Unless you’re living under a rock or completely off social media – which is the same thing – you’ll have seen a lot of tweets, updates and posts expressing frustration over the declining service levels of app-based cabs, particularly Uber. The internet universe is rife with angst. People are upset being stood up, being refused destinations, and having to pay extra for air-conditioning, so on and so forth.
It’s natural. Uber (and similar algorithmic cab-hailing services like Ola, Lyft, and Grab) came as a breath of fresh air to Indian users who were hassled by the thought of traveling by autorickshaws and cabs.
On the face of it, the offline style is simple enough. You must wave down an empty cab or rickshaw, negotiate the destination and price (in case you don’t have a meter regime like Mumbai), and that’s it. It’s not as if these methods are problem-free; getting a free rickshaw or cab during peak hour is never easy, negotiating for the price and the destination is another painstaking process altogether.
Uber set expectations at the get-go. It was sleek, modern and such an easy way to book yourself a ride. The initial incentives were high enough to ensure high quality of services – tidy cabs, predictable fares, digital payments, and no destination hassles – the algorithm did most of the dirty work for you. Drivers too, earned a decent living as compared to the pittance they earned as casual drivers for hire or in the old-style cabs. Even if he or she had to take a loan to buy a car, it worked out.
The Indian commuter seemed set. (I’m not talking of those who owned vehicles or used public transport – that market stayed pretty much as-is).
However, things do not seem ideal now. What may seem to a user as a service problem is, perhaps, just the market responding to incentives.
It’s the economy, stupid.
Most users of LinkedIn are loath to discuss what they consider as non-business topics, but businesses operate in the real world, where they’re affected by politics, culture, and the socio-economic environment they’re part of.
While the impact of rising oil prices is well-known (CNG has risen 37% in the past six months) what we often miss is the socio-economic disparity between customers and drivers. Over the past two years, while the salaried class has been relatively cocooned, many drivers have had to compromise greatly on incomes, particularly during the hard lockdown. 40% of drivers are migrants (as per pre-pandemic research) and are the sole breadwinners in their families (20%).
Drivers’ revenues were gradually reducing even before the pandemic. During launch, Uber’s driver recruitment programs were marketed well, with field agents accosting drivers at airports and public places and inviting them to presentations at five-star hotels. For drivers who earned an uncertain income (around Rs 12,000 to Rs 20,000 a month) a “gig” partnership, a seemingly high-tech algorithm, and incomes upward of Rs 60,000 were hard to refuse. But by 2018, Uber had increased its commission from the initial 10% to 35%, reduced its fare per kilometer, and implemented an entry fee, even as fuel prices were rising. The entry fee of Rs 5,000 is often too expensive for low-income aspirants.
The median income for an Uber driver in 2022 is reported to be Rs 25,000 a month. It isn’t hard to work out the math of the income disparity between the average Uber consumer and the driver.
Does this alter driver behaviour and service delivery? Quite obviously. The driver community will engage in practices that maximise their income and reduce their costs. These include, like users have reported on social media, trip cancellation, idling time to pick up, insisting on cash payment, negotiation on air-conditioning and other tactics we haven’t yet heard of.
As users, we feel short-changed. This is not what we pay for; this is not the service level we expect. We ask Uber to help, but that is just a band-aid on a waterfall – if there are no incentives, such tactics are likely to continue.
Algorithms follow the market
It takes Uber – or any platform app that’s based on AI – weeks if not months to calibrate, develop, test, and deploy algorithmic changes. It sometimes works on a change for several years before rolling it out. The rollout itself will be likely tested in select markets before it becomes universal.
But driver tactics – which we can call “the market” – does not take as much time to respond to the algorithm. Let’s just take two facets of Uber’s rules and examine how these incentivise tactical responses from drivers.
One of the most vocal complaints by users have been on cancellation by drivers. As of today, a driver cannot see the destination before he accepts a ride. There’s a reason for this – Uber would like to incentivise drivers to accept any destination that the app puts to them without exercising any discrimination. It’s ostensibly a good rule and creates a positive incentive for customers. Less ride refusals, quicker rides, better experience.
Uber has also created a disincentive for ride cancellation at both ends. If you cancel a ride after a driver is on his way to pick you up, you’re charged a cancellation fee, a part of which is paid to the driver. Likewise, while a driver cancels on you, there are greater disincentives which affect current and future incomes, his availability on the app and even disqualification.
In an ideal world, an Uber driver would never cancel a ride when the app is turned on. But it isn’t an ideal world – incentives have been warped, and a driver sometimes makes the decision to cancel despite the disincentives. Cancelling a ride may be more economical than traveling to a destination that is not profitable for him to go.
In turn, this causes aggravation at the user’s end. A user is unlikely to feel frustrated if the ride hadn’t been accepted at all; cancellation after acceptance means that the user loses time and has to make a repeated effort to finalise a ride.
Refining this tactic further, a driver adopts “idling” – a state at which he takes a long time to arrive at a pickup spot to induce a user-initiated cancellation and avoid his own penalties.
Could this inconvenience to the user have been avoided if Uber had transparent destination disclosure to the driver? Perhaps yes. They’re trialling it.
Customer is king, the queen, the prince and everything else.
Like in other platform apps, Uber brings a buyer and a seller together. And just like most platform apps, Uber is geared to favour the buyer – the user – and not the seller – the driver.
When it was introduced in the United States, Uber had a different idea. It wanted to break the heavily unionised taxi monopolies and allow idling car owners to earn income by matching a prospective rider with a car owner/driver with some time to spare. It termed these drivers “independent partners” and is steadfast on retaining that status and spare itself on minimum wage, worker safety and other regulatory compensations and protections. It’s not as if Uber does not have employees – all its “above-the-algorithm” workers have different compensation structures that are way more rewarding.
However, in India, the huge income disparity meant that there would be more drivers than car owners and that a huge percentage of these drivers will be from lower income groups. This made things easier for Uber and is likely to have been one of the reasons for its relative success as compared to the kind of opposition it has faced in cities like New York and London. Before the pandemic, drivers were plentiful and supply usually exceeded demand, which is likely to have given Uber the confidence to increase its incentive and reduce fares.
While the app is reasonably transparent to the user, it is relatively opaque to the driver. Destinations – the most important factor that would help a driver judge the economic benefit of a ride – is concealed. Penalties for cancellation are much more expensive for the driver, and digital payments are held for a few days before being paid out.
So how does the market respond to an unfair algorithm? By applying various tactics which we have discussed before.
What could Uber do?
The key to ensuring customer satisfaction lies in enabling transparency and decision making to the driver. The driver is the most important party in the platform; the experience of the brand and the app is directly linked to how the driver is interacting with the customer.
For starters, Uber could make its algorithm more transparent by allowing drivers to have a say in deciding incentives, destinations, penalties and other features of the app. Uber seems to have made a beginning by inviting selected drivers from across the country.
Another quick step would be to establish itself more emphatically as an intermediary between the driver and the user – which could even mean that the user-driver transaction is immediate, transparent and legal, and a driver does not have to insist on cash or off-app payment. There are other ways too – however, they must be driver-led, or at least driver-centric, to last.
Merely optimising customer care is not going to help.
Leave a Reply