AGILEVC My idle thoughts on tech startups

March 29, 2010

Let me start by saying that I’m a huge proponent of the “lean” startup philosophy for web software. Many, if not most, Internet-enabled companies can build significant value with a bootstrapped approach or small amounts of capital, especially compared to 8-10 years ago (e.g. before cloud computing, full open source stacks, self-service monetization like Adwords, distribution via social networks, etc). One of my favorite companies is a mobile gaming startup that never raised a dime of outside capital and reached a multi-million revenue run rate less than one year from inception. And even if a company ultimately transitions to a “fat” approach in order to dominate a huge market opportunity, I believe operating lean in the early phases pays real dividends in terms of discipline and focus.

In my prior post regarding the debate between “fat” and “lean” startup approaches, I highlighted my experience at PayPal which was undoubtedly a fat startup. At our “peak” in terms of fat, in late 2000, the company had a net burn in excess of $10M per month. In part this was because PayPal was built in the Web 1.0 days, in part due to some missteps we made and some idiosyncrasies of the company’s formation. Yet the company ultimately achieved >$100M in revenue and IPO in about 3 1/2 yrs, almost all of this in the post tech buble era.

It’s my firm belief that PayPal could only ever have succeeded as a fat startup, for several reasons:

1) Two Sided Marketplaces are HARD. Payments systems are inherently two-sided marketplaces, whereby you need to build a critical mass of both buyers willing to use the system and sellers willing to accept that form of payment. Having a million buyers and zero sellers (or the converse) is basically useless… it’s a classic chicken and egg problem. At PayPal we spent tens of millions in marketing through our referral program (get a friend to join we’ll give you $5-10), and this incentive was a key driver for both seller and buyer adoption. Would it have been theoretically possible to get a critical mass of both buyers and sellers without this marketing expenditure? Conceivably, but it might have taken too long to maintain either sides’ interest. A payments service with 10,000 buyers, 100 sellers, and slow growth doesn’t serve anybody particularly well and typically won’t make it.

2) Our Marginal Cost Was NOT Zero. The beauty of many digital businesses is that they have near zero marginal costs… serving an additional customer is simply a matter of a tiny bit of extra computing capacity. Payment processing has real transaction costs, so as PayPal grew usage we had direct costs associated with incoming payments via credit card or bank transfer (1). In the early days of developing our revenue model, we had tons of transactions going thru the system but comparatively little revenue coming in the door. Additionally, when you handle peoples’ money they typically expect real customer service via phone & email not just FAQs. We tried doing this in Silicon Valley initially but it was neither scalable nor cost effective, so after an unsuccessful attempt at outsourcing to a company in India, PayPal eventually opened its own ops/CS center in Omaha, NE. So we had significant costs for operating the system, which had to be funded with investor capital rather than revenue until we reached meaningful scale.

3) We Faced Competition Unconstrained by Capital. There were countless companies working on online payment products in the late ’90s and early ’00s. Many of these were startups similar to PayPal, but we faced two very large competitors who were essentially unconstrained in their capital resources. I don’t mean theoretical competitors, as in “Google or Microsoft could theoretically try to do this someday”… I mean actual competitors open for business with seemingly bottomless pockets. Billpoint was a service which was a joint venture between eBay and Wells Fargo, though eBay ultimately took over the organization. In addition to tens of millions (2) in funding from these two large public companies, Billpoint obviously had direct integration into the platform and other benefits of being part of eBay. The other large competitor was C2it, an in-house effort in which Citibank reportedly invested nearly $200M in capital before shutting it down. We can have a debate about the efficiency & nimbleness a startup possesses relative to corporate giants, but clearly without funding to build and operate the business PayPal never could have competed in this emerging online payments sector.

So what if PayPal was being launched today? Could the company have operated more efficiently than it did back then? Probably. But I’m doubtful these challenges could have been overcome in a “lean” approach. Would a company like PayPal get funded in today’s VC environment? Unclear. We certainly benefited from raising capital towards the tail end of the boom times, not just in terms of a heady valuation but simply attracting that much capital at a comparatively early stage.

Again, I’m not suggesting that most startups operate with a “fat” philosophy. But clearly not every set of interesting market opportunities can be tackled with a pure “lean” startup approach. Arguably all three dynamics above still exist for any startups trying to create a new payments system today. Similar issues face anybody trying to build the next great search engine (need critical mass of both consumers & search advertisers, large scale search indexing isn’t free, and Google & Bing exist).

So I for one celebrate and embrace lean operating philosophies, yet recognize that some startups will transition from lean to fat and some (even software-based ones) probably can’t ever be built without taking a fat approach early on.

Notes:

(1) At the time of our IPO, PayPal had gross margins of roughly 50% on a ble
nded basis. But you can basically think of this as a large very low margin business coupled with large extremely high margin business. When buyers paid with a credit card, PayPal’s margins were <20%.>90%.

(2) Billpoint was originally a small startup eBay acquired in 1999. eBay subsequently brought Wells Fargo in as an equity investor, though eventually they bought out WF’s 35% stake for $43.5M. Billpoint was never profitable, costing eBay $10-15M annually to fund operations. So the precise figure’s never been disclosed but my understanding is that a minimum of $90-100M of capital was consumed over the life of the company.