What It Is · Key Takeaways · Do's and Don'ts · Pair With · Things to Explore
On 1 November 1995 Slobodan Milošević, Franjo Tuđman and Alija Izetbegović arrived at Ohio’s Wright-Patterson Air Force Base, which they would not leave for twenty-one days. They were housed within walking distance of each other, kept away from the press, and they never convened in joint session. Richard Holbrooke's team shuttled between them. Boundary questions were settled in the map room, where the contested terrain could be virtually flown over on video screens, turning arguments about principles into decisions about specific hilltops. Negotiators worked on military annexes, constitutional provisions and the status of Sarajevo on parallel tracks the entire time. The final framework was initialed twenty minutes before the talks’ scheduled end. Holbrooke did not bring new arguments so much as new structure, as he testified in his memoir To End a War (ch. 16, p. 232): “Since November 21, 1995, ‘Dayton’ has entered the language as shorthand for a certain type of diplomacy—the Big Bang approach to negotiations: lock everyone up until they reach agreement.”
What It Is
Load is the aggregate cognitive and emotional burden the negotiation process imposes at any given moment. It is determined by how many issues are simultaneously alive, and how burdensome each one is. It is the first pattern in the Momentum triad, where Pathos meets Structure. The others are Velocity (6.2), which governs how fast actors have to move and decide, and Reversibility (6.3), which governs how costly it is to change course or reverse a decision. Load is a property of the process. The same issues can be manageable in one structure and unbearable in another. Load is different from Charge (9.1) which produces pressure through style aspects such as tone, confrontation, or physical presence. Load produces it through architecture. A party can be alone, without interpersonal exchange occurring, and still feel overwhelmed by the requirements of a complex RFQ. Confuse the two, and you risk intervening on style while the burden is generated by structure.
Low Load permits deliberation and psychological safety; high Load produces pressure, dependency and forced prioritization. Below a critical threshold, more Load just costs you efficiency: you need to work harder for the same quality. Above the threshold, behavior changes in kind. Analysis gets replaced by heuristics, engagement by avoidance, and evaluation on merit by a compulsion to make it stop. Below the threshold, actors risk mismanaging the deal; above it, themselves.
The problems of high Load are not carelessness. Bandwidth is finite in everyone, but there is no clear signal of where the limit is and taking shortcuts may not register as failure in the moment. When a counterparty overloads you while keeping their own load manageable, it’s not always detectable as foul play. Every request is reasonable, the questions are legitimate, and every individual document is defensible. The asymmetry exists only in the aggregate. It can cost you twice: first in the concessions you made to stop processing, and then on implementation when overlooked but consequential clauses in the agreement fall short. The failure usually gets attributed to bad execution or to sloppy governance, rarely to the process that produced it.
Key Takeaways
I. There’s almost no defense against self-imposed Load
Counterparty-imposed Load can be resisted. If you notice it and call it out, the counterparty must either concede scope or expose the tactic. Not so with Load you install and maintain yourself. It arises when your own side confuses activity with progress. Eager to “be responsive”, and impress and please the customer, sales people mobilize specialists, prepare material and drum up support from their principals. Internal advocates with deliberate and thoughtful questions are chided and lectured on their “sense of urgency”. It feels like winning. It costs the counterparty nothing beyond maintaining the appearance of receptiveness. It generates no resentment, because the burden is seen as underscoring the importance of the opportunity. Defending against this is hard, and the person best placed to do the defending is often least interested in doing so. Leadership can play a role here, critically questioning what concretely is moving in the counterparty’s decision process. For anybody else, it’s difficult and unpopular to step in front of the train.
II. Calmness can’t reduce what tone did not produce
The most common misdiagnosis is to interpret a structural condition as an emotional one. When the negotiation feels heavy, instinct attributes it to the interaction—which admittedly can get strained under the burden. If the problem truly is one of Charge (9.1), emotion produced by the interaction, that’s where the intervention should be. But no amount of warmth will help when the problem is there’s structurally too much to handle. Never mind the tone, what truly helps is a simpler agenda, a shorter issue list or fewer balls in the air. The number of topics, their interdependence and ambiguity, coordination burden… Load shedding can only come from restructuring these. Writing things down is more effective than calming down: a documented item, delegated to a named owner, removes it from working memory.
III. The timing of issues affects their Load
Even with an agenda and issue list in place, negotiators usually don’t decide upfront when in a session an item will be raised, nor how many in succession. It matters. Raised early, a fresh counterparty can analyze it. But raise it late in a long session, at the end of a multi-day process, or immediately after a difficult concession, and the counterparty just wants it to be over.
Early time spent on easy or minor items can create positive momentum, but it consumes bandwidth and time and comes out of a finite reservoir of counterparty capacity. By the time the important items come up, most of it may already be spent. Sequence the high-consequence items into the middle, where there is capacity left and the deadline is still distant. If you don’t control the sequence, decide in advance which items you don’t want to settle at the end, and say so while it is still a scheduling conversation.
IV. Ask who is carrying more before you ask where or who the Load comes from
Absolute Load is a poor diagnostic, because there is no objective reference point. Every serious negotiation is demanding, so the fact that you are working hard tells you nothing. So how can you tell who is carrying more? The observable signals are differences in preparation, in response time, and in escalation patterns, when one side frequently needs to check with principals while the other decides on the spot.
Carrying less Load is a structural advantage. When an asymmetry exists, the instinct is to look for the party that deliberately built it, but often there is none. Most processes are not chosen so much as designed by default around one party’s governance model, systems, review cycles and calendar. If that results in asymmetric burden, it’s not always manufactured on purpose. Still, a counterparty acting in good faith does nothing to reduce the burden. Without jumping to accusations, engage the counterparty to keep Load manageable. Between constructive parties, that should be uncontroversial. If the response is not constructive, you’ll have learnt something.
A Few Do’s and Don’ts
Don’t Read increased effort as proof of progress
High counterparty effort with warm reciprocation is consistent with a live deal. It is equally consistent with being the alternative option they keep open while they build their preferred deal elsewhere. Your own enthusiasm only says something about your own state, but it often gets interpreted internally as a signal about the counterparty’s. Before another wave of resources gets mobilized, establish the facts concerning the counterparty’s decision process: who confirmed what, where does that put us in the process, and what does the next round do? The pattern is the diagnostic: consecutive cycles of substantial effort against no concrete process step is a trigger to make further investment of time and effort conditional.
Do Write all live issues down, and give them an owner
A documented and delegated item does not need to be held in working memory. That frees up capacity. It gives you an overview you can keep track of, avoiding scope creep or misallocation of effort. The harder it is to maintain, the more you should want to, because the higher the need will be. Under high Load, the discipline is usually the first thing to go. But the tracking itself adds to the load, so make sure it doesn’t crowd out the real work.
Don’t Answer breadth with breadth
When a counterparty raises many things at once, the reflex is to answer all of them, thoroughly and quickly. It demonstrates control, and particularly in a client relationship, commitment and responsiveness. All true, but this reflex is the exploitable mechanism to overload. If the other side determines the scope and goes broad, they consume your capacity while preserving theirs. So before answering a question, narrow it down to a specific scope and ask how the answer is relevant to the process. If the breadth persists, call it out: “we’re carrying a lot at once, I’d like to propose we focus on the main ones first.” It’s an easy action that can’t reasonably be refused.
Do Ask about the consequences of a deadline
Deadline pressure is the oldest trick in the book. Don’t accept them without asking what, specifically, will break for whom on the designated date. Many deadlines don’t survive the question very well and get revealed for what they are: internally and somewhat arbitrarily manufactured. Ask factually though. If the question sounds like an accusation rather than a probe, it costs credibility and goodwill. And dismissed too easily as arbitrary or manufactured, you risk being perceived as an unconstructive counterpart who doesn’t understand their constraints well enough.
Pair With
Velocity (6.2). Overload is the product of the two. High load is manageable if there is sufficient time between gates. A fast pace is sustainable if the number of simultaneous items is OK. Neither high Load nor high Velocity alone reliably produces a collapse, but their combination does.
Reversibility (6.3). Load registers higher when decisions are perceived to be irreversible, even for individual items before they get added together. Building explicit reversibility into early, exploratory phases simultaneously reduces Load and increases quality of judgment. An easy structural intervention between parties negotiating in good faith.
Issue Decomposition (5.1). Decomposition converts a dense, highly interdependent problem set into a series of bite-size sub-problems. Insufficient decomposition, and all the complexity needs to be handled at once; too much decomposition and the switching costs cancel out the gains. Rightsizing is a function of the parties’ capacity: a decomposition that is logically correct but cognitively unmanageable is a bad one.
Things to Explore
Book
Pamela Herd & Donald Moynihan, Administrative Burden: Policymaking by Other Means (Russell Sage Foundation, 2018). It is no mean feat to write a book about a conceptual framework for policy making that gets 4.21 on goodreads.com and is described by one reviewer as a “banger”. The main topic is the burden a process puts on the person who gets to experience it, sorted into learning costs, compliance costs and psychological costs. A burden doesn’t have to be deliberately designed to be felt, they argue. Search handles: administrative burden, learning compliance psychological costs.
Article
Anandi Mani, Sendhil Mullainathan, Eldar Shafir & Jiaying Zhao, “Poverty Impedes Cognitive Function,” Science 341 (2013), 976–980. Indian sugarcane farmers were tested before harvest, when money was short, and after it, when it was not. Scores on reasoning and cognitive control were measurably better the second time. “Preoccupations with pressing budgetary concerns leave fewer cognitive resources available to guide choice and action,” the authors state in their introduction. Search handles: cognitive bandwidth scarcity, Mani sugarcane Raven’s.
Case study
The FCC broadcast incentive auction, 2016–17. To move television spectrum to mobile carriers, the Federal Communications Commission (FCC) first had to buy it back from hundreds of local broadcasters. Most of these had no auction expertise and no budget for consultants. The obvious process, asking the local broadcasters to name their price in a sealed envelope, would require them to solve a complex frequency allocation puzzle in addition to guessing what their competitors in the bid, neighboring stations using neighboring frequencies, would ask. The FCC still wanted to secure the lowest cost combination clearing target bands and re-packing the stations that continued, so they simplified the process for the bidders. They ran a series of rounds naming a certain price, the stations didn’t need to do anything if it was still above the amount they wanted, and they only needed to declare they were out if it dropped below. The FCC as process designer lowered the burden on the other side out of self-interest: overloaded bidders bid badly or stay away while the auction needed them to participate. Search handles: deferred-acceptance clock auction, obvious strategy-proofness.
And one you would not expect
The sterile cockpit rule, 14 CFR §121.542 (FAA, 1981). Below ten thousand feet, no flight crew member is allowed to do anything unrelated to the safe operation of the aircraft: no meals, no non-essential conversation, no reading. The rule dates from 1981, after the NTSB attributed crashes to crews who were competent and alert, but distracted talking about something else. When Air Florida 90 went into the Potomac in 1982, the NTSB concluded the crew had violated sterile cockpit regulations through the preflight checklist. Load management written into law, taking out in-the-moment judgment calls. Search handle: sterile cockpit rule critical phase of flight.
Load is pattern 6.1 of twenty-seven. The two axes, the nine categories and the full set are laid out in The Negotiation Pattern Language.