Ride Cannibalization — The Invisible Force Eating Operator Profits

By Rob Siliciano, senior economist at Zoba. Rob holds a PhD in Economics from Harvard University, and loves both urban mobility and subzero temperatures.

With mobility operators under increasing pressure to improve their fleets’ per-unit economics, it is our job at Zoba to provide operators with the tools and insights they need to increase rides, profits, and impact.

In this post, we’ll explore the impact “ride cannibalization” is having on operator profits and introduce a new metric, “Net Rides Per Deployment,” which we have developed to help operators reduce the impact of ride cannibalization by providing a better assessment of deployment profitability than traditional metrics.

Deployment Efficiency in Micromobility

Every day, micromobility operators across the globe move and deploy hundreds of thousands of scooters, bikes, and mopeds across cities — unfortunately, 1/3 of all deployments done today add no value to an operator’s network.

This is a massive problem for the industry, as deployments are often an operator’s largest expense. According to McKinsey, deployments, rebalances, and swaps make 40–50% of the total cost of a ride.

Given this, an operator’s deployment strategy can mean the difference between providing a profitable, sustainable service or suffering from negative unit economics and poor vehicle utilization. As operators look to cut costs, it’s no wonder many are taking a hard look at the efficiency of their existing deployment strategies — be they from a warehouse or as part of a rebalance. But what makes a deployment strategy good or bad?

At Zoba, we have spent a lot of time working with operators on these problems and have come to believe that a good strategy is one that:

1. Places vehicles in areas where users are most likely to ride them.
2. Prioritizes placements in areas where they are likely to remain in circulation.
3. Avoids over-supplying areas where there is already sufficient natural vehicle supply to meet demand.

It is on the third point — around over supply — that we see teams most often struggle. While straightforward on the surface, trying to estimate the number of vehicles needed to meet demand across dozens or hundreds of stations is no easy task. Throw in temporal variances such as weather, seasonality, and local events, and it becomes nearly impossible for a human to do effectively.

Ride Cannibalization — The Invisible Force Eating Operator Profits

Imagine a vehicle sitting at a station downtown, and you put a vehicle down next to it. An hour later, the new vehicle gets a ride while the first vehicle remains there, sitting idle. This is a scenario that plays out thousands of times a day in cities across the globe.

When teams deploy vehicles in locations where they are not needed, they make every vehicle at that location less productive. At Zoba, we call this phenomenon ride cannibalization.

Ride cannibalization is the reduction in rides (and therefore revenue) adjacent vehicles experience due to the introduction of another vehicle by the same operator.

Ride cannibalization is particularly costly for operators, as they incur the expense of deploying the vehicle while missing the potential revenue that vehicle could have captured if deployed in another location.

Unfortunately for operators, the current scale of this problem in micromobility is hard to overstate. Currently, teams spend thousands of hours per week deploying vehicles to locations that cannibalize their own fleets.

Looking across hundreds of millions of rides and deployments from operators across the globe, we estimate that as many as 1/3 of all deployments today add no real value to an operator’s network.

In locations where ride volume is high such as train stations, malls, and downtown areas, ride cannibalization is not a rare occurrence — accounting for half of all rides in high-demand locations.²

Perhaps driven by a bias to not miss out on rides, it’s a classic problem of right place, wrong time, and it is slowly eating away at operator profits.

The RPD Blind Spot: When More Scooters ≠ More Rides

While it can be easy for central operations teams from afar to assume over-supply is driven by a desire to minimize the number of deployment locations — why stop at two stations if you can stop at one? — we do not think that is the case³. Instead, we believe the issue is more insidious and driven by the metrics operations teams are using to measure themselves.

Metric traps can be hard to recognize, especially at the operational level, because they are often deeply hidden by the same metric-driven dashboards and KPIs operators rely on to track performance and identify issues.

When we see data-driven operations teams struggling with high levels of ride cannibalization, the root cause often lies in one of two places:

1. The overuse of heatmaps to inform deployment strategy (we explored this issue in a previous post)
2. The use of metrics such as Rides per Deployment (RPD) that look only at the impact of interventions on deployed vehicles.

Let’s look at an example: You are a market manager who needs to choose between placing a vehicle at Station A or Station B. Looking only at RDP, which station would you choose?

The current industry standard, Rides per Deployment, measures the number of rides a deployed vehicle receives over a set time period after deployment (typically 24 hours). It is a good proxy for the direct revenue captured by a deployment; however, RPD is limited in its use as a proxy for profitability because it fails to account for the cost deployments have on nearby vehicles.

When teams are unable to understand the impact their actions have on other vehicles in the fleet, they end up inadvertently making unprofitable decisions that cost operators time and money and ultimately reduce the quality of service for riders.

While Station A may seem like the obvious choice in the example above, in the next section, we’ll show why this may not be the case once ride cannibalization is considered and introduce a better metric that teams can use to assess the impact of their deployments.

Introducing Net RPD — A Better Way to Measure Deployment Profitability

This is why at Zoba, we recommend operators use Net Rides Per Deployment(nRPD)⁴ to assess deployment profitability. A new metric we’ve developed with our customers, nRPD, measures the rides a vehicle captures after deployment that would not have been served by existing vehicles.

Unlike RPD, nRPD accounts for the cost of ride cannibalization, providing operators with a far clearer view of the positive or negative impact specific deployments have on total ridership and revenue.

The impact that this shift can have on an operator’s business can not be understated — when leveraged effectively, nRDP-based deployment strategies increase total ridership while reducing variable costs at no additional cost to the operator.

Looking back at our example, you can see how the picture changes when the effect of ride cannibalization is considered. Now including nRPD, which station would you choose now?

By adopting nRPD, market operations teams can:

1. Measure the impact their work has on revenue growth.
2. Optimize deployment decisions for profitability.
3. Minimize time spent conducting unproductive tasks.

While this post focuses on the need for better metrics for assessing deployment profitability, in future posts, we will dive deeper into the specifics of how we calculate nRPD at Zoba and provide some examples and tools for operators looking to start calculating nRPD for themselves.

In the meantime, if you’re interested in learning more about nRPD and how Zoba’s decision-automation platform can immediately begin helping your team increase revenue and improve margins — click here to schedule a demo.

¹ At Zoba, we believe it is just not possible to intuit your way to knowing the optimal distribution of, say, over a thousand vehicles no matter how well you know a city. Throw in temporal variance (e.g. weather impacts, weekend-weekday variation) and you are faced with a problem set humans were simply not built to solve. Helping operators overcome this problem is why we founded Zoba.

² In competitive markets, operators sometimes cluster vehicles near other operators. This exacerbates the issue.

³We speak with operators in hundreds of markets every month and are continually impressed with the dedication and passion local operators have for the mission and vision of micromobility. At Zoba, we believe no one in micromobility works harder nor is doing more to move this industry forward. It’s why we’re committed to helping them make better, faster decisions by developing solutions that provide operators with recommendations unique to their market conditions and personalized to their goals.

⁴Alternatively, this metric could be called Variable Net Rides Per Deployment (vRDP) or Incremental Net Rides Per Deployment (iRPD).