Indulge me in a flight of fancy in which we pretend, for the sake of argument, that
a) Paypal is run not by Snidely Whiplash clones but
b) by smart geeks working with thin margins in a highly regulated industry where
c) customers are at risk essentially never,
d) merchants eat 100% of the risk if they stay in business, and
e) Paypal eats 100% of the risk if the merchant doesn't.
Why is Paypal very skeptical of pre-sales? Because, if the business fails (as new businesses often do), customers will file chargebacks. Their banks will hear "Internet merchant did not deliver as promised" and sustain the chargeback automatically. Paypal will lose that argument with the bank, 99.999% of the time, and have to seek restitution from the merchant.
Paypal has to do underwriting -- basically, guessing at probable risks and likelihood of partial repayment -- for new merchant accounts. What percentage of sales are at risk of chargeback in a pre-sales business? A Very High Percentage (TM). What is the probable chance of failure of a new business in developing a new product? Fairly high. Given product failure, what assets will be available to Paypal (in the Paypal account or the linked bank account) for automatic recovery from the failed business? Very Little (TM). What is Paypal's margin on this business? A fraction of a percent.
Now we break out the Hadoop cluster and use several billions of dollars of transactional data to construct a model of what the expected loss is, expected recovery given loss, and expected margin in event of non-loss is.
This puts us in an incredibly uncomfortable position as we do not feel that it's remotely in their jurisdiction to ask for a detailed budget of our business, any more than it is within our right to ask for theirs.
This communication is incredibly useful from Paypal's perspective among multiple axes:
1) It signals very strongly "We are not only unwilling to comply with the table stakes of every underwriting process for businesses everywhere, we are so inexperienced at business as to be unaware that this is table stakes, and accordingly you should dramatically revise upwards your estimate of our risk of failure."
2) It provides Paypal a simple, face-saving out for declining this business without having to say, in so many words, that "You seem, oh, 93% likely to ship this year. You get an A! This means, however, you are 7% likely to lose all the money, and we only make .9% margins, so this is going to be a No. We get that you don't like this. We don't like having to decline hundreds of dollars of revenue either, but we have the experience of losing hundreds of millions to fraud and know that some revenue just isn't worth the risk. We respect that you might not agree with this, but don't feel the need to spend additional resources paying for our computer programmers, underwriters, lawyers, and accountants to give you an expensive education in the realities of e-commerce on our nickel."
Let's talk about the difference between Paypal and Indiegogo:
1) You pay Paypal ~3% when their costs are probably approaching ~2%. Indiegogo would charge ~7% for the same thing. One of the luxuries when selling something which is five times as lucrative is that you can self-insure against project failure.
2) Indiegogo believes it has a different business model than Paypal and that they have a uniquely better understanding of the risks of crowdfunding, whereas Paypal has had their filters tuned by too many middle Americans selling Beanie Babies.
3) Paypal has lost hundreds of millions of VC money to fraud and Indiegogo hasn't. Paypal decisionmakers might at this point give Indiegogo the sort of look a school psychologist gives a C student with a drug habit who has just announced that they're taking a semester off to find themselves, man. They know which way this story is going to turn out, which is in its own special way as bad as not knowing how the story is going to turn out.
Are the risks larger because we are successful?
Ask a simple question, get a simple answer: yes! Paypal loses more on a $1,000,000-in-transactions account which goes bad than a $1,000-in-transactions account which goes bad, clearly. You might wonder "Well are we more risky than the same aggregate volume spread over N accounts?", in which case the answer is available to Paypal's Hadoop cluster but plausibly "Yes, with a p value which would make a statistician weep." Accounts which go 0-to-60 in processed transactions are hugely disproportionately likely to be outright fraud (Paypal has had many, many, many encounters with carders smart enough to have invented the suborn-a-botnet and make-a-lot-of-small-donations attack prior to having seen it on Breaking Bad). Additionally, it is quite plausible that Paypal could demonstrate that success is a curse to new businesses and most which blow up proceed to, well, blow up. (Which they would, of course, not love to disclose publicly.)
This dynamic is not unique to pre-sales or crowdfunding. It also explains Paypal's active hostility to many other business models, including money services businesses, third-party payment aggregators, and travel agents. (Most people don't immediately associate travel agents with having a lot of payment industry problems, but they do: they make lots of big-ticket sales but have low working capital, and if a cruise gets canceled or a hotel goes bankrupt or any of the standard vicissitudes of the industry causes them to eat a bunch of chargebacks all at once, they go bankrupt and their payment processor is on the hook for hundreds of thousands or millions of dollars as that bankruptcy causes cascading failure to deliver promised goods or services.)