The beginning life stage of a startup always seems to be different from the middle, in terms of the way it operates and - most importantly - how it acquires customers.
In the early days of Groupon, a sales force would have to actively call on vendors to get them on-board. Now, in some cities vendors are flocking to their platform without anyone calling on them. Conversely, consumers would not be able to claim their Groupon until a 'tipping point' number of consumers had signed on, and "the deal was on". Now, every deal "is on" because Groupon can guarantee huge numbers will flock to buy its coupons.
Similarly, OpenTable first started by installing a CRM system into the top restaurants in San Francisco - sometimes at zero upfront cost to the restaurant - before other restaurants flocked to buy into the system. Zaarly claims to have first built a community who bought into its concept 50 to 100 people in Chicago alone. These people bought in to it before there was even a product. When the marketplace launched, this group was the seed for attracting the rest.
I think we can often build our businesses with just the middle scenario in mind - without realizing the dynamics of the actors involved in that situation is different to the dynamics at the beginning. To be successful, we have to build a proposition that caters to the beginning scenario, before evolving to the middle.