4 ms·

Counter-argument: The work described here as "bullshit" -- administration, bureaucracy, ever-finer metrics of people and products -- actually DO generate "real"

by gipp 10y ago

Counter-argument: The work described here as "bullshit" -- administration, bureaucracy, ever-finer metrics of people and products -- actually DO generate "real" value in increasing the speed at which information asymmetries resolve themselves. E.g. people metrics actually make it harder to maintain the cocoons of bullshit the author describes, product metrics make the disparity between what's provided and what's desired more apparent more quickly, etc. The proliferation of these kinds of work is not necessarily out of "filling space," but because information technology has drastically decreased the marginal costs of this kind of information-only work, and so lowered the barrier of marginal utility required to justify it. So, we have information work where the workers actually carrying it out can't perceive the value, because it's too distributed and too many layers removed from the individual, but it's there. The apparent lack of productivity increase is because those metrics, too, are flawed and do not account for these kinds of value. Now, the question of whether that work is then worthwhile at a societal level is entirely different.

spitfire 10y ago
You wouldn't mind telling that to the Societs, or the Chinese would you? I see your point and sympathize with you, but accurate metrics are a necessary but not sufficient to gain results.
- mark_edward 10y ago
  Depends on what you consider to be measure of bureaucracy. If it's by the proportion of done aggregate work done in such tasks, China is probably less bureaucratic than USA considering how many of us are essentially bureaucrats and clerks and considering how many of them are farmers and line workers. If it's percent GDP I dunno, probably similar.
ArkyBeagle 10y ago
I can count on one hand the number of people I have encountered who could actually design instrumentation from scratch. It just. Isn't. Done.
fsloth 10y ago
I would claim that for analysis work to be worthwhile the organization must be more productive with the analysis than without it. If not, then the analysis is pointless. This does not apply to scientific research of course, but actual scientifically valuable research is a different thing entirely.
- gipp 10y ago
  Oh, absolutely. I'm not trying to make a blanket endorsement of, for example, "data-driven decision making" as a groupthink thing that everyone must do. That can, and does, lead a lot of unwary businesses down a lot of blind alleys. But the flip side is also true -- blanket dismissal of information work as "self-preserving bullshit" is equally, or more, naive.
  - fsloth 10y ago
    Sure - to be brief, the data must be actionable and understandable to be of value. A great example of value adding data gathering and analysis process is the work of Eniram - they take a ship, analyze it, and optimize it's usage to save millions in large shipping operations fuel costs globally. This after the shipping companys internal engineers have optimized the shit out of their operations. This is what value added data enrichment scheme looks like. I'm sure the analogous thing can be done to all sorts of organizations and processes. It's not about gathering obscure metrics, running them through excel and creating powerpoints with no actionable items.
zaroth 10y ago
Your argument boils down to 'the productivity metrics are broken'. But since I think we're talking about GDP basically, I doubt you're right. TFA argues that productivity is squandered by Office Space style ass-covering wastes of time. But I think even good intentioned beaurocratic nonsense can still account for the strong headwind against productivity. Take a look at the great improvements in supply chain management and JIT manufacturing techniques for true process improvements which result in measurable productivity rise. Kanban works there in ways it simply does not for software, for example.
- Bartweiss 10y ago
  Similarly, "good" overhead like A/B testing can suck up an enormous amount of time without producing anything useful. Badly-executed A/B tests are an easy way to show single-digit "gains" over and over again without ever producing actual movement, and since executing them well is genuinely difficult there are a lot of companies which are running frantically on a treadmill. There's a lot of business 'motion' that's fundamentally a mismatch between something looking good at an individual level, and achieving nothing on a corporate or national scale.
notahacker 10y ago
> So, we have information work where the workers actually carrying it out can't perceive the value, because it's too distributed and too many layers removed from the individual, but it's there. The apparent lack of productivity increase is because those metrics, too, are flawed and do not account for these kinds of value. I'm struggling to understand this bit. The ultimate purpose of information work is improve profitability, typically by making more efficient use of their inputs. The appropriate metric therefore surely is whether overall productivity actually changes, and the article argues that it tends to indicate that corporations are not becoming more efficient. Now I can see one tenable line of argument why work might be valuable to a company even as it drags down the measure of productivity: the information work is positional and therefore valuable for the individual corporation to ensure it doesn't fall behind the others even though the net result of the industry reinvesting productivity savings in additional information work is that generates the same returns as before. (The author of the article, incidentally, chairs a big group of marketing and ad agencies). Another would be the existence of diminishing returns to scale, so if firms are continue to grow they have to spend productivity savings from process improvements on additional labour to allow them to expand into more difficult markets. That certainly applies at a market level, but I'm not so sure whether it's true of individual firms. But I'm not sure whether those were what you were trying to imply?

by gipp 10y ago

spitfire 10y ago

You wouldn't mind telling that to the Societs, or the Chinese would you? I see your point and sympathize with you, but accurate metrics are a necessary but not sufficient to gain results.

mark_edward 10y ago
Depends on what you consider to be measure of bureaucracy. If it's by the proportion of done aggregate work done in such tasks, China is probably less bureaucratic than USA considering how many of us are essentially bureaucrats and clerks and considering how many of them are farmers and line workers. If it's percent GDP I dunno, probably similar.

ArkyBeagle 10y ago

I can count on one hand the number of people I have encountered who could actually design instrumentation from scratch. It just. Isn't. Done.

fsloth 10y ago

I would claim that for analysis work to be worthwhile the organization must be more productive with the analysis than without it. If not, then the analysis is pointless. This does not apply to scientific research of course, but actual scientifically valuable research is a different thing entirely.

gipp 10y ago
Oh, absolutely. I'm not trying to make a blanket endorsement of, for example, "data-driven decision making" as a groupthink thing that everyone must do. That can, and does, lead a lot of unwary businesses down a lot of blind alleys. But the flip side is also true -- blanket dismissal of information work as "self-preserving bullshit" is equally, or more, naive.
- fsloth 10y ago
  Sure - to be brief, the data must be actionable and understandable to be of value. A great example of value adding data gathering and analysis process is the work of Eniram - they take a ship, analyze it, and optimize it's usage to save millions in large shipping operations fuel costs globally. This after the shipping companys internal engineers have optimized the shit out of their operations. This is what value added data enrichment scheme looks like. I'm sure the analogous thing can be done to all sorts of organizations and processes. It's not about gathering obscure metrics, running them through excel and creating powerpoints with no actionable items.

zaroth 10y ago

Your argument boils down to 'the productivity metrics are broken'. But since I think we're talking about GDP basically, I doubt you're right. TFA argues that productivity is squandered by Office Space style ass-covering wastes of time. But I think even good intentioned beaurocratic nonsense can still account for the strong headwind against productivity. Take a look at the great improvements in supply chain management and JIT manufacturing techniques for true process improvements which result in measurable productivity rise. Kanban works there in ways it simply does not for software, for example.

Bartweiss 10y ago
Similarly, "good" overhead like A/B testing can suck up an enormous amount of time without producing anything useful. Badly-executed A/B tests are an easy way to show single-digit "gains" over and over again without ever producing actual movement, and since executing them well is genuinely difficult there are a lot of companies which are running frantically on a treadmill. There's a lot of business 'motion' that's fundamentally a mismatch between something looking good at an individual level, and achieving nothing on a corporate or national scale.

notahacker 10y ago

> So, we have information work where the workers actually carrying it out can't perceive the value, because it's too distributed and too many layers removed from the individual, but it's there. The apparent lack of productivity increase is because those metrics, too, are flawed and do not account for these kinds of value. I'm struggling to understand this bit. The ultimate purpose of information work is improve profitability, typically by making more efficient use of their inputs. The appropriate metric therefore surely is whether overall productivity actually changes, and the article argues that it tends to indicate that corporations are not becoming more efficient. Now I can see one tenable line of argument why work might be valuable to a company even as it drags down the measure of productivity: the information work is positional and therefore valuable for the individual corporation to ensure it doesn't fall behind the others even though the net result of the industry reinvesting productivity savings in additional information work is that generates the same returns as before. (The author of the article, incidentally, chairs a big group of marketing and ad agencies). Another would be the existence of diminishing returns to scale, so if firms are continue to grow they have to spend productivity savings from process improvements on additional labour to allow them to expand into more difficult markets. That certainly applies at a market level, but I'm not so sure whether it's true of individual firms. But I'm not sure whether those were what you were trying to imply?