Bargaining Theory for Enterprise Deals: Outside Options, Patience, and the Split-the-Difference Fallacy

TL;DR: Negotiation training sells tactics. Bargaining theory says the tactics arrive too late: the split is largely fixed before anyone speaks, by three things, who is more patient, whose outside option actually binds, and who faces the deadline. Nash (1950) derived the split from four axioms; Rubinstein (1982) derived the same split from a non-cooperative model of alternating offers; and Binmore, Rubinstein and Wolinsky (1986) showed the two converge, with the more patient party taking the larger share. Experiments complicate the picture. Ultimatum offers cluster near an even split across four countries (Roth et al. 1991), and an outside option moves the price only when it exceeds a party's equilibrium share (Binmore, Shaked and Sutton 1989). The field is messier still: across millions of eBay threads people split the difference even when theory says they should not, and more patient buyers pay 5.7 percentage points less (Backus et al. 2020); real bargaining leaves 12 to 23 percent of the gains from trade unrealized (Larsen 2021). For enterprise sales and procurement this reframes the quarter-end discount. The discount is not a tactic the seller chose; it is the buyer collecting the seller's impatience, priced by the sales quota. What an operator should change is the sequence and the timing, not the talking points.

---

The Deal Was Over Before the Call

A vendor's account executive dials into the renewal call for a six-figure contract with three days left in the quarter. The buyer's procurement lead has blocked forty-five minutes. Both sides have read the same negotiation book. Both have a list of tactics: anchor high, never make the first concession, use silence, trade nothing without getting something. Forty minutes later the seller has given up 22 percent on price and thrown in a quarter of professional services for free, and both sides describe the outcome as a hard-won compromise.

Almost none of what happened in those forty minutes explains the outcome. The 22 percent was set weeks earlier, by facts neither party discussed: the seller's quota closed on Friday and the buyer knew it; the buyer had a credible bid from a rival that would clear the switching cost; and the buyer could run the incumbent tool for another two quarters at no marginal cost while the seller could not run payroll on a deal that slipped. The tactics were theater performed on top of a result that patience, outside options, and the calendar had already written.

The claim is uncomfortable, and it is not a metaphor but a theorem. Bargaining theory, the branch of game theory that studies how two parties divide a surplus they can only realize together, says the division is a function of structural parameters that are mostly locked before the conversation starts. Nash proved it one way in 1950. Rubinstein proved it another way in 1982. The two proofs, built on completely different foundations, land on the same split. When the applied literature went looking for the theory in the wild, it found it half-confirmed and half-subverted in an instructive pattern that tells an operator exactly which levers are real.

We will build the argument in order: the theory taught properly, the outside-option result that most negotiators get backwards, the experimental and field evidence, the places where the models break, and then the part a founder or a procurement lead can act on. The through-line is a single reframing. In an enterprise deal, the discount is rarely a concession; the discount is a price the structure charges, and structure can be changed long before anyone opens their mouth.

The Theory, Taught Properly

Two people can jointly create value that neither creates alone. A buyer values a software platform at more than the seller's cost to deliver it; the difference is the surplus, or the gains from trade. Both prefer any agreement in that range to no deal. The question bargaining answers is not whether they trade, they should, but how they divide the surplus. Everything interesting lives in that division.

There are two ways to model it, and the history of the field is the story of making them agree.

Nash's axioms and the product that solves them

Nash did not model the haggling. In "The Bargaining Problem" (1950, Econometrica 18(2), 155-162) he asked a cleaner question: if a division is to be reasonable, what properties must it have? He wrote down four, and then proved that exactly one outcome satisfies all four.

The four axioms are worth stating in plain language, because each one corresponds to something an operator already believes. Pareto efficiency: do not leave money on the table; if both sides can do better, they will. Symmetry: if the two parties are interchangeable, same patience, same alternatives, they split evenly. Invariance to affine transformations: the answer cannot depend on the units in which we measure each side's happiness, so rescaling one party's payoff scale changes nothing real. Independence of irrelevant alternatives: removing options that neither side would have chosen does not change the chosen split.

From those four, Nash proved the solution maximizes the product of the two parties' gains over their disagreement values. Let d₁ and d₂ be what each side gets if there is no deal, the disagreement point, and let S be the set of feasible payoff pairs. The Nash bargaining solution is:

$$(u_1^{*}, u_2^{*}) = \arg\max_{(u_1, u_2)\, \in\, S,\; u_1 \ge d_1,\; u_2 \ge d_2} \; (u_1 - d_1)\,(u_2 - d_2)$$

The object being maximized, (u₁ - d₁)(u₂ - d₂), is the Nash product. Two features of it shape everything that follows. First, the disagreement point dᵢ is not decoration, it is subtracted from each side's payoff before the product is taken, so a party whose walk-away is better captures a larger share of the surplus. Raise your dᵢ and the maximizer shifts toward you. Second, the solution is symmetric in the two gains, so with equal disagreement points and equal stakes the split is even. The Nash product is the mathematical form of "your bargaining power is your next-best alternative," a sentence every negotiator has heard and few have priced.

Nash's result is an existence-and-uniqueness theorem about what a fair-and-efficient split looks like. What it deliberately omits is the process, no offers, no counteroffers, no clock. A skeptic in 1950 could reasonably ask why any real negotiation, with its bluffs and deadlines, should land on an axiomatic ideal. Answering that took thirty-two years.

Rubinstein's alternating offers and the price of impatience

Rubinstein modeled the haggling directly. In "Perfect Equilibrium in a Bargaining Model" (1982, Econometrica 50(1), 97-109) two players divide a pie by alternating offers: player one proposes a split, player two accepts or rejects; on rejection player two proposes, and so on, potentially forever. The only friction is time. Each player discounts the future by a factor δᵢ ∈ (0,1) per round, a dollar next round is worth δᵢ dollars now, so delay shrinks the effective pie for the impatient.

The model has a unique subgame-perfect equilibrium, and it resolves on the very first offer, no delay, no haggling, agreement in round one. The first mover proposes exactly the split that leaves the responder indifferent between accepting now and waiting to counter, and the responder accepts. Solving that indifference condition gives the shares:

$$x_1^{*} = \frac{1 - \delta_2}{1 - \delta_1 \delta_2}, \qquad x_2^{*} = \frac{\delta_2\,(1 - \delta_1)}{1 - \delta_1 \delta_2}$$

We should read the first equation slowly, because it is the most useful sentence in the theory. Player one's share rises as δ₂ falls, the more impatient the other side, the more you get. Your own patience δ₁ helps you through the denominator, but the dominant term in the numerator is the other party's impatience. Bargaining power, in the one model that derives it from first principles rather than assuming it, is almost entirely about who can afford to wait.

Two special cases make the formula concrete. If the two sides share a common discount factor δ₁ = δ₂ = δ, the first mover's share collapses to x₁* = (1)/(1+δ). At δ = 0, a party that cannot bear even one round of delay, the first mover takes the whole pie, which is just an ultimatum. As δ → 1, patience becomes symmetric and the share falls to one half. The first-mover advantage is real but it is entirely an artifact of friction; remove the friction and it vanishes.

The chart is not an approximation of anything empirical; it is the exact formula (1)/(1+δ) plotted against δ. What it shows is that the entire drama of "who moves first", the thing negotiation coaching obsesses over, is worth almost nothing once both parties can tolerate delay. At δ = 0.95, roughly a party that loses 5 percent of value per round of delay, moving first is worth about 1.3 percentage points of the pie. The lever that matters is on the horizontal axis, not in the turn order.

Loading diagram...

The bridge, and the limit that unifies the two theories

For thirty-two years Nash's axiomatic split and Rubinstein's strategic split were two different objects that happened to look alike. Binmore, Rubinstein and Wolinsky closed the gap. In "The Nash Bargaining Solution in Economic Modelling" (1986, RAND Journal of Economics 17(2), 176-188) they showed that as the time between offers shrinks to zero, the Rubinstein equilibrium converges to the Nash bargaining solution, the asymmetric version, weighted by relative patience. Model the discount factors as δᵢ = e⁻^r^_ⁱ^ ^Δ, where rᵢ is party i's interest rate and Δ is the interval between offers, and take Δ → 0:

$$\lim_{\Delta \to 0} x_1^{*} = \frac{r_2}{r_1 + r_2}$$

The share goes to the ratio of the other party's impatience to total impatience. A buyer half as impatient as the seller (r₁ = (1)/(2) r₂) captures two-thirds of the surplus. The convergence completes Nash's program, his own 1953 proposal that a cooperative solution should be derivable from a fully specified non-cooperative game. The axiomatic answer and the strategic answer are the same answer, and the weight that turns the symmetric Nash product into the real-world asymmetric one is relative patience.

Sutton's survey, "Non-Cooperative Bargaining Theory: An Introduction" (1986, Review of Economic Studies 53(5), 709-724), laid out this program for a generation of economists and is still the cleanest single entry point. The predecessor worth naming is Ståhl (1972), whose finite-horizon alternating-offer model Rubinstein extended to the infinite horizon that removed the artificial last-round advantage.

Outside Options Only Matter When They Bind

Here is where most negotiators, and most negotiation books, go wrong. The folk theory says a better alternative always strengthens your hand, so you should always develop and always brandish your outside option. The real theory says something sharper and, at first, counterintuitive.

Binmore, Shaked and Sutton tested it in "An Outside Option Experiment" (1989, Quarterly Journal of Economics 104(4), 753-770). Their result, now called the outside-option principle, is that an outside option changes the equilibrium split only if it exceeds what you would have gotten without it. Below that threshold the option is inframarginal, it exists, it is real, and it changes nothing. Above it, the option does not merely tilt the split; it pins your share to exactly the option's value plus a sliver, and no more.

The distinction between an outside option and a discount factor is the part to internalize. Patience operates continuously, a little more patience buys a little more share, everywhere on the curve. Unlike patience, an outside option operates as a threshold, it does nothing until it clears your no-option share, then it takes over entirely. Confusing the two produces a specific, expensive mistake: improving an alternative that was never going to bind, and paying for the improvement in time and money while the split does not move an inch.

BATNA is the outside option, and where the book disagrees with the theory

The practitioner name for the outside option is BATNA, the Best Alternative To a Negotiated Agreement, from Fisher and Ury's Getting to Yes (1981). The book did the field an enormous service by naming the concept and telling every reader to develop theirs before walking in. On the core point, the book and the theory agree completely: your power is your alternative, not your rhetoric.

The book and the theory part company on two things, and the gaps matter. First, Getting to Yes treats a better BATNA as monotonically good, develop it, improve it, and your position strengthens. The outside-option principle says improvement below the binding threshold is wasted, and improvement is only worth what it takes you past that threshold. Second, the book is nearly silent on patience, which the theory says is the other half of the answer and often the larger half. A buyer with a mediocre alternative but infinite patience beats a seller with a strong alternative and a Friday deadline. Getting to Yes would coach both sides to improve their alternatives; the theory says the seller's problem is the calendar, and no alternative fixes a calendar.

The Evidence, In Order

Theory this clean invites a test. When economists and psychologists ran the tests, they found the structural results largely intact and the behavior stubbornly human in a way that changes the operator's playbook. The evidence arrives in three layers: the laboratory ultimatum, the anchoring literature, and the field.

Ultimatum and dictator games: fairness shows up uninvited

The simplest bargaining game is the ultimatum. Güth, Schmittberger and Schwarze introduced it in "An experimental analysis of ultimatum bargaining" (1982, Journal of Economic Behavior and Organization 3(4), 367-388). One player proposes a split of a fixed sum; the other accepts, in which case it stands, or rejects, in which case both get nothing. The subgame-perfect prediction is brutal and clear: offer the smallest possible positive amount, because a rational responder prefers a penny to nothing. Almost nobody plays that way. Proposers offer far more than a penny, and responders reject low offers, choosing zero over a humiliating split.

The finding is not a fluke of one subject pool. Roth, Prasnikar, Okuno-Fujiwara and Zamir ran the game across four countries in "Bargaining and Market Behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: An Experimental Study" (1991, American Economic Review 81(5), 1068-1095). Modal offers sat near an even split everywhere, with means ranging from roughly 0.36 in Jerusalem to about 0.45 in Pittsburgh and Ljubljana. The cross-country differences in what proposers offered were real, but the relationship between offer size and rejection probability was similar across all four cities, proposers who offered less did so where responders accepted less, which points at differing fairness norms rather than differing rationality.

The gap between the modal offer, near 0.50, and the mean, dragged down toward 0.36 in Jerusalem, is the signature of a left-skewed distribution: most proposers cluster at the even split while a minority push lower and pull the average down. Camerer's synthesis in Behavioral Game Theory (2003) pools the accumulated evidence: across many ultimatum studies, mean offers land around 0.40 to 0.45, offers near an even split are almost never rejected, and offers below roughly one-fifth of the pie are rejected about half the time. The dictator game, same split, but the responder cannot reject, cuts giving sharply, which tells us the ultimatum's generosity is not pure altruism but partly the proposer pricing in the responder's willingness to burn the pie out of spite.

Anchoring: the first number bends the last one

Between structure and outcome sits a cognitive lever the pure theory ignores. Galinsky and Mussweiler, in "First offers as anchors: The role of perspective-taking and negotiator focus" (2001, Journal of Personality and Social Psychology 81(4), 657-669), showed that the party who makes the first offer pulls the final agreement toward it, and that the effect is large enough to swamp small differences in skill. The first number sets a reference point that both sides then adjust from insufficiently.

Anchoring and Rubinstein's first-mover advantage point in opposite directions, and the tension is instructive. The equilibrium first-mover advantage shrinks to nothing as patience rises; the anchoring first-offer advantage does not, because it works through cognition rather than through the discount factor. In a world of perfectly rational, patient agents, who speaks first barely matters. Yet in the world we negotiate in, the first credible number matters a great deal, not because it changes the structural split, but because it changes each side's estimate of where the structural split is. The lesson is not "the theory is wrong." The lesson is that the anchor is fighting for the gap between the true equilibrium and each party's belief about it.

The field: people split the difference even when they shouldn't

Laboratory games are small, brief, and artificial. The strongest test of bargaining theory came when two economists got data on millions of real negotiations. Backus, Blake, Larsen and Tadelis studied eBay's Best Offer platform in "Sequential Bargaining in the Field: Evidence from Millions of Online Bargaining Interactions" (2020, Quarterly Journal of Economics 135(3), 1319-1361), covering tens of millions of listings where buyers and sellers exchange offers.

Three findings matter for us. First, the structural prediction about patience holds in the wild: buyers who revealed patience by choosing the slowest shipping paid measurably less. For used goods, patient buyers obtained final prices 5.7 percentage points lower relative to a reference price, after controlling for experience, with no significant effect for new goods where the market price is well defined and there is little to bargain over. Patience is not a metaphor here; it is 5.7 points of margin.

Second, and against the theory, people split the difference. The modal buyer opening offer is half the listed price, and the modal counteroffer sits at the midpoint between the previous two offers. Rubinstein's model has no midpoint anywhere in it, the equilibrium split depends on discount factors, not on the arithmetic mean of the last two numbers on the table. Yet the equal split is a mass point that shows up again and again, which is why the authors connect their field data to the laboratory tradition on equal division. The even split is a focal point, in Schelling's sense: a number both sides can coordinate on without justifying, precisely because it needs no justification.

The power to constrain an adversary may depend on the power to bind oneself.

, Thomas Schelling, The Strategy of Conflict (1960)

Third, round numbers leak information. Backus, Blake and Tadelis, in the companion study "On the Empirical Content of Cheap-Talk Signaling: An Application to Bargaining" (2019, Journal of Political Economy 127(4)), showed that offers stated in round figures behave like cheap talk that signals flexibility: round-number listings draw larger concessions from the other side, and precise numbers hold firmer. A price of $10,000 says "I have not thought hard about this and I will move"; a price of $9,850 says "this number is load-bearing." The digits are a message, and most sellers send one without knowing it.

Larsen measured how much money the whole messy process leaves behind. In "The Efficiency of Real-World Bargaining: Evidence from Wholesale Used-Auto Auctions" (2021, Review of Economic Studies 88(2), 851-882), using roughly 265,000 alternating-offer sequences, he found that bargaining is inefficient but not catastrophically so: 17 to 24 percent of pairs with genuine gains from trade fail to reach a deal, and the process leaves 12 to 23 percent of the available surplus unrealized. Real bargaining captures most of the pie and wastes a meaningful slice, and the waste is not fully explained by the information problem that Myerson and Satterthwaite (1983) proved is unavoidable.

The pattern repeats in settings far from software. Keniston, in a 2011 working paper on the auto-rickshaw market in Jaipur, compared bargained prices against posted prices and found the outcomes of haggling track what a structural bargaining model predicts, with the division again shaped by the parties' patience and alternatives rather than by their eloquence. Across a used-car lot in the American Midwest, a rickshaw stand in India, and a software renewal in a video call, the same two parameters keep doing the work.

Where the Models Break

A theory earns trust by being clear about its failures. Bargaining theory has three that an operator should hold in mind, because each one is a place where doing the "rational" thing loses money.

The first is the split-the-difference regularity itself. The equilibrium split is a function of discount factors; the even split is a function of neither party wanting to argue. When both sides are drawn to the midpoint, the disciplined party wins by controlling what the midpoint is anchored on. If the current numbers on the table are $100 and $60, the reflexive split is $80. Move the anchor first, open at $120 rather than $100, and the reflexive split becomes $90. The split-the-difference instinct is not a law of fairness; it is a coordination shortcut, and shortcuts can be redirected by whoever sets the endpoints.

The second failure is that the models assume the surplus is fixed and known. In enterprise deals it is neither. Much of what looks like distributive haggling over a fixed pie is actually poor discovery of a pie that could be grown, a longer term, a wider deployment, a co-marketing arrangement that changes the disagreement points for both sides. Fisher and Ury's best contribution was insisting that parties expand the pie before dividing it, and here the practitioner book is ahead of the classical theory, which took the frontier as given.

The third failure is information. Rubinstein assumes both parties know each other's discount factors and payoffs. Real bargainers do not, and the resulting uncertainty is what produces delay, impasse, and the deals that should close but don't. Larsen's unrealized 12 to 23 percent is the visible residue of two sides guessing wrong about each other. The models that add incomplete information, the tradition running through Myerson and Satterthwaite, explain why even rational, well-intentioned parties leave money on the table. For the operator this reframes disclosure as a lever, not a leak, which is where we turn next.

What This Changes for an Operator

When we move from theory to practice, the work is translation. Every object in the model maps to something concrete in an enterprise deal, and the mapping tells us which levers to pull and, more importantly, which to stop wasting effort on.

The quarter-end discount is the buyer collecting the seller's impatience

The single most reframed object on that list is the discount factor. A software seller's patience is not a personality trait; it is manufactured by the compensation system. Sales quotas reset at period boundaries, and a rep who closes on Friday is paid differently from one who closes the following Monday. Oyer showed the macro version of this in "Fiscal Year Ends and Nonlinear Incentive Contracts: The Effect on Business Seasonality" (1998, Quarterly Journal of Economics 113(1), 149-185): nonlinear incentive contracts create predictable spikes in selling effort and price flexibility around period ends. The mechanics of why quotas are shaped that way belong to the study of incentive design, and a companion essay in this series takes up the principal-agent problem directly; here the only thing that matters is the consequence. At quarter-end, the seller's effective δ drops toward zero, and by the Rubinstein formula, a counterpart facing an impatient opposite number captures more of the surplus.

Which yields the reframe an operator should tattoo somewhere visible: the quarter-end discount is not a tactic the seller deployed. The quarter-end discount is the buyer collecting the seller's impatience, and the size of the collection is set by the quota structure, not by the negotiation. A seller who wants to stop giving away that 22 percent does not need better tactics. The seller needs to change the structure, smooth the quota, build enough pipeline that no single deal is a must-close, and decline to let the buyer put the decision on the quota boundary.

Multi-year deals are patience arbitrage

The trade between price and term is the cleanest patience lever in the enterprise toolkit, and it is usually mispriced. A buyer who commits to three years raises the seller's continuation value, less churn risk, lower future acquisition cost, smoother revenue, and in exchange asks for a lower price. The trade is not a favor either side does the other; it is an exchange of the buyer's patience for the seller's, and it should be priced as one.

The discipline is to set the term discount equal to what the extra patience is actually worth, which is roughly the seller's cost of capital plus its churn-risk reduction, not a round percentage pulled from a slide. If a seller's implied annual cost of impatience is 8 to 12 percent, a multi-year discount larger than that is the seller paying the buyer to accept money early, which is backwards. In practice we see three-year deals discounted at 25 or 30 percent because the number felt appropriately generous, when the defensible figure was closer to half that. The buyer who understands this asks for term length the seller values and pays for it in patience rather than in price.

Commitments work only when they are credible

Some of the strongest bargaining moves are things you do to yourself before the negotiation, to remove your own freedom to concede. A public price list, a no-discount policy, and a most-favored-customer clause are all attempts to make "I cannot go lower" true rather than merely said.

The theory of when such moves work is the theory of credibility, and a separate essay in this series is devoted to commitment and credible threats in pricing, so we will not re-derive it here. The one point that belongs in a bargaining essay is why a most-favored-customer clause makes a no-discount policy credible where a mere announcement does not. An announced no-discount policy is cheap talk; the buyer knows the seller can quietly break it under pressure. A most-favored-customer clause makes a discount to any one buyer automatically extend to others, so a single concession becomes expensive across the whole book, which makes the refusal to concede believable precisely because breaking it is now costly. Schelling's line captures the whole mechanism: the power to constrain the buyer comes from the seller's power to bind itself.

Information: sometimes you should show your constraint

The last lever is counterintuitive and underused: disclosing your own constraint can improve your split. The instinct is to hide everything, but a credible, verifiable constraint functions exactly like an outside option, it moves the equilibrium by changing what the other side believes is feasible.

A buyer with a hard budget cap of $180,000, who can prove the cap is real because it is board-approved and documented, has effectively committed to a disagreement point. The seller who believes the cap is genuine, and who prefers a $180,000 deal to no deal, will meet it, the buyer's disclosed constraint did the work that a bluff could not. The condition is verifiability: disclosure only helps when the other side cannot dismiss it as a negotiating posture. An unverifiable claim of a budget cap is just an opening offer wearing a costume, and experienced counterparts strip the costume off immediately. The parallel to the seller side is the credible pipeline: a seller who can show that walking away is genuinely fine has disclosed a high disagreement point, and the disclosure raises its share.

Run that diagnostic before the next renewal, on both sides of the table, and the negotiation stops being a performance and becomes an audit of a structure you can still change. Who is more patient, whose outside option binds, who faces the deadline, and what the round-number anchor is, those four answers predict the split better than any script. The buyer's outside option is the incumbent renewal or a rival vendor, net of switching cost; the seller's is the rest of the pipeline. Whichever of those is more real, and whichever side owns the calendar, will have written the outcome before anyone says hello. The value the platform actually delivers still sets the ceiling, a point developed in the pricing essays linked below, but within that ceiling, structure divides the pie.

Key Takeaways

1. The split is set before the talking. Nash (1950) derived the bargaining outcome from four axioms and Rubinstein (1982) derived the same split from alternating offers; Binmore, Rubinstein and Wolinsky (1986) proved they converge, with the share going to relative patience, r₂/(r₁+r₂) in the limit.
2. Patience is the dominant lever, and it is priced. In Rubinstein's model the first mover's share is (1-δ₂)/(1-δ₁δ₂), rising as the other side grows impatient; in the field, patient eBay buyers paid 5.7 percentage points less for used goods (Backus et al. 2020).
3. Outside options only matter when they bind. An alternative below your equilibrium share changes nothing (Binmore, Shaked and Sutton 1989), so developing or brandishing a non-binding BATNA is wasted effort, a mistake procurement teams make routinely with rival bids that lose once switching costs are counted.
4. People split the difference even when they should not. The even split is a focal point, not an equilibrium, and it is a genuine mass point in millions of field bargains (Backus et al. 2020); the disciplined party wins by controlling what the midpoint anchors on.
5. The quarter-end discount is the buyer collecting the seller's impatience. Quota structures push the seller's effective discount factor toward zero at period ends (Oyer 1998), so the fix is structural, smooth the quota, deepen the pipeline, and keep the decision off the boundary, not tactical.
6. Round numbers and disclosure are signals. Round-number offers leak flexibility and draw larger concessions (Backus, Blake and Tadelis 2019); a credible, verifiable constraint such as a documented budget cap functions like an outside option and can improve your split.
7. Bargaining captures most of the pie but wastes a real slice: 12 to 23 percent of the gains from trade go unrealized, and 17 to 24 percent of viable pairs fail to trade (Larsen 2021), most of it from each side guessing wrong about the other's patience and alternatives.

Further Reading

• Anchor Pricing and Its Limits, the first-offer anchor at the negotiating table is the same mechanism as the reference price on a pricing page, and it stops working in the same ways.
• Value-Based Pricing Operationalized, the buyer's outside option is their next-best source of the same value, so the split starts where value measurement leaves off.
• Discount Engineering: When Promotions Destroy LTV, what the quarter-end concession does to lifetime value once patience-collection becomes a habit the buyer learns to exploit.
• Switching Cost Engineering and Interoperability, why a rival's bid so often fails to bind as an outside option once migration and retraining are counted in.