In Donahoe's case, he has been trying to predict what major retail chains will do. Back in July 2011, he confidently told investors that eBay would announce a major retail trial by the end of 2011—that was Home Depot, which wasn't announced until January 2012—and that 2012 would bring 20 national retail trials. Last week, Donahoe slashed that figure almost in half, now saying that 10 to 15 retailers will be offering PayPal in-store this year. Asked to explain the different numbers, he told a TV interviewer: "We tend to under-promise and over-deliver. So no, I wouldn't say we're scaling back at all."
Setting aside the fact that he didn't answer the question, this is impressive spin. Donahoe had been promising 20 and he's now scaling it down to 10 to 15, which means he's now promising 10. That's not under-promising. It's over-promising. And it's not over-delivering, because none—beyond Home Depot—have been delivered yet. If Donahoe announces 28 retailers next week, that would be over-delivering.
The issue shouldn't be under-promising at all. It should be, "Why is he promising?" Even with eBay paying for everything, retailers will roll out—or not—at the last minute and on their own timetable. Investors understand that and would gladly accept, "We're talking with several major chains and we'll tell you as soon as we have anything concrete."
What Donahoe also didn't address are the growing security concerns of the Home Depot trial. Home Depot clearly is happy with the eBay-funded effort, given that it made the unusual move of expanding the PayPal offer to the full chain. But will other chains be as happy if their fraud losses are not covered by eBay? Maybe eBay should start releasing its PayPal fraud losses from the Home Depot rollout? Now there would be some stats that 10 to 15 chains might find very interesting.