Part five · Chapter 17

SEED THE HARD SIDE

A few people carry every loop: eBay sellers for PayPal, streamers for Whatnot, whales for OVRLND. Find yours and serve them by hand.

Andrew Chen's The Cold Start Problem calls it the hard side: in any network, one group is harder to win and does most of the work. In a marketplace it's usually supply. In a consumer brand's loop it's the small group of customers who bring most of the others. Find them before you tune the reward for everyone else.

Every loop in this book had one

The whales

OVRLND's average order fell from $45 in late 2025 to $22 to $24 by mid-2026 as order volume climbed: lots of small buyers, fewer big stacks. A segment of customers with $100-plus orders existed, and the founder's own read was that high spenders weren't being served hard enough.

The plan built them a branch of their own. Customers with an order of $100 or more, or $250 lifetime, stop getting the standard flow. They get a plain-text note and a text from the shop, their rank among entrants (top 3% of stacks, say), 24-hour early access to each new giveaway reveal, and a Founder's Stack bundle in place of the membership pitch.

None of that is a discount. It's status and access: the currencies from Pick the Currency that cost nearly nothing and are worth the most to the people who care most.

Pay the few in status. They're the ones who notice.

Measure the concentration

Rank last year's inviters by referred customers who completed a first order. Then read one number: the share brought by the top tenth.

Say 2,000 customers referred at least one person and the top 200 brought 60% of all referred customers. Those 200 are your sellers. A 20% bump in the reward for everyone mostly pays people who bring one friend a year. A monthly call, an early product, a code with their name on it, or a real share of the revenue they bring does more for the 200.

Seed by hand, then build the system

Whatnot's founders sold on their own first show. PayPal leaned on a dense group that already transacted every day. The common move is manual: go to the few people who already have an audience or a habit, and make the loop work for them one at a time before automating it.

Wrong for you if

Rank last year's inviters. If the top tenth brought under half of referred customers, your loop is broad rather than concentrated, and the general reward and the trigger matter more than a program for the few.

Do this

This is one chapter of Close the Loop, which is free and readable in full on a single page with no form in front of it.