One mega trend that we’ve been excited about recently is around the ways that technology is allowing for a re-organization of labor markets.  This has been a real trend since the first internet bubble (eLance, Kosmo), and is exploding at the moment.

What technology has done in these markets is streamline three things.

First, customer acquisition and ordering.  Being able to find providers and transact more easily, faster, and with less friction.

Second, coordination of resources.  Efficient routing and labor utilization.  Managing complicated workflows with multiple parties.

Third, enhancing the end product or service.  More on-demand.  Cheaper.  Faster.  More convenient.

What I’ve found interesting is that these forces have caused some markets to become much more fragmented by empowering the individual worker (eg Custommade) while other markets have become much more centralized by increasing the importance of the enabling company (Uber).

This got me thinking about what forces drive certain types of market structures to emerge.  At first, I thought the distinction would be skilled vs. unskilled labor.  Skilled labor markets would lead themselves to technology enabled fragmentation while unskilled labor markets would lead to technology enabled centralization.  But that didn’t seem quite right.

A conversation with my friend Aaron White led me to a slightly different take – that there is a distinction between craft labor and trained labor.  Craft labor tends to differ quite a bit from provider to provider, and the buyers of craft labor are very discerning, and can easily tell the difference between different providers.  Trained labor yields a service that is fairly similar from provider to provider, assuming that the provider is reasonably well trained.  Trained labor can be quite skilled.  Conversely, even unskilled labor can have elements of a craft-like market.

Some types of services fall in unexpected categories.  Programming is skilled, but for a number of different types of projects, is more trained labor than craft labor.  Hence the activity you see on elance where some low-end work can go to the lowest bidder.  Craft labor arises at the higher end of that category, or in more specialized types of projects.

I’d argue that house-cleaning is more crafty than one would think.  Don’t believe me? If you have a cleaner, haven’t you been able to immediately tell when a different person or crew did the job vs. your regular providers?  Buyers are super discerning, and providers usually aren’t really trained, they just do their job the way they think it should be done.

On the other end, I’d argue that driving, while skilled, it totally trained labor. You can tell the difference between a crappy provider and a good one, but as long as a driver clears a certain threshold, you still get from point A to point B quite reliably.

That’s why I think personally think Uber makes more sense than Lyft. The ethos of Lyft is more crafty, and more about empowering the individual driver.  Uber feels more like the consolidator – the common service layer that customers primarily interact with.  They are using technology to consolidate their labor market.

This is also why I think the feedback against many of the cleaning services has been so negative. Just check out the Yelp reviews – many many 1 or 2-star reviews.  Those companies are trying to consolidate a market for labor that is more craft-like that one would think.  The only way to overcome this would be to spend a lot more time and effort to train and enforce standards of consistency and quaility, which would really hurt the scaleability of the model.

It’s not a perfect theory, but it’s a lens that I’m starting to apply more and more when I see companies that are disrupting labor markets. I think it’s a really exciting theme that is leading to two very different outcomes in different sorts of categories.