Building the Next Gig Economy Startup

Let’s talk about what to do when the gig economy stops exploding and starts normalizing.

What Uber and Lyft accomplished— more than provide frictionless rides at lower prices — is they established and legitimized a contingent workforce dedicated to providing said rides. A gig workforce. I’ve gone through the process to become a Lyft driver. It’s kinda simple. Got a late(ish) model car and no marks on your background check? You’re in. You can drive this weekend.

Not long from now, your job might be a gig. You’re going get more freedom from your employer, which in most cases is going to be an app, a technology, a platform.

Make your own joke about our new app overlords, but that freedom is what’s fueling the gig economy. And when entrepreneurs connected the dots between money and freedom, they spawned the “Uber for X” startup movement. Starry-eyed founders stopped building the next Facebook, and instead started building Uber for Food, Uber for Housecleaning, Uber for Hugs. The movement jumped the shark with the Sliceline episode of Silicon Valley.

But now we’re seeing a narrowing of “Uber for X” startups underway. Entrepreneurs and investors are placing bets that the gig economy is about to normalize and become an even greater share of the real economy.

So two-sided market platforms are everywhere, because the technology to bring these platforms to life has become so ubiquitous that startups build the market before they figure out the market economics — which, by the way, is still a huge problem for both Lyft and Uber.

But what if we could build a minimum viable product (MVP) around a two-sided market and figure out the economics first. What would that look like?

I’m building one of these platforms, I’m advising two others at an earlier stage, and I’ve been trying to bring a third out of the idea stage for a few years now. Here’s what I’ve learned about building the basics of a gig market platform MVP.

We can fake everything to prove the concept. Building a two-sided market is a three-step process.

  1. We need to onboard providers and make them accessible.
  2. We need to discover consumers and connect them to the provider.
  3. We need to allow them to transact.

The one MVP feature that needs to be robust is the core offering, in this case, the tech to bring the two sides together in an efficient and transparent manner. But if we want to build a pilot to test the market economics, we can manually recreate the core offering too.

We can use spreadsheets to store our providers and our consumers and all their attribute information. We can use MailChimp to create mailing lists from those spreadsheets. We can use a combination of Zapier, Slack, and SMS to kick off a transaction. We can use Square to capture and transfer payment.

I can set this up in a few hours. One problem. It’s a huge, resource-intensive time-suck to operate in this manner. We’re never going to get to MVP with this much duct tape, but we can use the POC to start figuring out where the margins are.

Start local, small, and niche. The best way to test our proof of concept is to keep the boundaries closed, and we can start walling ourselves off by limiting our footprint. For a physical offering, we can choose a certain city. For a digital offering, maybe we limit participation to an invite-only beta. We can even use our personal network if it’s geared to the offering.

Then we need to go niche, limiting our providers to a single product or service, preferably the one that’s the easiest to deliver with the largest margins. Let’s give ourselves plenty of room for error and recovery.

The seller feedback loop is critical and where we should spend most of our time. If we don’t have verified, excellent, professional providers, we’re screwed. So most of our time needs to be spent on them.

We’ll need to make it easy to onboard, verify, and establish a baseline for trusted providers. In some cases, we’ll need to train them on our platform, not just the delivery, but our expectations.

We’ll also need to accurately score their performance every time they transact. This is the most delicate link in our chain, as we need a feedback system that is fair and thorough and true. We can’t have any bad apples amongst our provider population, but we also need to be very careful with calling a bad apple a bad apple.

We can’t drop the manual portion of onboarding, but we can streamline and automate a lot of it. We need to make sure our providers have the product inventory or the skills they claim to have. And that they’re not criminals or psychopaths.

We’ll need to verify, maybe background check, categorize, and qualify the providers in order to make the experience positive for the consumers. Automating this process is ultimately where our intellectual property and competitive advantage will be.

Reverse that logic for buyers, they need to be able to transact with as little friction as possible. This feels like doing it backwards, but we need to know a lot about the people we’re paying and very little about the people paying us.

If a marketplace is clunky for the consumer or we request too much of them, they won’t use it. My own data shows that “ease of use” is more important than “well-designed” by a factor of 2 to 1 and more important than “feature set” by a factor of 4 to 1.

Our platform should allow the buyer to transact with the bare minimum number of clicks to complete the transaction. If we’re good at what we do, we’ll figure out a way to get them to give us more information after the transaction is complete.

To monetize and model, follow in the footsteps of others. I noted earlier that I’m a driver for Lyft. I don’t drive, but going through the onboarding taught me what to do and what not to do. It also gave me a sense of what they offer their providers and how they motivate and retain them.

Do this. In fact, sign up for every gig market service you can, provider or consumer or both. You don’t have to transact, just reverse engineer the model on both sides and keep what works.

Incentivize providers for quality, loyalty, and participation, in that order. No good two-sided marketplace neglects its providers and I’d suggest using cash.

First, we pay bonuses for providers to get them started, like double their first transaction.

Pay bonuses for providers who make us look good. In fact, establish a pay scale based on quantity and quality of ratings and publish it.

Pay bonuses for providers who stick around a long time and become champions of our platform. A good stretch goal? Our top-level providers should be able to get rich off our platform.

Pay bonuses for providers who participate a lot, even if they’re not our top performers. These are our “glue” providers, filling gaps that no one else might.

Be the boutique option. This is probably the most important advice I can give you. A two-sided market will be expensive to run, especially at scale. This makes winning a price war difficult. Thus, a two-sided market doesn’t win on price, it wins on an element that the incumbents can’t provide.

Customers will pay more for time, convenience, and quality. If we can deliver those, we won’t necessarily be immune to pricing pressure, but it becomes a battle we can win.

Once we figure out the costs and the margins, we’ll have the confidence to automate the onboarding, the transaction, and the delivery, and we can keep the rest of our platform fairly small tech until we start to grow faster than the small tech can keep up. That’s a good problem to have.