​Are we overestimating the benefits of predictive analytics in the ridesharing industry?

Ridesharing companies are optimistic about the value of predictive analytics and its impact on their business, we dive into the repercussions of using predictive analytics in ridesharing

18Mar

The ridesharing industry has been looking for more innovative ways to compete against taxi companies and other traditional forms of transportation and data analytics has helped them earn a strong edge against more established competitors. New advances in predictive analytics have played an especially important role.

However, some experts may have put too much faith in the benefits of predictive analytics for this industry. We need to dig deeper to understand the benefits and limitations of predictive analytics for ridesharing companies trying to disrupt the market for transportation services.

Ridesharing companies are optimistic about the value of predictive analytics

Numerous ridesharing companies have stated that predictive analytics algorithms will play an important role in the future of their business models. One of the companies using big data to help them is Arity’s.

There are a number of ways that big data can be applied to improve the quality of ridesharing services. Some of these benefits include the following.

Predicting demand for future services and finding ways to meet it

Ridesharing companies tend to rely on contractors through the gig economy, rather than traditional employees. This somewhat limits their autonomy when it comes to setting shift schedules. Nonetheless, they can still try to control the number of drivers at a given time through financial incentives. The issue is that they need to know when demand is likely to be strongest.

As McKinsey has outlined, ridesharing companies can now use clustering data from different major cities to draw better conclusions about customer demand under certain conditions. This will help them decide when to offer better financial incentives to their drivers to encourage more of them to work.

Setting better customer prices

Setting prices is a very important part of operating a successful ridesharing company. According to The Definitive Guide to Uber, there are many factors that go into setting prices, including the distance of the trip, unanticipated events and whether the passenger changed their destination during the ride. Companies like Uber frequently change prices to reflect demand for their services and availability of drivers.

Preparing for the possibility that customer rebates will be needed

Weather patterns, major events and local trends can all make it difficult for ridesharing companies to meet customer demand. Ridesharing companies might know in advance that they will not possibly be able to have enough drivers on hand to meet the need for their services.

Predictive analytics algorithms make it easier for them to make these forecasts. These companies will be able to notify customers in advance of anticipated challenges that could bottleneck their ability to give patrons the service they usually guarantee. They can also budget for rebates to customers that they were not able to serve during this time.

What are the limitations of predictive analytics for the ridesharing industry?

There are clear benefits of using predictive analytics to optimize business models for ridesharing companies. However, ridesharing companies might be overestimating the viability of their big data solutions. Here are some limitations that they must be aware of.

Results could be tainted by data availability bias

Most customers are willing to provide data to ridesharing companies in return for higher quality service. At first glance, this sounds like positive news for the ridesharing industry. However, this means that a third of customers are not willing to share data through ridesharing apps. This could bias their data in a terrible way. It is possible that customers that don’t agree to share it have very different customer expectations. Companies that over-rely on this data could unwittingly alienate customers that prefer a different user experience if there a clear distinction in the homogeneity of customers that are willing to share data and those that are not.

Data might be less useful for dealing with customers in the suburbs

Ridesharing companies tend to operate in larger cities. A growing number of them are starting to expand into the suburbs around metropolitan areas.

The problem is that their customers still tend to be clustered in major urban districts. Ridesharing providers that rely on data might not realize that they are developing predictive analytics forecasts for suburban customers based on data from people living in more urban areas. There are clear behavioral differences between the two, but they might not be factoring for them properly.

Data on less precedented events might be unavailable or too limited

Predictive analytics models can be very reliable for ridesharing companies during normal conditions. However, they might not be so useful during major events, such as the Super Bowl. It takes time to collect enough data to make the kinds of forecasts that ridesharing operators need.

New regulations could make certain forecasts useless

One of the biggest benefits of predictive analytics models for ridesharing companies is the ability to set high prices to align with future demand. Unfortunately, that benefit my cease to exist in the future. A growing number of jurisdictions are considering laws to prevent price gouging.

The intersection of innovation performance and culture small

Read next:

The intersection of innovation, performance and culture

i