Library
Decades of research into revenue, sales, marketing and psychology
Discover more about the math and science driving everything from the click through the sale and beyond.
If you’ve ever watched every KPI go green while revenue fell behind plan, you’re in good company. Economists, statisticians and marketers, some of them with Nobel Prizes, have been studying why for more than a century. It just doesn’t make it into the marketing, sales, or revenue meeting. This library puts it in one place, 119 published works plus 19 books every revenue team should read. Click into any section and read as much as you want. Every entry links back to the original.
Every entry was checked against its publisher, a DOI or the original source in October 2026. The summaries are in plain English, and anything in quotation marks is the author’s own words. For what it all means for your team, read the articles.
Research
The research, by theme
Ten themes covering economics, statistics, marketing science and management. Click a theme to see what it’s about, the works broken down in plain English, and every other work in it with a line on why it matters to your revenue team.
Theme 1 of 10 · 12 works
Adverse selection
Why does optimizing to the average buy the bottom?
If a buyer can’t tell good from bad, they pay for average, and that one price ends up pulling in the sellers, applicants and customers you want the least. The good ones go somewhere that pays them what they’re worth, and you’re left with the rest. It all starts with Akerlof’s market for lemons, and it’s the math behind the Free Lamborghini. It’s a snake eating its tail. Every time you optimize to the average, the average gets a little worse.
Explained
- Why the Average Price Buys the Lemons What happens when a buyer can't tell a good car from a lemon? In 1970 George Akerlof showed that the price settles at the average and the good sellers leave. Two journals turned the paper down as too trivial. George A. Akerlof, 1970
- Why One Price for Everyone Doesn't Last Rothschild and Stiglitz showed that when insurers can't see who's risky, a single price for everyone falls apart. What survives is a menu that gets customers to sort themselves. Michael Rothschild and Joseph Stiglitz, 1976
- Why Banks Say No Instead of Charging More Stiglitz and Weiss showed that raising the price of a loan changes who applies and how they behave. Past a point, a higher rate makes the whole book riskier. Joseph E. Stiglitz and Andrew Weiss, 1981
Every work in this theme
- Akerlof (1970) ● One blended price or bid pulls in the worst slice of your applicants, vendors and leads.
- Arrow (1963) Cream-skimming got its name in 1963. If you don't sort your customers, a competitor will do it for you.
- Rothschild & Stiglitz (1976) ● Tiers and commitments sort your customers. Put out one flat offer and you're inviting competitors to take your best ones.
- Stiglitz & Weiss (1981) ● The terms you set decide who applies. Raise the price to cover the risk, and the safest borrowers are the first to walk away.
- Cutler & Reber (1998) This death spiral really happened. Flat pricing plus free choice sorts your costliest customers right into your richest offer.
- Einav & Finkelstein (2011) This is the counterpoint. Adverse selection doesn't show up everywhere, and sometimes your most eager buyers are your best, so measure it in your own book. It's also the clearest one-chart picture of the mechanism.
- Ausubel (1999) A high response rate doesn't mean quality. The worse your offer, the worse the people who answer it.
- Agarwal, Chomsisengphet & Liu (2010) A peer-reviewed study found the same thing. How you design the offer decides which customer you get.
- Karlan & Zinman (2009) Your terms change how customers behave after the sale, not only who buys. So don't blame everything on selection.
- Adams, Einav & Levin (2009) This one paper gives you two signals. The size of the ask tells you something, and a down payment works as a screen.
- Hendren (2013) Some segments can't be served profitably at any price, so turning them away makes sense. A "no" to those applicants isn't a lost sale.
- Cutler & Zeckhauser (1998) The death spiral told year by year, with the price and the enrollment. One price rule for everyone, and the healthiest customers left first.
Open Adverse selection: every work with its citation and what it found
Theme 2 of 10 · 11 works
Signals, screening and ordeals
When does friction sort customers, and when does it sort the wrong way?
So how do you tell the customers worth having from the rest? With a signal that costs the wrong customer more to send, with a menu that lets people sort themselves, or with a small hurdle only a serious buyer will clear. You’ll also find the cases where friction sorts the wrong way.
Explained
- A Signal Works When It Costs the Wrong Side More Michael Spence asked how the better side of a market gets noticed when quality can't be seen. The answer is a signal that's cheaper for the high-value side to send than for the low-value side to fake. Michael Spence, 1973
- A Small Hassle Did the Sorting In 400 Indonesian villages, households that had to travel and apply for a cash-transfer program turned out much poorer than households enrolled automatically. A little friction did the sorting, and adding more didn't help. Vivi Alatas, Abhijit Banerjee, Rema Hanna, Benjamin Olken, Ririn Purnamasari and Matthew Wai-Poi, 2016
Every work in this theme
- Spence (1973) ● A buyer signal like a call, a paid pilot or a data upload only tells you something if it costs a poor-fit prospect more than a good-fit one.
- Stiglitz (1975) Every hurdle you add to qualification costs money, so check whether it adds value or just relabels who gets served.
- Nelson (1974) Your own advertising is a signal too. What you offer tells the high-value buyer who you're for.
- Klein & Leffler (1981) Why would a buyer believe your quality promise? It takes a price premium plus visible, sunk brand investment.
- Milgrom & Roberts (1986) Pricing and visible spend are acquisition levers to you, but buyers read them as quality signals too.
- Zahavi (1975); Grafen (1990), with the dispute It's a great story with a live dispute. For your business decisions, Spence's version, that a signal has to cost the wrong customer more, is safer ground than "costly means honest."
- Mussa & Rosen (1978) Your good/better/best tiers are a screening device, and the limits on your entry tier protect your premium revenue.
- Nichols & Zeckhauser (1982) This is the theory behind "some friction sorts." A phone call or a booked appraisal is an ordeal that serious buyers will put up with.
- Alatas, Banerjee, Hanna, Olken, Purnamasari & Wai-Poi (2016) ● One small ask filters well. Pile on more friction and you add cost without getting a better mix.
- Deshpande & Li (2019): the counterpoint Friction can screen out the very people you want. So every time you add a step, check who drops out, tier by tier.
- Einav, Jenkins & Levin (2013) Scoring pays you twice, once for saying no to the worst and again for investing more in the best. That's "sort, don't reject."
Open Signals, screening and ordeals: every work with its citation and what it found
Theme 3 of 10 · 19 works
Power laws and concentration
Why is the average where almost nobody is?
In most books of business a few accounts carry most of the value. Where those lopsided distributions come from, how to measure them without fooling yourself, and why the average lands where almost none of your accounts actually are.
Explained
- The 80/20 Rule Isn't Pareto's Pareto found incomes falling on a straight line in country after country. Juran put Pareto's name on a universal rule, then apologized for it. The shape Pareto found is real, but 80/20 was never a fixed ratio. Vilfredo Pareto, 1896–97; J. M. Juran, 1951 and 1975
- Performance Isn't a Bell Curve O'Boyle and Aguinis looked at 633,263 researchers, entertainers, politicians and athletes and kept finding the same shape. A few performers produce most of the output. Ernest O'Boyle Jr. and Herman Aguinis, 2012
- Some of Your Customers Are Costing You Money Rank customers by profit and add them up. The running total climbs well past 100%, then falls back. Robert Kaplan found that shape in virtually every customer profitability study. Robert S. Kaplan and V. G. Narayanan, 2001; Kaplan and Robin Cooper, 1998
Every work in this theme
- Pareto (1896–97) ● This is the original evidence that value piles up in a heavy tail.
- Juran (1951; 1975) Treat 80/20 as a rule of thumb about concentration. There's no law that says the ratio is fixed.
- Zipf (1949) Rank your customers or keywords by revenue and you'll see this shape.
- Simon (1955) Why do your biggest accounts and channels keep pulling ahead? This gives you a mechanism for it.
- Price (1976) It's the formal version of the compounding you get from referrals, expansion and word of mouth.
- Barabási & Albert (1999) In your partner, affiliate and referral networks, most of the value goes to whoever is best connected.
- Gibrat (1931) When growth is proportional, you get a lognormal distribution. That's heavy-tailed, but it isn't always a true power law (see Mitzenmacher, 2004).
- Newman (2005) Start here for the math. It's the standard primer.
- Clauset, Shalizi & Newman (2009) Don't claim a strict power law off a chart. Their method fits the tail and tests the fit, which a business rarely needs. For the business point you only need to see how much of the book sits above the average: split it at the average, split each half again, and compare each part's share of the accounts with its share of the revenue.
- Mitzenmacher (2004) Before you call your customer value a power law, check whether it's lognormal.
- Taleb (2007) Your customer revenue lives in Extremistan, where one account can dominate the total.
- Taleb (2020) In a heavy-tailed book, your average LTV, average deal size and top-20% share are all unstable.
- O'Boyle & Aguinis (2012) ● Your reps' output is heavy-tailed too. So bell-curve quotas and splitting lead flow evenly will mislead you.
- Aguinis & O'Boyle (2014) It matters for how you hire, keep and route work to your top sellers.
- Schmittlein, Cooper & Morrison (1993) Your top-20% share depends on the time window, so read value over time instead of one month.
- Kim, Singh & Winer (2017) Even at the grocery store, the top fifth of buyers carries most of the volume.
- Sharp, Romaniuk & Graham (2019); Sharp (2010) This is the counterpoint. Concentration varies, and it isn't always 80/20. In the same paper's appendix, the top 20% of buyers carried 67% of sales at listed companies on average, and 58% at subscription firms. Measure your own book before you plan around a ratio.
- Kaplan & Narayanan (2001); Kaplan & Cooper (1998): the whale curve ● This is the power law in profit terms. Part of your tail is small, and part of it costs you money.
- Galton (1886; 1889) There are two lessons here. Last quarter's extremes tend to drift back toward average next quarter. And the bell curve comes from adding many small effects, while revenue compounds.
Open Power laws and concentration: every work with its citation and what it found
Theme 4 of 10 · 19 works
Customer value, acquisition and retention
Why does who you acquire decide what they're worth?
What a customer is worth to you depends on how you acquired them. The research on acquisition discounts, lifetime value, retention models and uplift shows why who you buy decides what you get, and why one blended churn rate can mislead you.
Explained
- A 35% Discount Buys Customers Worth Half as Much What are the customers you win with a discount actually worth? Michael Lewis followed them at a newspaper and an online grocer, and the deeper the discount, the less they were worth. Michael Lewis, 2006
- Why Retention Rates Rise Even When Nobody Gets More Loyal Why does a cohort's retention rate climb every year? Fader and Hardie showed that the customers likely to leave go first, so the ones who are left were always the stayers. Peter S. Fader and Bruce G. S. Hardie, 2007
- Don't Chase the Customers Most Likely to Leave Who should get your save offer, the customers most likely to leave or the ones the offer will actually change? In two field experiments, Eva Ascarza found that aiming at the second group worked better. Eva Ascarza, 2018
Every work in this theme
- Lewis (2006) ● The deeper the discount you acquire customers with, the less they end up being worth.
- Villanueva, Yoo & Hanssens (2008) The channel a cohort came in through predicts what it'll be worth.
- Datta, Foubert & van Heerde (2015) Judge your free trial on whether it pays back in lifetime value. Sign-ups alone won't tell you.
- Anderson & Simester (2004): the counterpoint Discounts don't always bring in worse customers. It depends on who gets the discount.
- Byers, Mitzenmacher & Zervas (2012) Customers who come in through deals behave differently, and so do the signals they leave, like reviews.
- Schmittlein, Morrison & Colombo (1987) How do you count live customers when there's no contract? This model does it.
- Fader, Hardie & Lee (2005) This gives you a practical forecast of repeat purchases and lifetime value.
- Fader & Hardie (2007) ● A cohort's retention rises with age because the leavers leave first. Nobody got more loyal. So read churn by cohort instead of as one blended rate.
- Fader & Hardie (2010) Value your subscription book with one blended churn rate and you'll get it wrong.
- Vaupel, Manton & Stallard (1979) Fader and Hardie took this from demography and applied it to customers.
- Gupta, Lehmann & Stuart (2004) Retention is the biggest lever you have on firm value. So the next question is whose retention.
- Reichheld & Sasser (1990) It's the most-quoted retention claim in business. The figures are case examples from the article, so don't treat them as a general law.
- Reinartz & Kumar (2002) Tenure doesn't tell you value, so put your retention effort where the value is.
- Venkatesan & Kumar (2004) Rank your accounts on expected future value instead of past spend or count.
- Ascarza (2018) ● Aim your retention effort at the customers it actually changes, weighted by value, instead of at the highest churn scores.
- Ascarza, Iyengar & Schleicher (2016) Reach out to save customers and you can wake up "sleeping dogs" who would have stayed anyway.
- Lemmens & Gupta (2020) Save your valuable customers instead of just chasing the ones most likely to leave.
- Radcliffe & Simpson (2007); Siegel (2013): uplift's four groups Only spend where the effort changes the outcome. That goes for your ads as much as your retention, since brand search reaches sure things.
- Yang, Eckles, Dhillon & Aral (2024) Validate your early signals against long-term value, then use them to decide who to target.
Open Customer value, acquisition and retention: every work with its citation and what it found
Theme 5 of 10 · 8 works
Selection bias and paradoxes
Why do you only see the customers you let in, and why do averages reverse?
You only ever see the customers you let in, and averages built from them can flip when you split them up. Heckman’s correction, reject inference, the winner’s curse, Simpson’s and Berkson’s paradoxes, and Wald’s bombers.
Explained
- When the Total Points the Wrong Way In 1973 Berkeley's graduate admissions looked biased against women. Department by department, they weren't. The difference was the mix, and the same thing happens in your churn rate. E. H. Simpson, 1951; P. J. Bickel, E. A. Hammel and J. W. O'Connell, 1975
- The Planes That Came Back In 1943 Abraham Wald worked out where bombers were most vulnerable using only the planes that survived. What he actually did is more useful than the story you've heard. Abraham Wald, 1943; explained by Marc Mangel and Francisco J. Samaniego, 1984
Every work in this theme
- Heckman (1979) Train a model only on approved or closed deals and it inherits the selection that produced them.
- Lakkaraju, Kleinberg, Leskovec, Ludwig & Mullainathan (2017) You never see outcomes for the files you declined or the leads your reps skipped. Differences between your underwriters or reps are your natural experiment.
- Hand & Henley (1993; 1997): reject inference It's the lending industry's name for "you only see the loans you made."
- Capen, Clapp & Campbell (1971); Thaler (1988): the winner's curse If you win a contested lead, keyword or media auction, it's bad news about your own estimate. Your A/B-test winners suffer the same curse (see Lee and Shen, 2018).
- Simpson (1951); Bickel, Hammel & O'Connell (1975) ● A blended rate like churn, conversion or approval can move the opposite way from every segment inside it, because the mix changed.
- Berkson (1946) If you only analyze the deals you won, you'll find trade-offs that don't exist in the market.
- Mangel & Samaniego (1984): Wald's bombers ● Your surviving customers can't show you why the others left.
- Bazerman & Samuelson (1983): the winner's curse in a classroom The curse doesn't need an optimistic room. Here the room guessed low and the winners still overpaid, because winning picks the top of the spread. More uncertainty and more competitors make it worse: more variants in a test, more bidders on a lead or a keyword.
Open Selection bias and paradoxes: every work with its citation and what it found
Theme 6 of 10 · 14 works
Measurement, experiments and ad platforms
Why do tests pick the wrong winner, and ad platforms the cheapest conversion when every conversion counts the same?
What controlled experiments actually show about advertising, and why ad platforms go find you the cheapest conversion instead of the best customer. From eBay switching off its search ads to how platforms deliver an A/B test, plus the trials where a better number meant a worse outcome.
Explained
- What Happened When eBay Turned Off Its Search Ads What happens if you just turn the ads off? In 2012 eBay switched off its brand-name search ads, then its other Google search ads across a third of the US, and found the ads returned a fraction of what the usual numbers said. Thomas Blake, Chris Nosko and Steven Tadelis, 2015
- Your Ad Test Compared Audiences, Not Just Ads When you A/B test ads on a platform, each version gets shown to a different, optimized mix of people. Braun and Schwartz showed that can change the size of the result, and even which ad wins. Michael Braun and Eric M. Schwartz, 2025
- The Drugs Improved the Number and More Patients Died In the 1980s, heart-attack patients were given drugs that suppressed irregular heartbeats, a warning sign of sudden death. A trial found the drugs improved the number and worsened the outcome. Debra S. Echt et al., 1991 (the CAST trial); Thomas R. Fleming and David L. DeMets, 1996
Every work in this theme
- Blake, Nosko & Tadelis (2015) ● Attribution gives your ads credit for buyers who were coming anyway.
- Lewis & Rao (2015) Be suspicious of any precise ROAS number you're handed, especially when your revenue is heavy-tailed.
- Gordon, Zettelmeyer, Bhargava & Chapsky (2019) A model built on platform data can't stand in for a holdout.
- Johnson, Lewis & Nubbemeyer (2017) It's a cheaper way to test retargeting, the channel most distorted by selection.
- Braun & Schwartz (2025) ● The platform picks who sees each ad, so a creative can win just by reaching better prospects. The targeting decides your test before the ad does.
- Ali, Sapiezynski, Bogen, Korolova, Mislove & Rieke (2019) Whatever your targeting settings say, the platform decides who sees your ad.
- Lambrecht & Tucker (2019) An algorithm that minimizes your cost drifts toward whoever is cheapest to reach, whatever they're worth.
- Kohavi, Tang & Xu (2020) Whatever metric you test on ends up deciding which customers you select.
- Hohnhold, O'Brien & Tang (2015) Chasing short-term revenue wins can teach your customers to tune you out.
- Athey, Chetty, Imbens & Kang (2025/26) It's the rigorous version of "signals," early indicators validated against long-term value.
- Fleming & DeMets (1996) Your MQLs, CTR and form fills are surrogates. Validate them against revenue before you optimize to them.
- Echt et al. (1991): the CAST trial ● It's the starkest case on record of a better number and a worse outcome.
- Lee & Shen (2018) The cumulative lift your testing program claims is inflated.
- Oldroyd, McElheran & Elkington (2011) Speed matters. When you fix bad routing, route on value with a short time-out so speed doesn't get lost.
Open Measurement, experiments and ad platforms: every work with its citation and what it found
Theme 7 of 10 · 19 works
Metrics and incentives
Why do measures stop measuring once they become targets?
Why a number stops measuring what you care about once people are paid to move it. Goodhart, Campbell and Kerr, the multitask problem, sales compensation, and what happens when the target and the goal come apart.
Explained
- The Most Quoted Line in Management Isn't Goodhart's Charles Goodhart was writing about the money supply in 1975. The sentence everyone quotes came from Marilyn Strathern in 1997, and she was writing about British universities. Both describe what happens when a number becomes a target. C. A. E. Goodhart, 1975; Marilyn Strathern, 1997
- Rewarding A While Hoping for B In 1975 Steven Kerr collected examples of reward systems that pay for one thing while hoping for another, from war and medicine to universities and business. Read it today and it sounds a lot like a sales comp review. Steven Kerr, 1975 (updated 1995)
- The Work You Don't Pay For Gets Dropped Holmström and Milgrom showed that incentive pay decides where people put their attention, on top of how hard they work. Pay on the part of the job you can count and the part nobody measures gets less of it. Bengt Holmström and Paul Milgrom, 1991
Every work in this theme
- Goodhart (1975) ● Once you make cost per lead the target, its link to revenue starts to break.
- Strathern (1997) It's the line everyone quotes, credited to the right person.
- Campbell (1979) You end up corrupting two things, the number itself and the behavior it was supposed to measure.
- Lucas (1976) The funnel conversion rates you estimated from history will shift once you change bids, pay or routing, because people respond to the change.
- Kerr (1975) ● If you pay per appointment, you get appointments.
- Ridgway (1956) It's the oldest management evidence that a single-number target pulls effort toward whatever's easy to count.
- Muller (2018) It's the book-length case for using metrics alongside judgment instead of in place of it.
- Manheim & Garrabrant (2018) It gives you the vocabulary to diagnose which failure your cost-per-lead target set off.
- The McNamara fallacy (Yankelovich, 1971) The first audience for this was a room of sales executives.
- Holmström (1979) Put your pay on signals that show effort toward value instead of on whatever is cheapest to count.
- Holmström & Milgrom (1991) ● When you pay hard on lead volume, effort gets pulled away from lead quality, which nobody measures.
- Baker (1992; 2002) Cost per lead is low-risk and badly distorted. Value measures are noisier, but they point the right way.
- Oyer (1998) Your quota cliffs change when revenue lands, not just how much.
- Larkin (2014) This puts a hard number on what your accelerators cost you.
- Misra & Nair (2011) It's field evidence that the structure of your plan moves revenue.
- Chung, Steenburgh & Sudhir (2014) This is the counterpoint. Well-designed quotas and bonuses do work, and different parts of your plan move different reps.
- Kishore, Rao, Narasimhan & John (2013) This is the multitask problem, caught in real field data.
- Bevan & Hood (2006) It's a documented case of an organization damaged by its own targets.
- Wells Fargo (2016–17) It's the public, documented extreme of what a volume target can do.
Open Metrics and incentives: every work with its citation and what it found
Theme 8 of 10 · 6 works
Matching and routing
Who should meet whom?
Who should meet whom? When your best leads should go to your best people, and what you give up when you route them by speed or availability.
Every work in this theme
- Becker (1973) ● If your great rep is worth more on a great lead than on a poor one, pair them up. Close rate × lead value is multiplicative, and that's the definition of complements.
- Gale & Shapley (1962); Nobel 2012 Matching is a solved problem you can design, whether it's territories, account assignment or routing, and your design choices decide who comes out ahead.
- Gans, Koole & Mandelbaum (2003) Read this before you challenge a speed-to-answer target. It's the base reference.
- Armony (2005) Routing built for speed loads up your fastest agents. It optimizes waiting and leaves value out.
- Armony & Ward (2010) Routing to whoever's been "longest idle" is the fairness default, and it has a measurable cost.
- Mehrotra, Ross, Ryder & Zhou (2012) It's the closest published evidence on routing by ability. You give up a little speed for better outcomes, and a smart hybrid rule gets you most of both.
Open Matching and routing: every work with its citation and what it found
Theme 9 of 10 · 4 works
Humans, algorithms and who decides
Who should set the objective?
An algorithm is only as good as the objective someone hands it. When algorithms beat human judgment, when people overrule them anyway, and what happens when a model is aimed at the wrong target.
Every work in this theme
- Hoffman, Kahn & Li (2018) Managers who overrode the test hired worse. Your job is choosing the objective and validating the score, not overriding it on instinct.
- Kleinberg, Lakkaraju, Leskovec, Ludwig & Mullainathan (2018) It's a template for testing a score against human judgment when you only see outcomes for the accounts you approved.
- Dietvorst, Simmons & Massey (2015) Expect your reps to drop a lead score after one visible miss, and plan for it.
- Obermeyer, Powers, Vogeli & Mullainathan (2019) ● It's the cleanest published example of the wrong objective. The algorithm worked, but it was pointed at a convenient proxy, the same mistake as optimizing cost per lead instead of value.
Open Humans, algorithms and who decides: every work with its citation and what it found
Theme 10 of 10 · 7 works
Correlation and causation
When does a number that moves with revenue actually move it?
When two numbers move together they look like a lever. That’s the trap. Nonsense correlations, Bradford Hill’s deciding question, and why a dashboard can’t tell you what would happen if you pushed.
Every work in this theme
- Yule (1926) If two of your metrics both trend up over the year, or both peak in Q4, they'll correlate whether or not one has anything to do with the other.
- Hill (1965) ● Hill's decisive question is the lever question. Before you push on a KPI, ask whether changing it would change revenue.
- Messerli (2012) A strong, significant correlation between averages tells you nothing about what happens when you pull one lever.
- Vigen (2015) Data dredging finds coincidences, so if you search enough dashboards you'll always find a "driver."
- Lewis, Rao & Reiley (2011) The people who see your ads are busier online in general, so they search, click and buy more whether or not the ad worked. Only a holdout can separate the two.
- Shmueli (2010) Your lead score can predict value well without anything in it being a lever. So use it to sort, and test before you steer with it.
- Pearl & Mackenzie (2018): the ladder of causation Your dashboards live on the first rung. Deciding where to spend is a second-rung question, and first-rung data can't answer it on its own.
Open Correlation and causation: every work with its citation and what it found
Required reading
Books every revenue team should read
Before any of the research, these are the books that teach the craft. For each one, what you’ll take from it and where it shows up in your numbers.
Shelf 1 of 6 · 3 books
Where it started
The idea that a response can be measured, and that an ad is a salesman whose results can be counted.
- Wilhelm Wundt · 1873–74 Principles of Physiological Psychology What a Click Actually Measures In 1874 Wilhelm Wundt set out to make psychology an experimental science. Every time you test a headline or a landing page, you're using his idea that a human response can be measured.
- John E. Kennedy · 1905 Reason-Why Advertising Make Sure the People You Reach Can Actually Buy John E. Kennedy called advertising salesmanship on paper. In the same short book he warned advertisers against paying to reach people who could never buy, and laid out a two-city test judged on goods sold.
- Claude C. Hopkins · 1923 Scientific Advertising Count Customers, Not Replies In 1923 Claude Hopkins argued that advertising had become a science, because a keyed ad could be traced, counted and tested. In the same chapter he warned that some replies are worthless, and that warning is the half most teams skip.
Shelf 2 of 6 · 4 books
Copy that sells
How the people who wrote for a living turned attention into a sale, and tested their way there.
- Eugene M. Schwartz · 1966 Breakthrough Advertising The Desire You Aim At Decides Who Shows Up Eugene Schwartz's argument in 1966 was that the want is already out there, shared by millions of people, and an ad's job is to find it and aim it at a product. It's also where the levels of customer awareness come from.
- David Ogilvy · 1983 Ogilvy on Advertising The Only Way to Know Which Ad Sells David Ogilvy built one of the great agencies on two convictions. An ad exists to sell, and you find out what sells by research, not by taste. Ogilvy on Advertising is his late-career summary of both.
- Joseph Sugarman · 2007 The Adweek Copywriting Handbook Every Line's Job Is to Get the Next One Read Joseph Sugarman sold everything from pocket calculators to BluBlocker sunglasses with ads he wrote himself. His handbook is about as plain an account as you'll find of how copy carries a reader from the headline to the order.
- Michael Masterson & John Forde · 2011 Great Leads Start From What Your Reader Already Knows Michael Masterson and John Forde reduced the opening of a sales message to six kinds of lead, ordered from the most direct to the most indirect. The right one depends on how aware your prospect already is.
Shelf 3 of 6 · 2 books
Why people say yes
The shortcuts and stories buyers use to decide, and what that means for the signals you send.
- Robert B. Cialdini · 1984 Influence: The Psychology of Persuasion Why the Turquoise Sold at Twice the Price Robert Cialdini spent nearly three years inside sales and advertising training to learn why people say yes. His principles explain a lot of marketing, and a few of his stories explain a lot about signals and quotas.
- Seth Godin · 2005; blog since 2002 All Marketers Are Liars, and his blog Decide Who It's For Seth Godin's All Marketers Are Liars made the case that people buy the story they already believe. His daily blog has been making the case for permission, trust and the right small audience for more than twenty years. If you only read one, make it the blog.
Shelf 4 of 6 · 4 books
How the best reps sell
What the research on top performers and big deals says about who sells, to whom, and what they get paid for.
- Matthew Dixon and Brent Adamson · 2011 The Challenger Sale Your Best Reps Sell Differently, Not Harder Matthew Dixon and Brent Adamson studied more than 6,000 sales reps and sorted them into five types. Every type showed up among the average reps, and one type dominated the stars.
- Neil Rackham · 1988 SPIN Selling Big Deals Need a Different Sale Neil Rackham's team studied more than 35,000 sales calls and found that the techniques built for small sales, closing above all, fail in large ones. A sales process tuned on your many small deals is tuned to the wrong end of your book.
- Chet Holmes · 2007 The Ultimate Sales Machine Go After the Buyers Who Buy the Most Chet Holmes's The Ultimate Sales Machine is a twelve-strategy playbook, and two of its chapters cover the Dream 100, a short list of the best buyers in your market, worked relentlessly until they buy. It's power law targeting, written for salespeople.
- Mark Roberge · 2015 The Sales Acceleration Formula Pay Your Reps for the Customers Who Stay Mark Roberge built HubSpot's sales team from its first hire to more than $100 million in run-rate revenue, and he ran hiring and comp the way an engineer would. The part to copy is the comp plan, which moved from paying on new signups to paying on customers who stayed.
Shelf 5 of 6 · 4 books
Keeping the right customers
Which customers to keep, what keeping them is worth, and why the average customer is the wrong one to build around.
- Peter Fader · 2011 Customer Centricity Find Your Best Customers and Build Around Them Wharton's Peter Fader argues that not every customer deserves your company's best effort. Customer centricity means finding the customers worth the most, sorting them by customer lifetime value, and building the business around them.
- Peter Fader, Bruce Hardie & Michael Ross · 2022 The Customer-Base Audit Know How Many Customers Carry Half Your Revenue Peter Fader, Bruce Hardie and Michael Ross set out the customer-base audit, a regular review of what your customers actually buy. It answers the questions most leadership teams can't, starting with how different your customers are and how each cohort behaves over time.
- Sunil Gupta & Donald R. Lehmann · 2005 Managing Customers as Investments Spend on Customers the Way You'd Invest in Any Asset Columbia's Sunil Gupta and Donald Lehmann argue that customers are assets, worth the present value of the profit they'll bring in, so the money you spend to win and keep them is an investment. Price every acquisition and retention dollar against what that customer is expected to be worth.
- Larry Selden & Geoff Colvin · 2003 Angel Customers & Demon Customers Run the Business as a Portfolio of Customers Columbia finance professor Larry Selden and Fortune's Geoff Colvin argue that every company has customers who build its value and customers who destroy it, and most managers can't say which are which. Their answer is to manage the company as a portfolio of customers, with someone accountable for the profit of each segment.
Shelf 6 of 6 · 2 books
Reading the numbers
Customers as a database, and the outliers that the averages hide.
- Kevin Hillstrom · 2006 Hillstrom's Database Marketing, and MineThatData Know Who Buys Again Kevin Hillstrom has spent his career treating customers as a database to be read: who buys again, who doesn't, and what a campaign really changed. His 2008 email test is still a standard teaching dataset.
- Nassim Nicholas Taleb · 2007 The Black Swan You Don't Have an Average Business Nassim Nicholas Taleb's The Black Swan is about rare events with huge consequences, and about the world where a single observation can outweigh all the rest. Customer revenue lives in that world.
The Nobel lineage
Four Nobel Prizes behind the same idea
The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel has gone to work in this library four times. The citations below are from nobelprize.org.
Often misquoted
Lines you’ve heard quoted, and what was actually said
-
"When a measure becomes a target, it ceases to be a good measure." (attributed to Goodhart)
That sentence is Marilyn Strathern's, from 1997. Goodhart's own 1975 wording: "Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes." The source
-
Pareto's 80/20 rule
Joseph Juran named the "Pareto principle," and later wrote that he had "mistakenly applied the wrong name." Concentration also varies by industry: closer to 60/20 in some packaged goods. The source
-
Wald told the Air Force to armor the places without bullet holes
The popular version is embellished. Wald's documented work estimated aircraft vulnerability from the planes that survived. The source
-
Keep 5% more customers and profits rise 25–95%
Reichheld and Sasser's figures are case examples from one article (their chart reads 25% to 85%), not a general law. The source
-
Costly signals are always honest (the peacock's tail)
The "handicap principle" is now openly disputed. Spence's version, a signal works when it costs the wrong sender more, is the safer ground. The source
-
Call a lead within an hour and you're seven times as likely to qualify it
Widely quoted from a 2011 study that compared firms. It shows leads go cold fast, but it doesn't prove cause and effect. The source
Get the week's articles and research in one email
The Power Law Performance newsletter puts this research to work on your numbers, where volume KPIs pick the wrong winner and the signals that sort value from volume.