Craft an Irresistible Offer: Make "Do Nothing" the Expensive Option

Mechanism: people tax "later" and "I have to do it myself" brutally. Waiting discounts a future benefit more than proportionally, and effort itself is scored by the brain as a real cost. So shortening the wait and taking work off them raises perceived value out of thin air — without touching your margin.

• Behavioural economics · the discount for waiting isn't linear (well supported): Green & Myerson (2004, Psychological Bulletin) review a large delay-discounting literature and find discounting is closer to hyperbolic — a short delay near the present destroys far more value than the same delay far out. Which means "a first visible result on day seven" changes how it feels more than 20% off does.
• Neuroscience · immediacy gets scored separately (hard evidence, contested reading): McClure et al. (2004, Science) found that when an option is available right now, dopamine-projection areas including the ventral striatum and medial prefrontal cortex are more active, while delayed choices lean more on dorsolateral prefrontal and parietal regions. Honest labelling: Kable & Glimcher (2007, Nature Neuroscience) explain similar data with a single valuation system, so the two-system reading remains disputed; the direction is what's solid. (cross-ref neuroscience, reward and discounting)
• Cognitive neuroscience · effort is a subtraction (hard evidence): Westbrook & Braver (2015, Neuron) review how the cost of cognitive effort is computed, pointing to striatal dopamine and the anterior cingulate jointly encoding "is this worth the trouble"; Kool et al. (2010, JEP: General) showed people systematically avoid the more effortful option even when the payoff is identical. The forms they have to fill in and the things they have to learn are real deductions.
• Large-scale field evidence · one step can be worth an order of magnitude: Madrian & Shea (2001, QJE) studied a company that switched its retirement plan from opt-in to automatic-enrolment-with-opt-out; participation went from about 37% to about 86%. Johnson & Goldstein (2003, Science) found organ-donation consent rates differ enormously by country default. The plan didn't change — only whether you had to act yourself.

Boundary: all four are perceptions, not promises. Shortening the wait means pulling the first real result forward, not shrinking the timeline in the telling; reducing their effort means you genuinely absorb that work and genuinely spend the hours. If you can't, leave that term alone — the one irreversible mistake in an offer is promising something you can't deliver. Budget ceilings are also genuinely hard: if you're above theirs, don't grind (the qualifying issue covers how to ask early).

Mechanism: a vague promise can't be evaluated, so it gets discounted by default; concrete items can be pictured, checked, and repeated to someone else. The same mechanism cuts the other way — people tend to average a set rather than sum it, so any weak item drags the whole thing down.

• Social psychology · adding a weak item lowers the total (hard evidence, the most counterintuitive line here): Weaver, Garcia & Schwarz (2012, JPSP) call it the "presenter's paradox": add a middling item to a set of strong ones and the audience's overall evaluation drops, while presenters uniformly expect it to rise (consistent across product bundles, résumés and arguments). The conclusion is blunt: delete the weakest line and the offer gets stronger.
• Consumer behaviour · the jump to "free" is disproportionate (reliable, don't abuse): Shampanier, Mazar & Ariely (2007, Marketing Science) found that dropping a price from a tiny positive number to zero produces a jump in choice far larger than the price difference — zero isn't "very cheap," it's a different category. So "that step is ours, no extra charge" beats "a discount on that step." Boundary: free also invites carelessness, so it's safest applied to reducing their effort (we'll do the migration).
• Cognitive psychology · what forms easily in the mind is more readily believed (real but modest): Reber & Schwarz (1999, Consciousness and Cognition) found that merely making a statement easier to read raised the rate at which it was judged true; Alter & Oppenheimer (2009, PSPR) review this family of fluency effects. That's why "the cash-flow sheet by Wednesday" is more credible than "improved financial visibility": one can be pictured, the other can't. (Same mechanism as the concreteness in the make-it-stick issue.)

Boundary: stacking usually fails not from too few items but from items that aren't theirs — before adding a line, ask which sentence of theirs it answers, and delete it if you can't say. The second failure is stacking up delivery debt: every line in the annex has to actually happen, and four you can do beat eight you can't. If you can't find four, you haven't yet worked out where they're stuck — go back and ask.

Two layers. First, uncertainty is itself penalised: unknown probabilities repel more than equivalent known risk, so removing one uncertainty raises perceived value even when the upside is unchanged. Second, signalling — Spence (1973, QJE) showed only low-cost parties can afford an expensive signal, which makes the guarantee itself proof of credibility: the fact that a competitor can't write that sentence is information.

• Neuroscience · when the odds are undefined, the brain treats it as more dangerous (hard evidence): Hsu, Bhatt, Adolphs, Tranel & Camerer (2005, Science) found that ambiguity (unknown probability), as opposed to known risk, raised activity in the amygdala and orbitofrontal cortex and weakened the value signal in the striatum — and patients with orbitofrontal lesions were largely insensitive to ambiguity. That's why "here's the number, here are the conditions" beats "don't worry": you're converting ambiguity into known risk. (cross-ref neuroscience, amygdala and uncertainty)
• Behavioural economics · the do-nothing default carries its own advantage (well supported): Samuelson & Zeckhauser (1988, J. Risk and Uncertainty) established status quo bias across experiments and real retirement-plan choices: labelling an option as the status quo significantly raises how often it's chosen. Your competitor was never the other quote — it's doing nothing.
• Behavioural economics · once it's theirs they won't give it up (hard evidence): the mug experiments of Kahneman, Knetsch & Thaler (1990, JPE) showed people demand markedly more to give up what they already hold than they'd pay to acquire it. That's the real force behind "use it first, pay later": when the trial ends, stopping is already a loss.
• Retail evidence · leniency raises purchases and returns (meta-analysis): Janakiraman, Syrdal & Freling (2016, Journal of Retailing) found more lenient return policies are associated with higher purchase rates and also with higher return rates; the net effect depends on category. Honest labelling: "hardly anyone claims a money-back guarantee" is mostly self-reported industry data and unreliable — price as if it will be claimed.

Boundary: a guarantee you can't carry is borrowing, at the worst rate available. Three rules: one, assume the claim rate is three times your estimate and only write it if you'd still survive; two, tie the guarantee to action on their side ("followed the agreed steps and still no X"), or you'll attract precisely the people who never intended to do the work (adverse selection); three, businesses with huge one-off delivery cost (custom builds, physical manufacturing) shouldn't play with full refunds — use staged payments instead. And the more specific the wording, the more force it carries: "refund if you're not happy" is far weaker than "if you've followed the steps and don't have X by day 30, you get every penny back."

Mechanism: price has no absolute scale, only relative position. Estimating an unfamiliar quantity, people start from whatever number is at hand and adjust back — and the adjustment is always insufficient, so whatever appeared first drags the judgment with it. Same with a set of options: they don't evaluate "is it worth it," they pick one from the set.

• Cognitive psychology · the original anchoring experiment (classic, direction very stable): Tversky & Kahneman (1974, Science) had participants spin a wheel for a random number before estimating the share of African countries in the UN — those who landed on a high number estimated systematically higher. Even an obviously irrelevant number drags the judgment.
• Field · professionals aren't immune either (hard evidence, closest to selling): Northcraft & Neale (1987, OBHDP) showed students and practising real-estate agents the same house, varying only the listing price: both groups' valuations were systematically pulled by it, and the agents generally insisted they hadn't been influenced. Which applies to you too: the first number they name is dragging you, so work out your floor before you walk in.
• Consumer behaviour · adding an option changes the share of the first two (peer reviewed): Simonson & Tversky (1992, JMR) demonstrated the compromise and attraction effects — introducing a third option significantly shifts choice shares between the original two. That, not folklore, is the basis for three tiers. Honest labelling: the much-quoted Economist subscription decoy case comes from Ariely's book demonstration, not a peer-reviewed experiment.
• A note on strength: the famous paradigm in Ariely, Loewenstein & Prelec (2003, QJE), where the last digits of an ID number anchor bids, has an inconsistent replication record (Fudenberg, Levine & Maniadis 2012 did not reproduce effects of the same magnitude). The direction is reliable, but don't expect a casual number to set a price — the sturdiest anchor is the one they work out themselves.

Boundary, and the law: inventing a "was" price and crossing it out is fraud, not anchoring — most jurisdictions explicitly prohibit fabricated reference prices. Only three anchors are real: the status quo cost they calculated themselves, a genuinely available higher tier that people genuinely buy, and a publicly verifiable price for the alternative. As for urgency, the mechanism is covered in the six-principles-of-influence issue (loss aversion plus reactance when freedom is restricted); the offer-specific rule is one line: a deadline must be tied to a real constraint you can say out loud (the schedule, a price rise, a limited slot), and you should explicitly allow them to take the later option. A deadline with no stated reason reads as invented, and usually is.