Rumors of the on-demand economy collapse were recently rivaled by Apple’s plans to invest $1 billion in Chinese ride-hailing platform, Didi. The funding goes against declining VC figures, as the tech giant makes strides to improve its prospects in the Chinese market.
Apple’s investment shows there is hope for the on-demand model yet, despite a recent slew of wounded unicorns and failed startups, forced to sell-up or pivot. The “on-demand apocalypse” – a term coined by Techcrunch reporter Josh Constine – marks a turning point for the this business model. And from the demise of so many, an evolved creature has emerged.
The survivors of the collapse of ‘Uber for X’ have learned from the fates of their fallen peers. This technology is no longer simply a middle-man, providing low-paid work to an unskilled workforce. New models recognize employee rights, make use of big data and machine learning, and represent an evolved offering. So, what have we learned, and what can we expect for the future of on-demand platforms?
Failures Of The Shared Economy
In 2015, investment into the sharing economy-styled startups reached new heights, with a plethora of new businesses helping to transport you, your belongings, and even your Granny from A to B. This model of employment has transformed many industries, providing an on-tap supply of workers, at a lower price, with little risk for consumers or businesses. By the end of the year, however, this new and shiny concept had begun to lose its sparkle, and startups failed to live up to initial buzz and over-optimistic valuations.
In April, Shuddle, an “Uber for kids” transport service folded after just two years and $12 million in funding, as competitors drove low rates and unprofitability for the business. A fundamentally niche and limited product, that had ridden the wave of Uber success, in reality failed to provide necessary growth.
Shuddle’s fate echoes that of SpoonRocket, the on-demand meal delivery service that was forced to sell up due to insufficient financing. SpoonRocket transitioned its customers to competitor Sprig, a more expensive service, focused on higher quality meals. In many cases, unprofitable business models and competition have driven low rates, this has been met by contractor demands, requesting improved payments and employee benefits.
Last year, Homejoy’s home cleaning service crashed under worker lawsuits — with contractors demanding the same rights as employees. Instacart is currently experiencing a similar employee dispute, and has also been forced to drop prices to remain profitable.
A common trend between failed platforms was relatively easy access to initial funding for models that, later down the line, struggled with scale, margins, and employee satisfaction. As Wired reporter Davey Alba said, “the nature of on-demand necessitates a high volume of transactions on low margins.” In order to protect themselves financially some are deserting this model, for instance on-demand parking valet apps Luxe, Zirx and Valet Anywhere, now opting for the security of monthly subscriptions.
Those that have succeeded with the on-demand model have focused on employee empowerment, using big data to connect skilled workers for fairer rates. As Alba reports, this is not really a sharing economy – it is service, and this calls for an evolved tool, prioritizing quality and trust.
Evolved On-Demand Business Models
A rise of workers finding new employment through shift work and online platforms is replacing the traditional workplace structure. As Bloomberg columnist Justin Fox put it, “Jobs are out. Gigs are in.” According to software company Intuit, by 2020, so-called “contingent employees” will comprise over 40 percent of U.S. employment.
However, in a saturated market startups need to re-evaluate just how scalable their “Uber for X” is. The original taxi-hailing service has also diversified, now offering UberEats, a courier delivery service, and even a special “ugly Christmas sweater” temporary delivery service.
The on-demand graveyard also shows just how important it is to retain employees, and the survivors reveal a growing trend of empowerment, in a service-focused industry.
Homejoy’s former New York City GM Katie Shea, cofounded on-demand home cleaning service Slate. Slate started off a decade ago as a laundry delivery app. Fast Company described its primary focus as “customer convenience”. Slate marks a trend of new on-demand startups that are service first, and tech second, in contrast to failed businesses like Homejoy.
In New York City, new platform Coopify, currently in development, aims to connect popular home cleaning and caregiving co-operatives with consumers. The co-op structure means employees are owners of the platform. Workers define their own rates and working conditions. They will be able to receive payments, work with peers to manage their schedules and also access client ratings. The co-op style business reportedly relies heavily on word-of-mouth, and recommendations.
Recommendations mean trust, and this is something that on-demand platforms have started to realize. Techcrunch reported the emergence of the “trust economy” in the evolution of on-demand businesses.
Beyond the basic 5 star review, the Internet offers a growing online resume. Sites such as LinkedIn allow employees to build a history, to flaunt their achievements and third-party recommendations provide a level of transparency and security that has previously been lacking. Today employees have identities, histories, and begin to build and showcase their own careers.
This concept has sparked renewed interest in waning investors. In April, identity verification startup Onfido raised $25 million to grow its service, helping on-demand businesses to conduct background checks on potential employees using machine learning algorithms.
Machine Learning Predictions Augment The On-Demand Offering
The emergence of machine learning in developing on-demand platforms shows how a previously binary matching system can be enhanced with big data.
A recent example of this is AirBnb’s new pricing recommendations; using local event data and hotel availability to suggest smart rates for listings. At the unveiling of this new feature VP of engineering Mike Curtis reported hosts that set prices in line with suggestions are four times more likely to receive bookings.
Using big data to understand demand is also a tool utilized by Uber, who adjust rates based on peak travel times. As an extension of this on-demand courier service Rapidus collects driver data to understand their routines, suggesting pick-ups that match their schedule. This helps the Californian delivery service to reduce costs, using drivers that may already be traveling that way.
Pre-made meal delivery service Maple serves 800 meals an hour from four kitchens across New York City. It uses historical demand and menu popularity to estimate kitchen workloads, and divides orders between drivers travelling optimized routes which also take into account waiting times, and location proximity.
Combining supply data and consumer specified parameters new businesses can better match this to the demand. This type of technology can be used to ensure that the more skilled or experienced workers are paid higher rates, from customers that are willing to make the investment.
This requires matching up numerous variables of data — location, ability, qualifications, experience, availability, matched against the demand parameters of project complexity, pricing ranges and more. Creating audiences, noting trends and suggesting pricing – something comparable to programmatic advertising algorithms, eBay buying tools, or even the tagging and recommendation selection process that gave birth to House of Cards. Combine this with the transparency of the emerging trust economy, and you have far more intelligent tool, and happier workers.
This will be the next generation of on-demand growth. Much more than simply “Uber for X”, on-demand is is maturing. And now it’s sacrificed a handful of startups, it’s just morphing into something a little more refined, evolving to suit a growing workforce of freelance employees.