The course of their growth and extension never did run smooth.
In March 2016, much to the consternation of the investment community, The New York Times declared “the end of the on-demand dream”, stating, “across a variety of on-demand apps, prices are rising, service is declining, business models are shifting, and in some cases, companies are closing down”.1 The debate on the future of the on-demand gig economy has continued since, particularly focusing on the platforms which drive it, the feasibility of their extensions, and the justification of their valuations. By 2019, a study of 250 platform companies over the past two decades reported that more than 80 percent of them had failed, with their average lifespan edging five years and many collapsing by their third year due to a lack of users or funding, or both.2
The excitement around platforms has led to the inevitable rush to ‘platformise’ everything. The valuations of these digital platforms have often skyrocketed, based on the belief that network benefits will accrue and they will be able to extend themselves across different market opportunities. However, the reality appears to be quite different. Many platform companies have not really succeeded in expanding their offerings, a fact which becomes strikingly apparent in the case of what I refer to as ‘physical service platforms’. These are companies that provide physical, human-based services, where if not the full service, at least a part (particularly, the ‘last mile’ act of the service), is fulfilled by a human being and mostly on site—such as dropping off a parcel at the door, chauffeuring a passenger to a destination, or cleaning the windows of a premises.
Physical service platform delivery is distinct since such services cannot be provided otherwise, either digitally (by a bot) or remotely (by a call-centre agent). The rise of these physical service platforms has led to the development of an ‘on-demand’ app ecosystem or the gig economy. Uber is the classic example, but there are many more spanning a plethora of offerings, including travel and transportation, food delivery, valet parking, healthcare, and professional services.
The declining cost curve of high-volume, product-based businesses leads to greater profitability for market leaders. This idea took centre stage beginning with the Profit Impact Market Strategy (PIMS) studies of the early 1970s, which led to the famous BCG model or the early versions of the GE strategy model being applied across companies making and selling products. However, when service-based businesses, including physical services, were analysed, we often found rising (instead of falling) cost curves. Unfortunately, with increasing talent requirements, the quasi-normal talent distribution forces service providers to relax the selection criteria, which leads to lower skilled talent being hired, and the consequent lowering of overall service delivery and quality. Such insights caused many scholars to question whether product-based strategies could be applied wholesale to areas like service marketing and management. It seems that we are back at the same crossroads, trying to force-fit our experiences from product-based platforms to service-based platforms—in a sense, creating a new ‘Procrustean bed’.3
What I would like to bring to this debate is an analysis of why physical service platforms, despite their huge acceptance, are not doing as well compared to their digital product cousins, and what firms can do to manage the challenges that accompany the efforts to extend and grow them.
Challenges facing physical service platform companies
Unlike their digital counterparts, physical service platforms derive limited benefits from the core tenets of network economics, which include network effects (increasing value by increasing the number of users), platform stickiness, and user loyalty.
RESTRICTED GAINS FROM NETWORK EXTERNALITIES
In a completely digital platform business like Windows or iOS, the cost of delivery is minimised as firms perform more and more deliveries. Even for a platform like Amazon, which involves physical delivery of products, the firm can optimise such deliveries, since the same delivery service extends across different products that can be stored and stocked in the same way. However, when a firm is providing a human-based service, it cannot do any of these things.
Additionally, digital platforms are one-to-many (like YouTube where the same content can be consumed by multiple users at the same time) or many-to-one (like Amazon, where multiple products/vendors can supply to the same user simultaneously). However, physical service networks are essentially one-to-one networks, which means that a particular service personnel like a delivery person or a caregiver can serve only one customer at a time. Consequently, the benefits from network effects are limited.
IMBALANCE BETWEEN USER VALUE AND PLATFORM PERFORMANCE
Given the absence of unit-one costs and limited network effects, physical service platforms are compelled to squeeze service providers to enhance their business results. At some stage, this will no longer be an attractive option for the service providers, especially if they were initially incentivised to come on board. As a result, the firm loses the better service providers, and service quality and consistency begin to suffer. Additionally, it starts to lose the economies of scale in hiring/purchasing, training, and delivering the service. To top it all, the platform also starts charging higher prices, making it a less attractive proposition for customers.
Take Luxe for example. It offered a very interesting proposition when it started in the U.S.: “Fall in love with parking”. It became popular as the valet service was quick and economical. However, as the business started growing, the quality of the service began to deteriorate and the cost of providing the service started to rise. There came a point when it flipped over and became unattractive to both car owners and valets. Hence, this logic of squeezing one side or overcharging the other side is not extendable. There is a limit.
PLATFORM LEAKAGE, DISINTERMEDIATION, OR CIRCUMVENTION
It occurs when both sides have established a relationship and developed sufficient trust to bypass the platform and match themselves off-platforms. Disintermediation is perhaps the biggest challenge that physical service platforms face, since it happens with their highest value customers and has serious implications for business results. In the case of a physical goods service provider, say Amazon, there is little reason for customers to go to the actual merchants, because Amazon can provide the levels of assortment and selection that they seek, while receiving and passing on large discounts to them. The vendors also gain from associating with Amazon, since it gives them volumes while taking care of the pain of physical distribution too. However, typically for every physical service delivery platform, the actual service provider and the service consumer can very easily have a side deal with no impact on service, but there will be a significant price impact on both.
Urban Company (formerly ‘UrbanClap’) is an Indian gig marketplace founded in 2014 that offers home maintenance and repair, along with salon and massage services. The platform has made multiple attempts to foray beyond beauty into wellness by offering personal training, yoga classes, and diet instruction. However, each of these forays has failed. Imagine having found a trainer one is satisfied and wanting to continue with—what is the incentive to maintain the relationship with the platform? In contrast, Amazon’s ‘subscribe and save’ strategy actually helps consumers move many of their fringe purchases like detergent powder or diapers to the platform.
LACK OF PLATFORM LOYALTY
Physical service platforms such as Uber and Zomato frequently offer discounts to lure and retain their customers. When they cannot sustain it anymore, or even before that, someone else comes up with a discount. Hence, they are unable to create much customer stickiness. This also lowers loyalty, leading to consumers’ multi-homing across competitive platforms.
How digital product platform companies extended their presence
Success in platform extension—whether for physical service or digital product platform providers—does not depend on network effects alone, because networks do not seamlessly transition from one application or use case to another. Instead, a study of digital platforms provides insights on critical factors enabling network extension.
DEVELOPING THE TECHNOLOGY CORE
Google, Amazon, and Apple are examples of companies that have managed to extend across markets; these platform extensions were not just due to the power of their networks, but rather because they created a technology-based core product or competence that offered them access to newer opportunities. When Amazon was a bookseller, it built enormous ‘physical + digital distribution’ muscle that was the stepping-stone to becoming a marketplace for all products and services.
Google has grown based on a very clear understanding of how to use data to arrive at insights. Google search is not an intellectual evaluation of which website has better content, nor was it based on how often a search term appeared on a webpage (as was the case then). Instead, it was built on proprietary PageRank technology that determined a website’s relevance based on the number of pages, along with the importance of the pages, which linked back to the original site. From then on, organising and interpreting data ended up being Google’s core capability, which then became platformised as a business model, providing it with the opportunity to create multiple businesses beyond just web-based search to the mobile platform, location-based services, videos, and many other offerings.
Similar to Google, it is this preoccupation with providing an unparalleled user experience through integrated hardware and software that has allowed Apple to constantly push the boundaries of devices and use cases for best results. In the process, Apple has entered multiple new areas, and upended many industries and existing business models to become the most valuable company in the world.
Successful digital platform companies have extended across different markets not just due to their networks but because of the capabilities they built over and above their technology core, which they were then able to use to develop other use cases for existing users who gained value from these new ways to utilise the platform.
COLLABORATING TO SUPPLEMENT CAPABILITIES
In most cases, digital platforms extended into new markets not just by using their existing core, but by further enhancing it with additional capabilities. Amazon built AWS (Amazon Web Services), its hugely successful cloud business, on the foundation of its own requirements for a scalable technology infrastructure. Such additional capabilities may be built over a period of time, but more often than not, they are acquired through partnerships and alliances. Netflix collaborated with the United States Postal Service to home-deliver DVDs, in the process setting up the disruption of the home entertainment industry. Google created the Open Handset Alliance (OHA) and got HTC as a key member to set itself up in the mobile space. Apple partnered with multiple app developers to develop the iPhone user experience, as did Sony when it moved from CE (consumer electronics) devices to gaming with the PlayStation. And when Facebook felt threatened by new photo-sharing and messaging apps, it promptly acquired the leaders, WhatsApp and Instagram.
India’s telecom provider Jio is another interesting case study, as it has managed to extend beyond the traditional model of just being a conduit for connectivity through a series of carefully selected investors and strategic partners. Facebook was one of Jio’s first investors, and its ownership of WhatsApp and Instagram gave Jio the leverage into consumer apps, where its earlier attempts (like JioMeet and JioChat) had failed. Similarly, Google’s entry as a strategic investor gave Jio a play at the device and OS (operating system) layer, while its partnership with Microsoft provided additional inroads into the web services and cloud layer (along with Google Cloud) and enabled it to have a competitive position against Amazon. Jio’s collaboration spree did not stop there. It has gone on to build a similar relationship with Qualcomm, giving it a strategic foothold at the core technology infrastructure level. The slew of investments exceeding US$20 billion has not only helped Jio rid its balance sheet of huge debts that it incurred while developing a comprehensive platform offering, but has also given it the technological heft to extend and take on the might of both Walmart and Amazon in the retail space.
In fact, it would be fair to say that most extensions in the digital platform world emerge from strategic alliances and partnerships. What then behoves the on-demand service delivery platforms to believe that they can seamlessly extend to other functionalities? As the above-mentioned failed experiences have proven, it is not that simple.
What can physical service platforms do?
Given that there is no exponential growth and profit in the physical service platform business, the revenue is never going to come at zero cost. Therefore, the only way ahead is to grow the number of users and red ce the cost of service delivery. However, as I have higlighted, growth is limited due to a lack of extendability and you end up squeezing your provider to lower costs. This pulls the platform into a different competitive ar na where it is competing as a physical arbitrator and/or se vice contractor. So what can the platform do?
INVEST IN CORE TECHNOLOGICAL CAPABILITY
Physical service platform firms need to identify their core competence, and then invest in building it up further. Their energy should not be exhausted on blindly growing the numbers but should instead be channelled towards growing the capability to do so.
At the core, we have an issue with how to deliver a physical service on a digital platform. Unlike a product, a service is intangible, inseparable, variable, and perishable– qualities which make it difficult to monitor and offer service delivery guarantees. On the flip side, the platform continues to funnel resources into building physical capacity by onboarding and enabling more service partners. What remains equally, if not more important, is for the platform to develop suitable technology tools to train and equip vendors while also monitoring their quality of service delivery. Unfortunately, this is challenging to most firms. For example, even after a decade of operations, Uber has no means of letting riders find out about car and driver quality beyond the superficial ratings.
PROVIDE ENTIRE SOLUTION STACK
When the company’s energy is spent more on the non-technology, non-platform side, fungibility becomes a big challenge. Take the example of Urban Company again. It started by providing basic massage and salon services, for which it had to bring in one set of partners–masseurs and beauticians–and define the specific delivery criteria. However, when it extended from these to home cleaning services, a completely different set of partners came in, thus requiring a completely different pricing structure and service delivery timeline. Whereas a masseur or beautician can individually take multiple calls per day, home cleaning typically requires a group of personnel who end up taking the entire day to provide a single service. In fact, there is no similarity in delivery across the multiple offerings Urban Company has, and it is not really technology that is driving the business as much as its ability to bring in different service providers. Urban Company should thus invest in identifying the quality parameters and create norms across the different service delivery models. They are not technological capabilities that allow easy transition from one use case to the other.
Instead, like Subway, physical service platforms need to ‘productise’ their services. Subway produces high volumes of customised sub sandwiches and meals with great variety, yet uses very few raw materials to do so. This is a classic case of mass customisation, which requires product modulation (i.e., breaking down the sub into its key ingredients like bread, fillings, and sauce), process modulation (i.e., breaking down the process of sandwich- making into clear and simple steps), and finally a means of integrating both in a seamless manner. The Subway model does not extend to pasta. However, Singapore’s zi char (Chinese home-style dishes cooked to order) brand Wok Hey has similarly created a classic mass customisation model using stir-fried noodles and India’s Barbeque Nation chain provides a wide range of appetisers through an over-the-table live-grill format. If any of these chains were to extend their offering to a different cuisine (as they have often tried but failed), it would require an entirely different combination of product and process modulation, and their consequent integration.
DELIVER END-TO-END SERVICE
Some may assume that because of its success, Uber’s model can be simply forklifted into another domain. It cannot. Instead, what businesses need help with is the polar opposite: not the forklifting, but the unpacking of Uber to identify the two to three most applicable elements from its operating model and growth execution that can generate opportunities in its particular industry.4 At the same time, there is a need to be logical in terms of deciding what layers can be added. Uber teamed up with Cargo in the U.S. to make it easy for drivers to sell snacks to their passengers, and with Kellogg’s in India to provide commuters with breakfast on the go. I did not find it surprising that both efforts failed.
Physical service platform companies need to etch the importance of the service design factor onto their DNA, specifically how different elements of one service do not necessarily extend across others. Ola in India has a fabulous navigation system for its ride-hailing services, but this fell apart when it attempted to branch into food delivery. The same goes for many other gig economy ventures. Each service has a completely different design, which needs to be drawn up holistically. Rather than trying to do it all by themselves, physical service platforms need to start thinking like Alibaba and Tencent. These Chinese giants today are less of individual platforms, but have instead grown into extensive competing ecosystems with their proprietary apps and payment systems, while at the same time partnering with critical applications and service providers. Such ecosystems have their advantages since they ensure quality control while continuously tracking customer movement and behaviour. At the same time, they are able to build Artificial Intelligence and machine learning tools which help increase platform relevance. Although one can question the resulting privacy concerns, the use of advanced technologies provides a phenomenal value-add for the platform by ensuring service delivery across offerings.
ENSURE STICKINESS
Unlike Uber, Ola and other physical service platforms, ecosystems like Alibaba have enabled seamless transition from one service to another, ensuring that each is not a separate customer journey. This calls for placing the customer journey at the core of platform design and making the logic of customer experience (rather than network and access) central to determining platform offerings. The many failed efforts and humongous amount of wasted capital suggest that often, this has not been the case. Physical service platforms are not going to be able to extend into other delivery models unless they keep the customer experience central to their plans. Each service involves not just different vendors, but also different user expectations, thus creating more complexity. Urban Company cannot hope that just because it has an app that home users love, it will be able to satisfy every home maintenance need with the same level of detail. The requirements for getting sofas upholstered are quite different from those for getting a manicure!
DON’T FORGET THE HUMAN FACTOR
The moment we get into service-led businesses and platforms, we are dealing with marketplaces that require humans to fulfil some specific job requirements. Consequently, various factors that are irrelevant when dealing with inanimate products take on primary importance when dealing with people, and these can have serious implications. Imagine being stranded and missing your flight because the Uber driver cancelled the trip, after making you wait for 30 minutes! Service delivery platforms thus need to plan and provide for this uncertainty while designing their offering.
The future of physical service platforms
Upwork was created in 2014 through the merger of the two largest freelancing websites, Elance.com and oDesk.com. Elance was known for matching creative professionals, and oDesk did the same for technical jobs like software development and programming. Today, despite being the largest freelancer marketplace in the world and having a two-decade-long head start, the company is fighting with hundreds of competitors and still draws more than two-thirds of its projects from creative and software development, areas that correspond to Elance and oDesk’s original strengths respectively. Upwork’s arduous journey towards broader acceptability and profitability is perhaps a grim testament of the travails of a service platform.
To summarise, physical service platform firms need to temper their expectations. Setting an incorrect goal can lead to a cascading downward spiral, as the recent implosion of WeWork demonstrates. We have long acknowledged that strategic management of a service business is different from that of a product business. The same difference holds and is amplified across digital product and physical service platforms. Decisions regarding service delivery and customer journey are critical, and the overall platform strategy needs to be framed in this context, rather than in general terms. It calls for laboured execution through detailing, capability- building, and extensive collaborations. In the words of the controversial yet popular American pastor John Gray, “Everybody wants the platform but nobody wants the process.” The path towards the growth and extension for physical service platforms will remain long and arduous, and such companies would be well-advised to keep in mind the considerations outlined in this article as they draw up their plans.
Prakash Bagri
is Associate Dean – Corporate Engagements and Clinical Associate Professor of Marketing at the Indian School of Business
Endnotes
1. Farhad Manjoo, “The Uber Model, It Turns out, Doesn’t Translate”, The New York Times, March 23, 2016.
2. Michael A. Cusumano, Annabelle Gawer and David B. Yoffie, “The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power”, New York: Harper Business, 2019.
3. A ‘Procustean bed’ refers to a scheme or pattern into which someone or something is arbitrarily forced.
4. Sameer Patel, “In Defense of the Uber-ization of Everything”, TechCrunch, June 20, 2016.