Accounting For Cost Of Collections In LTV


This post first appeared on the VCCircle blog here. I have added some additional thoughts below based on some feedback and questions I got.


In an earlier post I had talked about a key difference between developed and developing economies being that startups in developed economies can buy products and services on a purely variable cost basis. In this post I am going to focus this observation on cost of collections and how it impacts life-time-value (LTV) of a customer. LTV and corresponding customer acquisition costs are critical to determine long term financial viability and strength of the business. The general method to calculate LTV is to have a model for lifetime revenues and subtract from it cost-of-good-sold (COGS) to come up with the LTV number. In economies like the US it is typical to not include collections costs in these calculation because regardless of the type of business (B2B, B2SMB, B2C) or transaction amount ($0.99 song on iTunes to $100K B2B billing) the collections costs is purely variable and in steady state is about 1-2% of revenues so it doesn’t really materially affect the LTV calculation. This has been made possible by the omnipresence of electronic payment methods where the marginal cost of closing a transaction is minuscule; a reliable and enforceable credit rating system for consumers and business; and finally a law enforcement framework which provides the necessary the carrots and sticks to consumers and businesses to follow through on contractual obligations.

This framework does not necessarily apply to developing economies, certainly not to India. Vast majority of the transactions in India are in cash with very little penetration or growth of electronic payment systems. As a result most collections in the B2C space are done at retail locations or using COD etc which in turn require follow up reconciliation and management. But the key characteristic is that there is a FIXED minimum cost per transaction after which the cost becomes variable. In other words:

  • For electronic payment methods, cost of collections = 2% of transaction amount
  • For cash collection methods, cost of collections = MAX of (Rs100, 1% of transaction amount)

This has a very meaningful impact on LTV calculations especially when transaction amounts are reduced. Let me illustrate this with a comparison between three scenarios, one where collections is purely variable cost, one where there is a fixed minimum but low number of transactions and one where there are high transactions.

As can be seen from the table, the 3rd scenario dramatically changes the complexion of the business even though in all three businesses the revenues and COGS are the same.

Getting rid of that minimum amount per transaction in India is non-trivial. As I mentioned earlier this will require electronic (automated) payment methods, which in turn require robust credit rating institutions and which in turn require reliable law enforcement. All this is beyond the scope of the startup ecosystem. At Canvera we focus heavily on high transaction amounts for cash transactions, provide incentives to make electronic payments, only go after business that we see annuity value in so that over a period of time we can optimize the cost structure, and try to collect advance deposits where possible. This is work in progress and after 3 years of investments in technology, processes and infrastructure we unfortunately still don’t have the scale and cost structure that a purely electronic payment network can provide. But we are acutely aware of these costs and factor them in our calculations.

Additional comments

After the post first appeared on the VC Circle blog I got quite a few questions and I’d like to address one of them around the minimum cost per transaction I talked about earlier. I’ll consider two separate cases: COD and Retail payments.

Minimum cost per transaction in COD

There are many sources of this minimum cost and these include:

  1. Cost for a person to show up at a residence/office is fixed and not dependent on the amount being transacted.
  2. Customers refusing the product when the delivery/cash collection person shows up or simply having to make multiple attempts at delivery/collections
  3. Fraud: for the economics to work the person delivering cannot make a lot of money. At the same time increasing their “throughput” means allowing them to handle larger and larger sums of money which in turn creates perverse incentives. And adding more checks and balances adds more cost. There are other sources of fraud that are inherent with cash handling and reconciliation even after the money has been safely delivered to an office.
  4. Cost of reconciliation and properly matching with the right order. This is a fixed cost independent of the transaction amount
  5. Fake currency: as the business scales this issues shows up very quickly – in fact in all of Canvera’s regional offices have fake currency detectors since there was non-trivial fraud
Minimum cost per transaction in Retail environment
Retail environment is much cheaper on a per transaction basis but does require the fixed upfront cost of the setup of the retail location.  The retail location does of course serve many purposes  (marketing, product display, sitting area, warehousing, customer service etc) over and above collections so it is difficult to tease out the collections cost. But the per-transaction cost is much cheaper than that of COD. In the case where say an FMCG company supplies to a kirana store, the store owner absorbs the cost of collections and a portion of the equivalent rental costs of the location need to be used to calculate the cost of collections. Of course for family owned property all these costs are never considered. For a brand that sets up its own retail locations, the cost of the property is shared with many other functions outside just collections and it is well known that the real estate costs are key to determine overall profitability of each individual store.

5 thoughts on “Accounting For Cost Of Collections In LTV

  1. Good point made here Dhiraj. I must say though that all the top flight ecommerce companies I have met are very conscious of both their cost of collections and their cost of logistics related to COD (over 40% returns in some businesses) and they are instituting processes to reduce their costs. Some of them are already rolling out their own logistics frameworks to reduce their per-pickup and per-delivery costs. In an economy with as much “black” money as we do, I am afraid COD will linger for a while.

    • Absolutely, I am sure each company HAS to track and minimize, but I also think there is a PER TRANSACTION Shannon lower bound (hey, some times I need to prove I have a EE background 🙂 ) to the costs as long as cash is involved. To push it further below requires a tectonic non-linear change in the operating environment as I detailed in the post or a complete switch to a leveraged offline model such as FMCG companies do with “kirana” stores (pushing the last mile costs of delivery and collections to the store owners and also increasing the per-transaction amount by doing bulk shipments to the stores).

      It sounds enticing simple that as these costs are reduced economics will start working, but there is a assumption of the fact that they can indeed be reduced within the same model to a largely variable cost structure.

      I do agree with you that for at least the next decade cash (and therefore COD for mail-order/e-commerce companies) will continue to rule, the question is with the LTV implications as I have laid down above what are the capital requirements for the operating costs of these businesses without a change in the model.

      And I won’t go in to customer mindset issues which is another huge barrier to cross.

  2. Agree with your point that people are under-estimating costs in these businesses and that scale may not mean much. There is atleast one evolving view I have heard that we are experiencing game theory now. Some of the leading ecommerce startups are raising ob$cene amounts of money – one use of funds is to deter competition and the other is to fund the quest for profitability. So in some ways you can build a category leader simply by dint of raising a lot of money first and not by virtue of superior execution. Time alone will tell.

  3. Pingback: Grokking India: High fixed cost per transaction « Biz Tech « – India Internet, mobile, consumer tech, business tech

  4. Pingback: Why I don’t believe in COD | In a startup frame of mind

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s