The problem with 'Shadow Work' is that it rarely benefits the people who have to actually do the work.
My company has a homegrown A-B testing platform. Without getting too far into the technical details, the process typically occurred as follows: A business unit would suggest a performance increasing hypothesis to the statistical analyst ("let's change the font size from 12 to 14, I bet that'll increase conversions"), who would clarify the hypothesis into a statistically correct terms ("let's set 1% of our traffic to size 14 font and segment the resulting revenue this way") and submit a ticket to the engineering team, who would set up the test using XML, then ran a script that would convert that XML into MySQL insert statements, which populated a table that the web application would read from.
Then we had layoffs and engineers were canned. To mitigate some of the work for the engineering team's now reduced staff, I gave all the analysts the necessary permissions to set up and execute the XML -> MySQL translation themselves. On the one hand, they appreciate the direct control. On the other hand, they've fucked up production more than once with malformed XML and the like, which is something they never had to worry about before, which in turn has caused huge headaches for my team.
Is this 'shadow work'? All I know is that my company used to employ more people, and both engineers and analysts were happier. Now nobody is happy, even if the productivity per person is higher, except maybe our executives who have squeezed out more productivity out of fewer employees.
Hmm, maybe I should see what those OWS guys are babbling about after all.
There is a defensible reading of that story which goes "Our engineering and marketing teams stumbled into a ridiculously inefficient process to do something critical for the business. After a bit of reorganization, while the new process has some kinks in it, we're increasing sales faster than ever while simultaneously not wasting thousands of dollars using senior engineers as glorified typists."
n.b. I don't want to criticize engineering choices from afar, but if people can bork production with malformed XML, that suggests opportunity for further process improvement to either check XML or render it unnecessary. Visual Website Optimizer, for example, mostly abstracts that away.
P.P.S. Quantify the problem to management, fix it, get heavily rewarded. You could even use the proceeds to take a trip to Wall Street, if that floats your boat.
- I gave font size as very easy example, and there are dozens of products that make that very easy. But a lot of our testing ends up being very heavily backend-oriented (e.g. "let's try this new optimization engine combined with this optimization engine"). So any third party tools probably won't be plug and play unless we made some fundamental changes to our web applications.
- And I have quantified making those changes long before the layoffs (although I had another in-house solution in mind, not one that used third-party tools), because even if 1% of engineering time is spent executing scripts like robots, that's 1% that could be better deployed into a profit center for the company. The problem is, like I'm sure many engineers understand, is dealing with executives who don't quite grok the technical limitations of the current system, and who have a hard time prioritizing anything that doesn't directly lead to more money in the bank account.
- And this is my main point... so we had layoffs, and ended up coming to a defacto more efficient solution, which would be okay if me or anyone else involved in this solution was compensated accordingly. Instead it's executives that will get lauded for cutting costs, even though the current solution now introduced a high risk of impacting our ability to actually execute on the aforementioned initiatives that would lead to directly more revenue.
- So basically, what I'm most frustrated about, is this: I had a proposal that would have improved an inefficient process. That proposal got ignored, then we had layoffs, then we were forced to improve on that inefficiency in a way that wasn't that much more efficient. If we ultimately get this all to work, the executives get bonuses for increasing profit by lowering headcount. If it doesn't, they get cut loose with what I assume is some generous compensation package.
I'm not trying to just rattle some populist chains here or complain about "the 1%." My point is, it's not the actual elimination of jobs and creation of 'shadow work' that is a problem, but the context of which is eliminated. In most cases it sucks for the people whose jobs got eliminated, and it sucks for the people who now have do the shadow work because they're not seeing the rewards of the new efficiency (e.g. lower prices at the grocery store). Instead, some layer of people highly detached from the process typically gain the most benefits.
Interesting. I'd much rather have root on my machines, bag my own groceries, pump my own gas, find my own books and tickets on the internet, and drive my own car. Not only is it less expensive and more convenient, it is more egalitarian than ordering around clerks or gas station attendants. It is also more socially responsible, as you are less of a burden on society.
The author's position on "shadow work" appears to be opposition to productivity and self-reliance, along with a pining for servants and dependence.
I'm not sure it's cheaper. I didn't get charged extra for having someone check me out at the grocery store, but the baggers are faster than I am.
If I'm paying them a fair wage for their time, how can you call me a burden on society? Is your boss a burden on society when he tells you to do your job?
Are you against all forms of division of labor or just the convenient ones?
I think the self-service lines vs. serviced lines is a bad example. When I'm in a serviced line, I'm basically left to do nothing, while the cashier scans my items. The cashier is unnecessary, because I'm already there. There are two people to do the job of one person.
Now, interestingly, I really shouldn't be there. The cost of shopping for groceries on a whole is _very_ expensive, time wise. I would rather sit at home (or do something else productive) than walk around the grocery store and wait in lines for 45-75 minutes.
However, I still don't use grocery services, and I know very few people per capita willing to let others shop for their groceries, even though many of them bring home over $75 an hour. Why is this?
I think an economist would tell you it is human irrationality. However, I think there are a few factors - a big one is that a change is as good as a rest. Doing 40 hours development and 10 hours running errands is a nicer balance than 50 hours development, and probably with similar productivity. There's an overhead to hiring help too - setup and management. You're also becoming more removed from everyday life. That might be a benefit or a drawback.
That said, I recently hired a servant to take care of all those things that I don't like to do (I live in a low-wage country). Shopping, Cleaning, Laundry etc. About $400 a month, although it could be a lot less. Wouldn't go back.
Grocery bagging specifically is an interesting point culturally. Bagging helpers do exist in the UK, but 99% of the time you will bag your own groceries as the cashier scans them. It gives you something to do while waiting, and enables you to pre-organise the bags any way you like.
There is something of a time loss if you can't keep up with the cashier, but that's not often much of a problem. The few times I or my husband have started bagging stuff in grocery stores here in California, though, we've gotten strange looks - and often, at Trader Joe's, where the cashier often bags after scanning everything - words of thanks.
I'm with you: if we were seeing a discount at the checkout when I self checked out, then maybe I would argue differently.
When I buy tickets on the internet, I get charged a convenience charge... costing me more for doing it myself. When I drive myself, I have no doubt that everything in, it costs me more than catching the bus.
Unfortunately these are all things that aren't necessarily obvious. "Doing it yourself" is a great idea, but it's not necessarily cheaper.
The baggers are faster than you only in the rare scenario that you are the only customer in line. Self-service checkouts are, at least in theory, always available. Not only do grocery stores rarely keep all checkout aisles staffed, but they don't even set complete elimination of lines as a goal: If there's ten counters, three are open, and each of the three has three people in line, that's just fine from the store's perspective. They'd rather you spend your time waiting than have their counter staff waiting around when customer volumes are low.
Self-checkout works because it's cheaper not only cheaper indirectly (through lower prices) but directly as well: The customer is given the choice of waiting in line for a human or doing it themselves, and they can ponder all of the economic variables, like their preference for not doing manual labor versus the value of their time.
I'm not sure how old you are, but gas stations used to have 2 prices. One price for self service and one for 'full' service. Over time I'm guessing more people used the self service to the point where even offering full service didn't make sense.
One day soon, RFID will be doing all of this. We won't need cashiers or baggers or the customer to do anything. You'll be told (based on the RFID tags in your basket) what you owe, slide your card, then walk out of the store. It'll be much faster than what we do today.
Bagging the items will be optional, or you can just bag them yourself as you shop.
Interesting that you think RFID will detect which items you owe but that you will still have to swipe a physical card to pay. How would RFID accurately distinguish your products from another customer's? What about fruits and vegetables, or other items either untagged or charged by weight?
My company has a homegrown A-B testing platform. Without getting too far into the technical details, the process typically occurred as follows: A business unit would suggest a performance increasing hypothesis to the statistical analyst ("let's change the font size from 12 to 14, I bet that'll increase conversions"), who would clarify the hypothesis into a statistically correct terms ("let's set 1% of our traffic to size 14 font and segment the resulting revenue this way") and submit a ticket to the engineering team, who would set up the test using XML, then ran a script that would convert that XML into MySQL insert statements, which populated a table that the web application would read from.
Then we had layoffs and engineers were canned. To mitigate some of the work for the engineering team's now reduced staff, I gave all the analysts the necessary permissions to set up and execute the XML -> MySQL translation themselves. On the one hand, they appreciate the direct control. On the other hand, they've fucked up production more than once with malformed XML and the like, which is something they never had to worry about before, which in turn has caused huge headaches for my team.
Is this 'shadow work'? All I know is that my company used to employ more people, and both engineers and analysts were happier. Now nobody is happy, even if the productivity per person is higher, except maybe our executives who have squeezed out more productivity out of fewer employees.
Hmm, maybe I should see what those OWS guys are babbling about after all.