Ten books and essays built the vocabulary of loops. Here's what each one gets right, and the three things none of them had to deal with: margins, the law and proof.
You don't need a new theory of loops. The good ones exist. What they were mostly written for is software, where the next user costs almost nothing to serve and a reward can be a feature flag. A consumer brand ships a box, pays for it, and runs its promotions under sweepstakes, advertising and consumer law. So take the theory whole, then add three things.
Ries named three engines of growth. The sticky engine grows when you add customers faster than you lose them. The viral engine grows when each customer brings more than one more. The paid engine grows when a customer is worth more than they cost to buy, so the margin funds the next purchase. His advice was to tune one engine at a time.
For a DTC brand, all three run at once, and the sticky one is inside the other two. A customer who churns after one order can't invite anyone later, and can't pay back a paid acquisition either. Retention Is the Fuel builds on this.
"Growth Loops are the New Funnels" argued that funnels describe growth as a straight line, when the fastest-growing companies run closed systems: an input goes in, an action happens, and the output becomes the next input. Each company's loop is specific to its product, and the loops, not the channels, are what compound.
That's the drawing in chapter 2. What it leaves to you is the cost of each turn.
Skok's formula showed that how fast a loop turns can matter as much as how many people each turn brings. Loop Math uses it and adds the caveat for physical goods: below a K of one, cycle time changes when the loop pays, not how much.
Chen, who spent years at Uber and then Andreessen Horowitz, describes how networks begin in a small, dense group he calls an atomic network, how one side of a network (the hard side) does most of the work, and how growth tips, accelerates, hits a ceiling and eventually becomes a moat. His earlier essay "The Law of Shitty Clickthroughs" argued that every marketing channel's response rates decay as the channel gets crowded.
PayPal's eBay sellers and Whatnot's streamers are both hard sides. Seed the Hard Side applies the idea to a brand's own best customers.
Ellis coined "growth hacking" and worked with Dropbox in its early years. The book argues for a cross-functional growth team, a clear idea of the moment a product clicks for a new user, and a high tempo of small tests. Its most useful idea for loops: don't pour customers into a product that hasn't clicked for them yet.
Penenberg told the early stories: Hotmail's footer, PayPal's email payments, eBay's sellers. His point was that the strongest loops spread as a byproduct of using the product, not as a separate act of promotion. The Loop Gallery returns to Hotmail.
Eyal's Hook Model has four steps: a trigger, an action, a variable reward and an investment. The investment is something the user puts in (data, content, progress, reputation) that makes the next trigger more likely. It's a loop inside one customer, and it feeds the loop between customers. OVRLND's lifetime entries in The Game Is the Loop are an investment in exactly this sense.
Meadows' Thinking in Systems (2008) describes reinforcing loops that amplify and balancing loops that resist, and warns that delays between cause and effect make systems overshoot. Senge's The Fifth Discipline (1990) names a pattern he calls limits to growth: a reinforcing loop runs until it triggers a balancing one, and pushing harder on the reinforcing loop makes things worse. The fix is to find and remove the limit. When Loops Break is this idea applied to fraud, the law and saturation.
Collins introduced the flywheel in Good to Great (2001) and gave it a whole book in Turning the Flywheel (2019). Amazon's version, which Jeff Bezos sketched in 2001, runs from lower prices to more customers to more sellers to lower costs, and back to lower prices. Collins' advice: write your own flywheel with only a few components, in an order where each one drives the next.
Cialdini's Influence (1984) explains why double-sided rewards work (reciprocity), why winners and testimonials pull people in (social proof), why waitlists hold (commitment) and why drops sell out (scarcity). Berger's Contagious (2013) explains why some things get talked about: they make the sharer look good, they're triggered by everyday cues, they carry emotion, they're visible, they're useful, and they come with a story.
Take the theory whole. Then price it, clear it and prove it.
| Gap | Why it matters for a consumer brand | Where this book handles it |
|---|---|---|
| Margins | Every referred customer carries cost of goods, shipping and a reward. A loop can raise your blended acquisition cost if the currency is wrong. | Loop Math, Pick the Currency |
| The law | Prizes, chance and purchases can make a lottery. Messages sent for a customer fall under consent rules. Rewarded posts need disclosure. | The Game Is the Loop, When Loops Break |
| Proof | Referral codes get used by people who were coming anyway. Only a holdout says what the loop added. | Count the Loop, Not the Link |
If you sell software or a service with near-zero marginal cost and no chance-based rewards, the books above cover most of what you need, and you can skip to part two for the case detail.
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.