it as unfair. But there are a lot of business models, where the goal is to get a foot in the door, before monetizing in the future (freemium comes to mind). And if it financially doesn't make sense, I'm doubtful that at scale everyone will be bidding $1 and providing great code.

Hopefully, though, the system won't make it hard for smaller development firms, which may struggle most with concessionary pricing.

I think the main problem is that while the pr is fantastic, with doing this for free (actually worse than free, the single dollar incurs the overhead of payment, taxes etc - without actually providing any of the benefits of payment) - they've not successfully tried out making a commercial market place. They've found a way to solicit free work.

While certainly interesting in its own right (how can we get random people to contribute solid open source modules to government IT projects?) - this isn't actually a success for growing a marketplace where contractors can participate and make money from these projects. Unless they can figure out something to fix the pricing. Perhaps having 10.000 such projects available would lead to most of them having a more real competition for "for profit" bids. Or maybe it would just turn into some kind of Amazon Turk dystopia with desperate people contributing code that takes much more time evaluating than it would take to write a proper solution.

I'm really exited this went so well (They did it; the winning bid provided a great solution) -- it will be interesting to see where this ends up going (and how they plan on handling support and project management as/if this model picks up steam).

These are tax dollars at work. The lower the bid the better.

I don't know that I buy that. Cheaper isn't always better.

I think what the parent means to say is: "The lower the bid ceteris paribus (all else being equal) [0], the better." In this particular case, the article states: "Not only did Brendan Sudol meet the requirements of loading the data, the new code had 100 percent test coverage, an A grade from Code Climate, and included some new functionality to boot."

[0]: https://en.wikipedia.org/wiki/Ceteris_paribus

They very clearly laid out the requirements for the project.

Why is this a question or a problem at all? If a business requires overpriced contracts to survive... then they should die, right? It's like photography. Photogs were complaining about hobbyists willing to enter the market for free. How Windows shipped with free photos from non-professionals, and how that hurts them. ... So what? If your work is something someone will do for fun, fame, etc. then why should you be paid for it?

So long they have QC controls in place, it sounds great. In fact, they should consider allowing the price to go negative as it's clear people are willing to _pay_ to do this work.

> If a business requires overpriced contracts to survive

Greater than $1 is now overpriced?

If there is supply at $1... yes ;)

Of course, there might be supply at $1 for one such project, not tens of thousands of projects.

The market is probably around -$500 to -2000 so yes.

If the government is getting better code and paying less than it otherwise would, I call that doubly a win. I don't think there's anything "illegitimate" about doing it for $1 - if people consider the resume padding to be worth doing the work for, why not let them?

This who depend on contracts for their survival have many many other potential clients who make money through profit rather than taxation. I'm fine with solving tragedies of the commons more cheaply.

If there are so many people willing to donate quality code that there's nothing left for legitimate businesses, maybe those businesses are doomed to failure?

To anyone who might argue that the resume-building bidder has taken food off the plates of other contractors, consider this: thanks to the low bidder, the government can now afford to process that data, build that resume, pay the low bidder, and give money away to the other bidders for free.

> and give money away to the other bidders for free.

I'm not sure what you mean. Surely the money will be spent, but this wasn't some kind of lottery, the remaining money isn't split among those that lost the auction...?

I was being facetious. There is no "remaining money" because the total cost of that work was $1.

As per previous comment threads this is currently just an example of experimental bias.

http://www.smbc-comics.com/?id=3025

Although I have heard of this behaviour before in government bidding, the fame makes it meaningless atm.

The 10th iteration will/would be interesting.

"the experiment also validated the core concept that open-source micro-purchasing can work, and it’s a thing we should try to do again"

This leap in logic seems entirely unsupportable. This was not a "purchase". It was a donation, very much within the long tradition of donations to open source projects.

There were plenty of bids before it dropped to $1 (Note: I work at 18F)

But those other bids dont matter. Only the winning one does

My point is that the other bids offer some support to the concept, regardless of the final bid and PR.

Previously: https://news.ycombinator.com/item?id=10471677

The big question is: will this work on scale?

Sure, 1 programmer can use this donation to pad his resume, but will this work if everyone is doing it?

Well if it doesn't scale, presumably the market price will converge on average back towards what is actually fair.

There seem to be so many things wrong with this.

How about if the person didn't deliver? Would the project be delayed and reauctioned?

Haven't the problems of lowest-bidder quality been illustrated over and over?

Wouldn't they garner a lot more good will if they just ask for volunteering?

I'm glad the US Government adheres to and pays the minimum hourly wage for work such as this.

Quoting the blog post on the the $1 bid: "It was a plot twist that no one here at 18F expected. This unexpected development will no doubt force us to rethink some of our assumptions about the reverse-auction model."

* I work at 18F

Reply on news.ycombinator.com