Strike Network Postmortem

The article was originally written on November 2, 2011. The application was written during August and September of 2011.

Strike Network started as a mobile site similar to 4square. After realizing that the audience was tiny and without an app there was no lock in (and with an app, there's a barrier to entry too), decided to try making a facebook app.

The facebook app was a litte more complex than most fb apps. There was a store search, a store solidarity registry where people would pledge support, a directory of players, and some news. The idea was to get people to declare their support of the workers on strike. The idea turned out to be a dud - people didn't understand the concept, or told me that others wouldn't understand the concept.

More useful was a store search to find stores that were not on strike. This was challenging because building that list is difficult. Getting the list was hard, and I never got one, because the UFCW only has lists of their organized stores available -- so you can find another organized store that's not being picketed. Even within the received data, there's a lot of data clean-up. So I added a feature to let users add stores. I don't think anyone used it, because the effort required was quite great. (At best, I'd expect one or two stores to submit their data per week. There are a lot of barriers to participation here. So I ended up entering data for stores near my users.)

FB apps are also a barrier to entry. Of my FB "friends", maybe 10% to 15% use apps, and of those, only a few use them regularly. Usually, it's zynga style games. (At the time, the intrusive "watch me read news" apps weren't so omnipresent.)

Design-wise, I noticed that the more popular games have tutorials built-in. So Strike Network was at a disadvantage - people didn't comprehend it. Interactive explanations seem to be important. Also, the eye candy matters - this means funny animations, slick graphics, and cool looking effects. There are a lot of unpopular FB games that are okay to play, but lack eye candy.

Conceptually, a registry of solidarity declarations for workers on strike didn't work. The only people who'd understand are union members, leftists, and some college students. Of these, you need to find the people who use facebook and use apps on facebook. We got 27 users after reaching out to around 100. Only a handful were people outside of the initial 100 contacts. So there was almost no viral lift, though there was immediate growth.

Today, people don't understand strike solidarity, which, having grown up in a part of LA that (unknown to me) had a high level of union membership, seemed like a simple idea. In the past 30 years, the number of industrial strikes has declined, and among those, very few have been community supported strikes. It's not the 1970s anymore. Apps with mass appeal need to work with what people understand, and people do not really understand strikes. Without more widespread union membership, "solidarity" is at best a nice concept. It cannot be realized by the 85% of people without a union job. An educational app might be more appropriate.

The app was too complex, with too many features (though it was my intent to make it simple). Occupy Wall Street, for example, led to the creation of many software tools, and only a few, like Occupy Together, really attained popularity. (Software development around Occupy Wall Street would be something worth researching.) The simpler tools were more popular than the more complex, richer tools. People used the tools to satisfy key needs, like finding and Occupy and avoiding duplication of regional organizations. The tool itself wasn't technologically innovative -- but it was the right tool in the right situation. When you try to introduce new technological ideas into a new social context, there might be too much novelty, and it can be confusing.

The virality of making pledges of solidarity didn't work - people didn't get it. (A pledge of solidiarty simply means not crossing the picket line - the effect is to deprive the business of profits.) The reason why viral virtual gifts work is because "gifts" communicate something people understand. The same goes for "meters" and "quizzes". "Pledges" to union members was interpreted as gifts of money, so that didn't work.

The context of a strike is important: a strike within a community needs community support, and that means individuals and organizations. In L.A. there is a lot of individual support, but that can be a barrier to growth, because solidarity is difficult between an individual and a union. Solidarity is possible between an organization and a union, so getting people involved in organizations that are friendly to unions and would support strikes is important. It also creates a context where labor unions can then support the community organization. There is a desire for progressive/left political community organizations as was demonstrated by Occupy Wall Street, which ranged from leftist to centrist, and even center-right or libertarian in some locales. There may be more potential for apps that help establish or strengthen community organizations through community organizing for progressive change.

Technological prior art: There was almost nothing "new" in this application, and most of the ideas were copied from other sites. The mobile checkin idea was from Foursquare, of course. Mapping actions was from Platial (a defunct mapping site that let you add information to maps). Declaring solidarity was from the social networks, particularly Myspace's comments, but also from "guestbooks" from the 90s, and Indymedia. Finding stores was from Yelp and the phonebooks. There were also some odd features in the code, like searching for similar sounding place names, and caching calculated distances to speed up searches, which I'm sure someone has already patented because it's obvious and useful.

[Ed. 20/Aug/2012 - after re-reading this blog, I have some new opinions. The project suffered a lot from my own myopia - it was designed without any feedback. Much of that was due to lack of planning; the system was designed as it was coded, more or less. As ideas emerged, code was written. The paradox, however, is that without going through the process of simultaneous design and coding, it would have been difficult to know what features were possible. While you can always design something, you can't always code it -- it might just not be possible, or it might run too slowly. The other problem is that the program, while not large, had too many features. A good, simple application has few features, but has all the necessary features. The hard part is determining what is necessary.]

[Revised on 9/23/2012]

I've added a dump of the database table structures, and a few graphics, so programmers can get an idea of what the app was. I cannot upload the entire source, due to some information that might be proprietary to the UFCW; it's also pretty sloppy. Unfortunately, the FB API changed, and the app doesn't work anymore.

Please note that all content is copyright 2011 John Kawakami. This code was written on my own time, without any knowledge by, or input from the UFCW or my employer.

Epilogue: Due partly to organizing by the LA Co. Fed, other unions, notably SEIU and the postal carriers, and community organizations, a strike was averted, and there was no repeat of the 2003 grocery store strike in Southern California. So the application had not been used. Later, when the contract for the Food 4 Less was negotiated, I think they got a better pension deal, so overall, things were going well for the UFCW (in some ways).

[Revised on 6/15/2015]

Removed first sentence of intro. Added the time breakdown for work performed. 59 hours. Not sure I have the code anymore. It was mostly jQuery and PHP.

banner.png55.26 KB
big-icon.png14.54 KB
homepage.jpg55.23 KB
1.instructions.png7.92 KB
2.instructions.png9.69 KB
strikenetwork.sql_.txt7.6 KB
Strike Network Project - Time Breakdown.pdf53.98 KB