Ideas from Eat People, in plain words
Thirteen rules from Andy Kessler's book Eat People in everyday language, what the provocative title really means, where they fit the ladder, with rupee examples.
First, what the title means
The title is meant to shock, but it is not literal. When Andy Kessler says “eat people”, he means: replace human labour with productivity and technology. Find work that a machine, software or a better way of doing business can do more cheaply, and build a business that does it. In an interview about the book he gave the examples of bank tellers replaced by ATMs and travel agents replaced by online booking, and argued that new and better jobs are created for the people who make the tools (Black Enterprise, May 2011). Nobody is being eaten. Jobs and tasks are.
Read it with one Indian caution. Kessler writes from American technology and venture capital, and he is openly opinionated, including about government. In India labour is plentiful and often cheap, many families depend on the very jobs his rules target (counter staff, helpers, agents, small middlemen), and moving to a better job is neither quick nor certain. So on this page we read “eat people” as eat tasks, not people: use technology to remove low-value work, and move people to work that earns more. Where one of his rules clashes with the site’s standard definitions or with Indian reality, we say so. Our definitions stay as they are.
About the book and the author
Eat People: And Other Unapologetic Rules for Game-Changing Entrepreneurs is by Andy Kessler, published by Portfolio, an imprint of Penguin (Portfolio/Penguin, New York, 2011; the hardcover’s ISBN is 978-1-59184-377-1, per its Open Library record). The publisher’s page gives the ebook date as 3 February 2011 and a paperback on 31 July 2012.
The publisher describes Kessler as a former analyst, investment banker, venture capitalist and hedge fund manager. He co-founded the fund Velocity Capital Management, wrote the earlier book Running Money, and has written opinion columns for The Wall Street Journal. The book’s heroes are entrepreneurs he calls “Free Radicals”, who create wealth by making things cheaper and more plentiful and by overturning slow, costly ways of doing things. It is organised as a set of rules, one per chapter.
What this page is. Below are the book’s 13 rules (12 main rules and a final one about money), explained in our own words. Each rule’s name is Kessler’s chapter title, which we quote; everything else is ours unless we say it is his. For each rule we give the idea, why it matters, where it fits the ladder, any clash with Indian reality, a made-up Indian example in rupees, and one thing to try. We attribute detail to Kessler only where we could check it in his own words or in reviews and interviews (see sources). For some rules we could check only the title; we say so, and the explanation is our reading. Read the book for his full argument and stories.
Then go deeper with three articles that apply the most useful ideas to Indian businesses:
1. Look for work you do once and sell many times
Kessler’s rule: “If it doesn’t scale, it will get stale.”
The idea. Kessler defines scale as something you invent once and then sell a million times. He contrasts a lawyer or doctor, who serves clients one by one, with a chip designer or software writer, whose work is done once and sold millions of times (Black Enterprise). A reviewer sums up the chapter as: rising volume, falling prices and a market with a huge appetite (David Seah’s review).
Why it matters. When most of your cost is paid once (design, software, a mould, a brand), every extra sale costs little, so profit grows much faster than sales. When every sale needs the same hours again, income grows only as fast as you can add people or machines.
Where it fits the ladder. It explains why the higher rungs can earn more: an own-brand manufacturer (3.5) sells one design many times, and product start-ups (4.2), own R&D (4.3) and deep tech or platform companies (4.4) live on it. A job worker (3.1) sells hours, which scale only by adding machines. See unit economics and the article Markets that get cheaper with scale.
India check. Plenty of good Indian businesses don’t scale this way and don’t need to: a busy shop, a 3S dealership, a skilled job-work unit. The rule is a lens for spotting where large wealth comes from, not a verdict on every business.
Hypothetical Billing software for kirana shops
A small team spends ₹12,00,000 building billing software. Shops pay ₹300 a month, and serving each shop (hosting, support calls) costs ₹60 a month.
- Contribution per shop per month: ₹300 − ₹60 = ₹240
- Per shop per year: ₹240 × 12 = ₹2,880
- Shops needed to earn back the build cost in a year: ₹12,00,000 ÷ ₹2,880 = 416.67, so 417 shops
- With 5,000 shops: 5,000 × ₹2,880 − ₹12,00,000 = ₹1,32,00,000 in the first year. The software was built once.
Try this: list what you sell. Mark each item "done once, sold many times" or "done again for every customer". Can any of the second kind become the first?
2. Use what’s getting cheap to save what’s scarce
Kessler’s rule: “Waste what’s abundant to make up for what’s scarce.”
The idea. In the reviewer’s summary, Kessler argues that each economic era has something whose price falls sharply (steel once, computing power later), and that the winners use that cheap thing freely to save what stays scarce, chiefly people’s time and ingenuity (Seah).
Why it matters. Businesses often guard the cheap thing and waste the expensive one: saving on phone data while a salesperson spends hours writing orders by hand.
Where it fits the ladder. Distributors (1.3) and authorised dealers (2.2) with sales teams, and no-stock sellers (1.1) who sell through phones and catalogues. In India today, mobile data, digital payments and phone cameras are cheap; skilled people’s time is not.
Hypothetical A distributor’s salesman and 400 shops
Writing each shop’s order by hand takes the salesman 12 minutes. If shops send orders on the phone from a price list, checking each takes 3 minutes.
- Order-taking now: 400 × 12 = 4,800 minutes, which is 4,800 ÷ 60 = 80 hours a month
- With phone orders: 400 × 3 = 1,200 minutes, which is 1,200 ÷ 60 = 20 hours a month
- Hours freed: 80 − 20 = 60 hours a month
- At 30 minutes per visit to a new shop: 60 × 60 ÷ 30 = 120 new-shop visits a month
Try this: write down the three things your team spends most time on. For each, ask what cheap tool (phone, app, shared sheet, UPI link) could take half of it.
3. Do one thing better than anyone, and plug into others’ chains
Kessler’s rule: “When in doubt, get horizontal.”
The idea. A published excerpt from the chapter says getting horizontal means doing one thing better than anyone else and fitting into a product or process while others do everything else; he sees it as how newcomers beat large companies that try to do every step themselves (book excerpt).
Why it matters. A specialist that serves many customers gets more volume in its one step, so its costs fall and its skill grows faster than an all-in-one firm’s.
Where it fits the ladder. Job workers (3.1) and component suppliers (3.3) that become the best at one process or part for many OEMs; C&F agents (1.4) and logistics specialists; technology licensees (4.1) who buy a proven design instead of inventing everything. See make or buy.
India check. Climbing the ladder often means doing more steps, not fewer: moving from job work to own brand (3.5), and many Indian groups grew by owning their supply chains. Horizontal is a strong route, not the only one. And a specialist with one buyer is not horizontal at all; it is dependent (see customer concentration and One big customer).
Hypothetical A machining shop that specialises
Each machine set-up costs ₹8,000. A shop making many kinds of parts runs batches of 1,000; after specialising in one part family for four customers, it runs batches of 5,000.
- Set-up cost per part, small batches: ₹8,000 ÷ 1,000 = ₹8
- Set-up cost per part, large batches: ₹8,000 ÷ 5,000 = ₹1.60
- Saving per part: ₹8 − ₹1.60 = ₹6.40
- On 20,000 parts a month: 20,000 × ₹6.40 = ₹1,28,000 a month
Try this: name the one step you do better than anyone you know. Who else, outside your current customers, needs exactly that step?
4. Listen to the people at the edge
Kessler’s rule: “Intelligence moves out to the edge of the network.”
The idea. Kessler’s point is that good decisions no longer come only from the top. The people using a product or service, at the “edge”, have more and more say, and he advises against starting a business that doesn’t tap their knowledge (Black Enterprise).
Why it matters. Users and front-line sellers see problems and ideas first. A business that collects and acts on what they see improves faster than one that waits for head office.
Where it fits the ladder. For an OEM (3.5), the edge is its authorised dealers (2.2) and 3S dealers (2.3), who hear every complaint. Platform companies (4.4) go further and let outsiders build on them. See service and warranty.
Hypothetical Dealers spot a recurring fault
A pump maker sells 30,000 pumps a year. Its 120 dealers report the same seal failure, and a design change cuts warranty claims from 4% to 2.5% of pumps. Each claim costs ₹1,800.
- Claims before: 30,000 × 4% = 1,200
- Claims after: 30,000 × 2.5% = 750
- Saving: (1,200 − 750) × ₹1,800 = ₹8,10,000 a year
Try this: ask five customers or dealers this week, "What is the one thing you would change?" Write the answers down and act on one.
5. Wealth comes from productivity
Kessler’s rule: “Wealth comes from productivity; everything else is gravy.”
The idea. The book’s central claim, in the reviewer’s words, is that rising productivity, meaning more output per worker, is what creates wealth; the reviewer also quotes Kessler’s short definition: doing the right things while doing things right (Seah). In his interview he says productivity is the only way to create wealth (Black Enterprise).
In standard terms. Labour productivity is output divided by the labour used: units, or rupees of value added, per worker or per hour. It is not the same as profit or turnover: a business can sell more by adding people without any rise in output per person.
Where it fits the ladder. Every level, but most visibly on the shop floor: job work (3.1), contract manufacturing (3.2) and commodity units (3.4), and sales per staff member in shops (1.2). See capacity and yield and the article Eat tasks, not people.
Hypothetical A namkeen packing line
Six workers, each paid ₹600 a day, pack 3,600 packets a day by hand. With a semi-automatic filling machine, the same six pack 5,400.
- Output per worker before: 3,600 ÷ 6 = 600 packets a day
- Output per worker after: 5,400 ÷ 6 = 900 packets a day
- Rise in productivity: (900 − 600) ÷ 600 × 100 = 50%
- Labour cost per packet before: 6 × ₹600 ÷ 3,600 = ₹1
- Labour cost per packet after: 6 × ₹600 ÷ 5,400 = ₹0.67
Try this: pick one product and work out output per worker per day for last month. Measure it again in three months.
6. Make the product fit people, not people fit the product
Kessler’s rule: “Adapt to humans; don’t make them adapt to you.”
The idea. We could check only this chapter’s title, not its detail, so this explanation is our reading: the products that spread are the ones that fit how people already behave, instead of asking them to learn new habits, fill long forms or change how they pay.
Why it matters. Every extra step loses customers. Small frictions (a form, a login, a language they don’t read) cost more sales than most owners guess.
Where it fits the ladder. No-stock sellers (1.1) and shops (1.2) selling online, and product start-ups (4.2) designing for Indian users: local languages, voice, scan-and-pay UPI, cash on delivery where trust is low.
Hypothetical A shorter checkout
Out of 1,000 shoppers who start checkout, 60% finish a five-step checkout and 75% finish a two-step one with UPI. Each order earns ₹120 of contribution.
- Orders with five steps: 1,000 × 60% = 600
- Orders with two steps: 1,000 × 75% = 750
- Extra contribution: (750 − 600) × ₹120 = ₹18,000 for every 1,000 shoppers
Try this: watch three customers buy from you, without helping. Note every place they hesitate, and remove one.
7. Eat tasks: replace low-value work with technology
Kessler’s rule: “Be soylent: eat people.” This is the rule that gives the book its title.
The idea. Kessler says most technology puts someone out of work, that this is how an economy becomes more productive, and that entrepreneurs should look for jobs that can be made unnecessary, while making sure their own business isn’t the one being made obsolete (Black Enterprise). In a Wall Street Journal column published as the book came out, he split workers into “creators”, who raise productivity, and “servers”, who serve them, and predicted that many server jobs would be replaced by machines, computers and new business models (WSJ, 17 February 2011).
Why it matters. If a task can be done more cheaply by a machine or software, someone will do it, so it is better to be the one who does.
Where it fits the ladder. Manufacturers deciding on machines (3.1 to 3.4), distributors (1.3) with warehouses and billing staff, and new-product founders (4.2) whose product replaces a manual task. See machines and ROI and payback.
India check, fairly. This is where the book clashes most with Indian reality. Labour is cheaper here, so a machine that pays back in months in a high-wage country may take years here; many livelihoods depend on exactly the jobs he calls replaceable; and “better jobs will appear” is a long-run argument that doesn’t help the worker who loses one this year. The productive response is the same as his, with a different emphasis: remove the task, keep and retrain the people where you can, and judge machines on payback, quality and speed, not on cutting heads. The article Eat tasks, not people works through it.
Hypothetical A carton-taping machine
A ₹3,00,000 machine does the taping that keeps two helpers busy. Each helper costs ₹15,000 a month, and the machine costs ₹40,000 a year to run.
- Labour cost of the task: 2 × ₹15,000 × 12 = ₹3,60,000 a year
- Net saving: ₹3,60,000 − ₹40,000 = ₹3,20,000 a year
- Payback: ₹3,00,000 ÷ ₹3,20,000 = 0.94 years
If the two helpers move to dispatch and delivery instead of leaving, the saving shows up as more work done with the same team, not as a lower wage bill.
Try this: list the tasks that take the most hours but need the least judgement. For the biggest one, price a machine or app, and plan what the people doing it would do next.
8. Let prices, not guesses, decide
Kessler’s rule: “Markets make better decisions than managers.”
The idea. In the reviewer’s reading, the chapter is about markets in general, not just the stock market, and its main point is that markets discover prices better than a manager can by guessing (Seah).
Why it matters. Owners often set prices, stock levels and product choices from habit or a hunch. A small test, or a live market price, gives a better answer.
Where it fits the ladder. Commodity manufacturers (3.4) and traders (1.3 to 1.5) who buy and sell at market prices; and new products (4.2), where testing two prices beats arguing about one. See unit economics and pricing.
India check. Some Indian markets are thin, regulated or easy to sway (a few big buyers, controlled prices, seasonal gluts), so a market price can mislead. Use it as evidence, not as an order.
Hypothetical Testing two prices online
A snack brand shows one pack to 2,000 visitors at ₹249 and to another 2,000 at ₹299. The pack costs ₹150 to make and deliver. 3% buy at ₹249 and 2.4% at ₹299.
- Orders at ₹249: 2,000 × 3% = 60; contribution 60 × (₹249 − ₹150) = ₹5,940
- Orders at ₹299: 2,000 × 2.4% = 48; contribution 48 × (₹299 − ₹150) = ₹7,152
Fewer people bought at the higher price, but it earned more.
Try this: pick one decision you make by habit (a price, a reorder quantity, a scheme). Design a small, cheap test that would let customers answer it.
9. Back exceptional people
Kessler’s rule: “Embrace exceptionalism.”
The idea. The reviewer says this chapter discusses a model of human intelligence and how employers screen for ability (Seah). Beyond that we could not check the detail, so our reading is: in work where one person’s ability multiplies results (design, research, software), the best people are worth far more than they cost, so find them, pay them and give them room.
Why it matters. Paying more for a much better person can lower the cost of each result.
Where it fits the ladder. Own R&D (4.3), deep tech (4.4), and design and tooling teams at own-brand manufacturers (3.5). The legal points about hiring tests that the reviewer mentions are American and don’t carry over to India.
Hypothetical Two mould designers
Designer A costs ₹1,20,000 a month and finishes 3 mould designs a month. Designer B costs ₹70,000 and finishes 1.5.
- Cost per design, A: ₹1,20,000 ÷ 3 = ₹40,000
- Cost per design, B: ₹70,000 ÷ 1.5 = ₹46,667
The dearer designer is cheaper per design, before counting quality.
Try this: for your most important role, measure results per person, not cost per person.
10. Win customers, not favours
Kessler’s rule: “Be a market entrepreneur and attack political entrepreneurs.”
The idea. We could check the chapter title but not its detail; the reviewer notes Kessler’s strong dislike of those who take value through tolls, fees and politics without creating it (Seah). Our reading: build a business that wins because customers choose it, not one that depends on protection, licences, quotas or subsidies.
Why it matters. Profits that come from a rule can vanish when the rule changes. Profits that come from customers last as long as you keep serving them well.
Where it fits the ladder. Commodity (3.4) and own-brand manufacturers (3.5), importers (1.5) affected by duties, and new-product companies applying for incentives.
India check. We don’t share the book’s hostility to government. In India, schemes such as production incentives, MSME support and public procurement are legal and often useful, and using them is not wrong. The safer test is: would this business still stand if the scheme ended?
Hypothetical How much of the profit is a scheme?
A unit sells ₹5,00,00,000 a year and makes ₹10,00,000 profit from its operations. A scheme pays an incentive of 4% of sales.
- Incentive: ₹5,00,00,000 × 4% = ₹20,00,000
- Profit with it: ₹10,00,000 + ₹20,00,000 = ₹30,00,000
- Share of profit that comes from the scheme: ₹20,00,000 ÷ ₹30,00,000 × 100 = 66.67%
Try this: work out your profit with every subsidy, incentive and protected price taken out. If it is small, make raising it this year's main goal.
11. When one more copy costs almost nothing, price for the flood
Kessler’s rule: “Use zero marginal cost to create a flood (or someone else will).”
The idea. Marginal cost is the extra cost of one more unit. Kessler points to digital goods, where selling one more song costs nothing (Black Enterprise). Our reading of the rule: when the next unit is nearly free, a low price that brings a flood of users can earn more than a high price, and if you don’t do it, a rival will.
Why it matters. With a near-zero marginal cost, almost the whole price is contribution, so volume, not price per unit, decides profit.
Where it fits the ladder. Product start-ups (4.2) and platform companies (4.4) selling software, content or digital services. For physical goods marginal cost is never near zero; idea 1 applies instead. See unit economics.
Hypothetical An exam-preparation app
Content costs ₹40,00,000 a year to make. Each extra user costs ₹8 a year (servers, messages). Suppose 5,000 students would pay ₹999 a year, or 40,000 would pay ₹299.
- At ₹999: 5,000 × (₹999 − ₹8) − ₹40,00,000 = ₹9,55,000
- At ₹299: 40,000 × (₹299 − ₹8) − ₹40,00,000 = ₹76,40,000
The low price wins only if the 40,000 really come, so test it before betting on it.
Try this: for a digital product, work out the cost of one more user. If it is a few rupees, ask what price would bring ten times the users.
12. Own the connection to your customers
Kessler’s rule: “Create your own scarcity with a virtual pipe.”
The idea. Kessler’s “virtual pipe” is the link between a business and its customers that the business controls. He gives the examples of a media company that owns a channel, site or magazine and rents space to advertisers, and of Apple’s online music store (Black Enterprise). Whoever owns the pipe decides who else gets through it, and on what terms.
Why it matters. The customer relationship is often worth more than the product. If someone else owns it, they can squeeze your margin.
Where it fits the ladder. A distributor (1.3) that owns the ordering link with its shops; a master franchisee (2.5) that owns a region’s customer base; and platform companies (4.4). It also explains the risk on platform-led businesses: the platform owns the pipe. See When technology replaces the middleman.
Hypothetical A distributor’s own ordering app
A distributor’s 600 shops order on its app. It rents 5 banner slots to brands at ₹15,000 a month each, and the app costs ₹3,00,000 a year.
- Banner income: 5 × ₹15,000 × 12 = ₹9,00,000 a year
- After the app’s cost: ₹9,00,000 − ₹3,00,000 = ₹6,00,000 a year
Try this: ask, for each of your main customers, "Who do they contact first when they need more: me, the brand, or a platform?" That is who owns the pipe.
13. Money flows to the highest returns
Kessler’s rule: “Money sloshes to the highest returns.” This is the book’s last rule.
The idea. We could check only the title, so this is our reading: investors’ money moves to wherever it earns most. That means a niche with unusually high returns will attract competitors, and that your own money in the business should earn clearly more than it would somewhere safer.
Why it matters. High returns rarely last on their own. And if your business earns little more than a fixed deposit, you are taking risk and doing the work for very little.
Where it fits the ladder. Deciding whether a dealership or outlet is worth it (2.2 to 2.5) and raising money for new products. See ROI and payback and investment math.
Hypothetical A dealership against a safe deposit
A dealer has ₹1,20,00,000 invested and makes ₹18,00,000 net profit a year. A safe deposit might pay 7% a year.
- ROI: ₹18,00,000 ÷ ₹1,20,00,000 × 100 = 15%
- The deposit would earn: ₹1,20,00,000 × 7% = ₹8,40,000
- Extra earned for the risk and the work: ₹18,00,000 − ₹8,40,000 = ₹9,60,000 a year
Try this: work out last year's ROI on everything you have put into the business, and compare it with what the same money would earn safely.
Sources
We used these to check the book’s facts and which ideas are Kessler’s. The explanations and examples on this page are our own.
- The book: Penguin Random House’s page for Eat People by Andy Kessler (Portfolio; dates and description), the Open Library record of the 2011 Portfolio/Penguin hardcover (publisher, place, year, ISBN and the table of contents), and the Internet Archive’s catalogue record (same table of contents).
- Kessler in his own words: an interview, “Looking for the Next Big Thing”, Black Enterprise, 4 May 2011 (ideas 1, 4, 5, 7, 11 and 12; also the title’s meaning and his co-founding of Velocity Capital Management); and his Wall Street Journal column “Is Your Job an Endangered Species?”, 17 February 2011 (idea 7; the paper’s page may need a subscription).
- Reviews and excerpts: David Seah’s review of the book, 17 March 2011 (ideas 1, 2, 5, 8, 9 and 10, and the “Free Radicals”); Publishers Weekly’s review, February 2011 (the earlier book Running Money, and the book’s opinionated tone); and a published excerpt from the chapter on getting horizontal (idea 3).
Some reader summaries attribute specific pricing rules to the book, for example how much volume should rise when you cut prices. We could not confirm them in Kessler's own words, so we have left them out. For ideas 6, 9, 10, 11 and 13 we could check only part of the chapter, or only its title; where we say "our reading", the interpretation is ours.