South African #ecommerce: The (new) old hype of MSP (Multi Sided Platforms)

The recent announcement of Sagarmatha establishing a MSP (Multi Sided Platform) as a “first” in South Africa has drawn a large amount of criticism from people in the know. Novices (such as the Black Consumer Council) slated MSP as a progressive and disruptive move with the purpose of generating over 5,000 jobs, supporting ANC president Cyril Ramaphosa’s “YES Campaign” of economic growth in South Africa.

The simple definition of a multi sided platform (MSP) are really technologies, products or services that create value primarily by enabling direct interactions between two or more parties or participant groups.

In short: MSP’s are nothing new and precede the introduction of the “super-galactic highway” as recently slated by a number of prominent (but possibly ignorant) well-known industry players. With the above example, South Africa had MSPs for a very, very long time:

  • Your traditional shopping malls: retail stores and consumers
  • Bidorbuy.co.za has been bringing buyers, sellers and affiliates together since 1999. By extension any other marketplace such as eBay and liberally even classified sites such as Olx or Gumtree
  • Any home-sharing business bringing together owners and renters (Airbnb and by extension any property-site such as PrivateProperty.co.za)
  • Any drive-sharing business bringing together drivers and passengers (Uber, Lyft, Taxify)
  • Any social media site (Facebook, Twitter) bringing together users, advertisers, 3rd party content developers
  • Android OS bringing together handset manufacturers, SDK developers, application developers and users
  • Any gaming platform (Xbox, Playstation, Nintendo) bringing together gamers and developers
  • Any payment service (bidorbuy Payfast, Paypal) offering services to merchants and consumers
  • Any ticketing platform (Computicket) bringing together event organisers, event venues and audiences
  • Any professional industry platform (LinkedIn, Xing, Careerjunction) bringing together professionals, employers and recruiters

MSP as an e-commerce marketplace

Bidorbuy, Amazon or eBay are typical players in this segment. The value to customers on one side of the platform increases with the number of participating customers. This indirect network-effect is established as more sellers will be attracted due to the increasing number of buyers. In short: The more buyers a MSP has, the more sellers it will attract and vice-versa.

This “chicken-and-egg” scenario is a double-edged sword which creates a high barrier of entry for competitors. This is visible how large players such as eBay, Amazon, Alibaba, Rocket Internet compete with each other and how companies such as Flipkart benefit from high valuations and funding runs and others horribly failing (Naspers: Kalahari.com, Markafoni, SACamera, 5rooms, Kinderelo and Style36)

High switching costs are not necessarily a barrier of entry and this was visible when group-buying was “a thing” a few years ago where Groupon and LivingSocial were the early leaders of daily deals. By now, hardly anyone uses Groupon and I have not heard from LivingSocial in years (they used to have 70m active users in 2013). In short, an MSP such as Groupon could gain sudden market-dominance but then will fall on it’s knees as the indirect network-effect diminishes. In the daily deal scenario, buyers moved off the platform as they often felt “cheated” by daily deal offers and their sellers. This in turn challenged profits of sellers having to offer deep discounts in order to be featured on daily deal sites.

The most important aspect of any MSP is however the economy of scale. The upfront development of a marketplace will result in significant costs and many companies fail. Naspers for example has tried for a good decade to position themselves as a dominant e-commerce player in South Africa with Kalahari.com and eventually had to close doors. Subsequent to Kalahari’s closure, Naspers ploughed in excess of R2bn (USD 200m) into Takealot.com which after 3 years is still not breaking even and will probably need to see another R1bn funding run in the next 12 months.

When a new MSP in the form of Sagarmatha (a bundle of digital media, some ecommerce in the form of Loot.co.za and a list of prospective but unknown acquisitions) surfaces with a R50bn (USD 4bn) valuation and then attempts to raise a minimum of R3bn (USD 300m) via a private placement with a listing on the JSE, the industry wakes up.

While MSPs generally work very well (look at LinkedIn’s or Google Android OS diversified sources of revenue), there are many cases where it makes commercial sense to restrict the number of sides in a MSP model. The most prominent example would be Microsoft Windows with it’s biggest failure being the openness to OEMs and it’s subsequent move into hardware. There is hardly a Microsoft product (who remembers the Microsoft Zune?) which had a long-lasting success.

While I do hope that our new African unicorn in the form of Sagamartha is not just smoke-and-mirrors. Financial experts highlighted the large debt burden of the digitial media companies and the oversubscription of current net-asset value (R 0.34/share vs R39.62/share in private placement) and the R50bn valuation with negative cashflow.

South Africa certainly needs a new large e-commerce player, but it is very doubtful that Sagamartha will be able to pull this off as their JSE listing has now been postponed for a 2nd time.

MSP’s are certainly high-risk and high-reward – perhaps it is no coincidence that Sagarmatha takes it’s name from the Nepali “Mount Everest” as it is a tough mountain to climb, while others refer to it being a business venture bringing in a mountain of debt.

Functioning MSPs such as bidorbuy, eBay, Amazon or Alibaba are very defensible businesses as long as they offer evident value propositions for every side. It is naive to think that success for an MSP is a rule — it is more an exception as there are many examples of failing MSPs.

 

Inspiration and credit for the above, thanks to HBS Multi Sided Platforms.

Print Friendly, PDF & Email