[Also published on VCCircle]
There has recently been increased discussion, and mainstream press reporting, on the adoption of a ‘marketplace’ model (vs. an inventory model) by e-commerce companies (e.g. these two articles in Mint: Mint 1 and Mint 2). This discussion reflects an underlying presumption that one model is better than the other. In framing the issue as a comparison of the two approaches, I think the dialog fails to address the more important question of why this shift is taking place and whether there are other approaches that can address the underlying challenges.
The shift towards ‘marketplaces’ is taking place as companies try to find a new balance between the following priorities:
- Maximizing capital efficiency
- Maximizing customer delight (selection, post purchase experience etc), and
- Minimizing logistical complexity (which helps to maximize scalability)
The need to find a new balance is triggered by scarcity of capital. As long as capital was freely available, most ecommerce companies focused heavily on the customer experience, which was best served by an inventory model. As capital tightens, these companies must now balance the need to delight customers with the need to build a viable business.
What are marketplaces?
Let me start by defining what I believe to be true online marketplaces. These are platforms that enable a large, fragmented base of buyers and sellers to discover price and transact with one another in an environment that is efficient, transparent and trusted.
- Efficiency is a function of liquidity (enough buyers and sellers) and an effective price discovery mechanism (e.g. an auction).
- Transparency is ensured by applying the same set of rules to all participants, and because buyers and sellers know who they are dealing with.
- Trust is provided by features such as buyer and seller ratings, reviews, and integrity / guarantee of payment.
Marketplaces are difficult to execute against because they require adequate and simultaneous liquidity on the buyer and seller side. Once adequate liquidity has been established and the ‘flywheel is spinning’, these businesses exhibit strong network effects (because a market that has the most buyers will attract more sellers, and the increasing base of sellers will in turn attract more buyers). So once a marketplace becomes dominant, it scales organically and often exhibits ‘winner take all’ characteristics. Additionally, because marketplaces are essentially technology platforms that provide tools for buyers and sellers to participate and a trusted environment that facilitates price discovery and transactions (vs. actually being responsible for fulfilling transactions), they can scale very rapidly.
We’ve seen all of these dynamics play out at close range as a result of our investment in the Indian Energy Exchange (IEX; www.iexindia.com). IEX operates an electronic market for power in India and has emerged as the dominant power exchange in the country with deep liquidity.
The take-away is that when you get marketplace business models right, they are profitable, scalable, defensible and highly valued. Which is why contrasting the inventory model with a marketplace model makes for an exciting debate.
The inventory model
In India, there is no question that being in control of the product (i.e. having physical inventory) enables a superior post-purchase consumer experience. If you have the product in your control, then (assuming your systems and processes are robust) you: (i) have visibility into your stock level, (ii) know where the product is physically located, and (iii) control the pick, pack and ship process. This means that you minimize the likelihood of accepting an order only to later discover that you don’t have the product. It also means that you can optimize dispatch time. The bottom line is that being in control of the product enables you to deliver faster and with higher accuracy, and respond effectively to customer inquiries about shipping status. Given the correlation between delivery times and return rates that we’ve observed (i.e. long delivery times are clearly correlated with high return rates), this is really important.
The problem is that being in control of the product has meant that companies compromise capital efficiency – because they buy product from vendors up-front, thus tying up capital in inventory, while at the same time exposing themselves to inventory mark-down risk. This can get ugly – which is why it makes sense to explore other approaches, one of which is a marketplace model.
Marketplaces in ecommerce – how different are they really?
The reality is that most of the marketplace models we see in ecommerce are not ‘platforms’, as described earlier. For example, in ecommerce marketplaces the prices are fixed, not discovered, and the ecommerce company is responsible (from the customer’s perspective) for several aspects of the post-purchase experience, such as fulfillment and customer service. The reality is that to the customer, many of these marketplace companies look identical to inventory-led ecommerce businesses. In other words, these models are simply one possible response to the constraints and challenges of traditional inventory models. And the marketplace model is not without its downsides – for example shipping costs are higher because multi-product orders are fragmented across vendors and shipped separately. And this in turn may lead to customer dissonance because a customer won’t receive their entire order at one time.
There are other solutions [Note that for purposes of this discussion I am not considering FDI related implications on company structure.]
Other possible ways of mitigating capital intensity while remaining in control of the product include (but may not be limited to) vendor credit, consignment sales (where products are in the possession of the ecommerce company but are not paid for upfront) or back-to-back purchasing (where the ecommerce company places the order on a vendor/supplier after receiving an order from a consumer). For example, ASOS, a UK-based online lifestyle retailer, has net working capital of less than 2% of sales while operating an inventory model. Similarly, Shoppers Stop in India has a negative working capital model – again despite being an inventory-led business.
Focus on the substance, not the glossy headlines
This is a meaty and critical subject for any company involved in online commerce. We’re encouraging our companies to experiment with strategies that resolve the trade-offs outlined in this post because we think companies that successfully do so will have more attractive scale and economic characteristics over the long-term. The purpose of the post is not to take sides on the inventory vs. marketplace model debate or address the pros and cons of each approach in detail – rather it is simply an attempt to surface the underlying issues that are driving the evolution of how ecommerce companies operate in India.