Uber for Support?

Many thanks to Monico Mallari for suggesting this topic. The question is whether a Uber-inspired approach to support, with contractors bidding on cases, as it were, would work. Interesting idea.

There are many similarities between assisted support and assisted transportation (that is, Uber and taxis):

Requests arrive at random intervals.
Customers prefer not to wait.
There are times when demand outstrips supply.
Service providers have different specialities/territories/vehicles that match some customers’ requests better than others.
Some customers want to pay less and will accept sharing a ride, or using an online community, even if it takes a little longer and it is a little less comfortable.
Some customers are willing to pay more to get a black car, faster service, or a more experienced helper.

There also have stark differences:

Support customers usually hold a contract that promises them certain terms of service (SLA). Travelers pay per ride.
For complex support at least, the resolution time is uncertain. With a car, traffic can make the trip much longer than usual but we usually have a pretty good idea of how long it will take when the trip starts.

With that, can one use subcontractors to bid on cases? One of my clients built its business on the idea, so it is absolutely possible. Interestingly, customers love the ability to choose the support engineer among those available at the time or the request, or to wait for one they know or has better customer ratings. (Uber doesn’t give customers a choice, not yet at least, and when it does you can remember that you saw the idea here first.) And workers fiercely compete for high ratings so that their profiles are featured higher in the queue of available workers, ensuring that they get a steady stream of customers.

But there are caveats. I see four that matter for support:

If you sell support contracts (or contracts that include support, as is the case for SaaS services) in which you commit to specific response times, you must set up a way to guarantee a timely response, especially if the response targets are tight. This means that you may need to compensate people for being available, at least during low-volume times that are not attractive for workers. (Busy hours are less of a problem; many folks want to work when they are guaranteed to get business.)
For complex support, you need to define how the customer will continue to engage with the owner if the case needs multiple interactions, a common situation. If workers work irregular hours, this is a challenge.
You need a pricing model that is fair to the workers. This is difficult for complex support. We all have stories about the 100-hour case that would not end.
It can be challenging to motivate contractors to participate in knowledge management activities or even keep good, reusable case notes. If I write down all I know, will you need me tomorrow?

That said, the idea has legs, at least for moderately complex cases and non-elite customers.

Would love to hear about any experiments under way around these ideas.