As we enter what we’re calling phase II of local the on-demand economy (LODE), the environment will be defined by vertical expansion, M&A activity, and a fair share of flameouts. We’re already seeing the latter, with companies like HomeJoy and GoodEggs.

In terms of expansion, we expressed at BIA/Kelsey NOW that in addition to different business verticals, LODE will move up-market to higher-end professions. It won’t just be about drivers and dry cleaning, but also lawyers and architects, as we’re already seeing.

But today, one of the biggest question marks is about vertical business categories. And one of the highest spending local verticals is home services. So we see Amazon enter the space, and more recently Google. I’ve been asked several times lately about the latter.

Quality Control

To begin, Google Home Services will be a sort of reverse auction model where leads are aggregated through search and then sent out for the quickest and cheapest local providers to win the business. The nuts and bolts are described well at Search Engine Land.

This makes Google’s move into on-demand home services fundamentally different than on-demand models by Uber and others. Its more grounded in desktop search, versus the pure LODE model of tapping an in-app button to summon something.

In home services, this desktop approach makes sense versus more “immediate” categories like food and rides. There is of course a longer consideration cycle, which is precisely where desktop search traditionally plays a key role. So the use case is well aligned.

But applying an on-demand model to home services has proven to be difficult as shown by HomeJoy. Though the on-demand model generally has attractive unit economics, those don’t play as well with products that have a wider quality variance.

In other words, Uber’s model works great given a fleet of service providers that are relatively interchangeable. That’s makes their supply/demand matching algorithms work so well, as they’re governed by things like proximity and minimum wait times.

Home services conversely have lots of variance in quality level from one provider to another. It can also be a recurring service, where someone is coming into your home. So trust and loyalty are important on an individual service provider basis.

These are the home services realities that players like HomeAdvisor have carefully built a business around. They’re not only dealing with depth, nuance and quality variance for home service pros, but also a very broad spectrum of jobs and customization.

This not only make service providers less interchangeable, it challenges quality control and breeds complexity. And complexity in on-demand matching algorithms can equal margin depletion, especially with Google’s low-touch approach.

This represents another key challenge of home services in that margins are typically thin to begin with, so there’s not much room for error. This is what killed HomeJoy, despite the common cry that it was because of policy battles around employee/contractor status.

Leaky Faucets, Leaky Business Models

Another key challenge is the “leakage” problem of service providers going directly to customers after the first job is done. This is of course inherent in the type of service where, again, there is often recurring work when trust and relationships are built.

This obviously sidesteps the Google’s and HomeJoys of the world. To avoid it, Google will have to offer tools that keep service providers on the platform, such as insurance, lead flow, payment processing, and arbitrating disputes (like Airbnb does).

Google isn’t necessarily following the pure on-demand model a la Uber, and has its own flavor of home services that merges with search. But it will have to be careful of the nuances, logistics and margin compression that have challenged home services for years.