No stage of the sales process carries more mythology than closing. The word conjures a library of named moves: the assumptive close, the alternative close, the puppy dog close, the Ben Franklin close. Generations of salespeople were trained to believe that deals are won or lost in the final conversation, by the seller who asks in exactly the right way. The evidence says something close to the opposite. In complex B2B sales, the deal is substantially decided before the final meeting, the biggest threat at the finish line is not a competitor but a customer who cannot decide, and the behaviors that actually move a late stage deal have far more to do with reducing the buyer's risk than with the phrasing of the ask.

This article lays out what the research actually supports about closing, why so much traditional closing advice fails in complex sales, and a working model (the Last Mile Check) that a seller or a sales AI can run on any late stage deal.

The most important study on closing techniques remains Neil Rackham's Huthwaite research program, published in SPIN Selling: an observational study of about 35,000 sales calls, conducted over 12 years across 23 countries. Rackham went looking for the correlation everyone assumed existed, between the use of closing techniques and sales success. He found it, but only in small, low value transactions. In larger sales the correlation reversed. The most successful calls in major sales contained the fewest closing techniques, and heavy closers performed worse as deal size grew.

The explanation Rackham offered still holds up. Closing techniques are pressure devices. Pressure works, modestly, when the decision is small, the risk is low, and the buyer can comply just to end the conversation. In a major purchase, pressure raises the perceived stakes of a decision that is already stressful, and sophisticated buyers read the technique for what it is: manipulation. The bigger the decision, the more a hard close signals that the seller's interests and the buyer's interests have diverged.

Rackham's second contribution matters just as much for practice. He distinguished a genuine advance (a specific commitment that moves the sale forward: a signed pilot agreement, a meeting booked with the economic buyer, legal review started) from a continuation (a pleasant ending with no commitment: "send me the proposal and we will take a look"). Sellers who could not tell these apart consistently overestimated their pipelines. Closing, properly understood, is the discipline of converting continuations into advances all the way through the deal, not a speech act at the end of it.

If closing techniques are not the answer, what is the actual problem at the end of a deal? The largest modern dataset on this question comes from Matthew Dixon and Ted McKenna's research for The JOLT Effect (2022), a machine analysis of more than 2.5 million recorded sales conversations. Two findings reframe the closing stage entirely.

First, between 40% and 60% of qualified deals in their data ended in no decision: the buyer expressed intent to purchase, often verbally committed, and then simply failed to act. Not lost to a competitor. Lost to inaction. For most B2B sellers, "no decision" is the largest competitor they face, and it is invisible on most loss reports.

Second, when Dixon and McKenna decomposed those no decision losses, only 44% traced back to a genuine preference for the status quo, the buyer concluding the current state was good enough. The other 56% were caused by customer indecision: buyers who agreed the problem was real and the solution was better, and still could not bring themselves to sign. In their data, 87% of all opportunities showed moderate or high levels of indecision somewhere in the buying group.

That split matters because the two failure modes require opposite treatments, and most sales training only equips sellers for one of them.

Status quo preference is a business case problem. The buyer does not yet believe the cost of staying put exceeds the cost of changing. The standard playbook (quantify the pain, build the ROI model, create urgency, sell the cost of inaction) is the correct medicine here, and earlier articles in this series (see the business case article, number 12) cover it in depth.

Indecision is a different disease. Dixon and McKenna found it is driven not by doubt about the company's outcome but by personal fear: the fear of messing up. The buyer worries they will pick the wrong option, that they have not read enough, that the promised benefits will not materialize and the failure will have their name on it. Once a buyer has established purchase intent, their psychology shifts. They stop optimizing for success and start guarding against blame.

Here is the trap: the standard playbook makes indecision worse. Dialing up urgency and fear of inaction on an indecisive buyer adds pressure to exactly the anxiety that is paralyzing them. In the JOLT dataset, high performers behaved differently once intent was established. The authors summarize the pattern as JOLT: judge the level and source of indecision, offer a recommendation instead of an open menu, limit the exploration of endless additional information, and take risk off the table with safety nets. Sellers who pumped the fear of missing out at this stage closed fewer of these deals, not more.

The practical implication is that a seller at the closing stage must first diagnose which disease they are treating. Urgency for status quo preference. Safety for indecision. Applying either medicine to the wrong disease is worse than doing nothing.

The other structural fact about closing in modern B2B is that the final decision is made in rooms the seller never enters. Gartner's buying journey research puts the typical buying group for a complex B2B solution at six to 10 decision makers, and finds buyers spend only about 17% of the total purchase process time meeting with potential suppliers, a figure that drops to roughly 5% to 6% per supplier when several are being compared. The closing conversation you plan so carefully is a small slice of a process dominated by internal meetings, independent research, and consensus building you do not witness.

Conversation intelligence data points the same direction. Gong Labs' analysis of 10,332 sales opportunities found deal value strongly correlated with the number of buyer side participants involved in won deals, and found no downside to win rates from adding buyer participants to meetings. Deals that close have wide, multithreaded contact across the buying group; deals that stall are typically anchored to one enthusiastic contact who cannot carry the internal argument alone.

So a real closing strategy is mostly an arming strategy. The question is not "how do I ask for the order" but "what will my champion say when the CFO raises the risk question, and have I given them the material to say it well." Multithreading, champion enablement, and mutual action plans are closing behaviors, executed weeks before the close.

Pulling the evidence together, here is a compact framework for any late stage deal. Run four checks, in order. The deal is ready to close when all four pass, and the failing check tells you exactly what work remains.

1. Commitment check. Is the most recent buyer commitment an advance or a continuation (Rackham's distinction)? List the specific, dated commitments the buyer has made: access granted, stakeholders introduced, legal engaged, budget confirmed. If the last three interactions produced only warm words, you do not have a closing problem, you have a qualification problem, and asking harder will not fix it.

2. Coverage check. Map the buying group against Gartner's six to 10 expectation. Who signs, who pays, who implements, who can veto? For each: have they engaged, and does your champion have materials that survive the internal meeting without you? A deal single threaded through one contact is not in the closing stage, whatever the forecast says.

3. Diagnosis check. If intent is established but the signature is not moving, classify the resistance. Status quo preference sounds like "we are not sure this is a priority." Indecision sounds like "we just need to compare a couple more options," "can you send more case studies," or silence after a verbal yes. The first calls for a sharper cost of inaction case. The second calls for the opposite: less pressure, more safety.

4. Safety check. For indecisive buyers, engineer the risk out of saying yes. Make a clear, personal recommendation for one option instead of presenting three. Close down open ended evaluation by agreeing on what information would actually change the decision, and a date on which the evaluating stops. Offer downside protection sized to the fear: a phased rollout, a paid pilot with defined success criteria, an opt out clause, professional services attached to the riskiest part of the implementation. The goal is that the individual buyer can defend this decision if it is ever questioned.

Ask directly, and drop the choreography. The evidence against closing techniques is not evidence against asking. A clear, direct request ("are you ready to move forward, and if not, what specifically remains") is honest information gathering. What the data indicts is scripted manipulation, which suppresses exactly the risk conversation an indecisive buyer needs to have.

Replace trial closes with a mutual action plan. A dated, shared document listing every step from today to go live (security review, legal, procurement, signature, kickoff) converts vague momentum into a series of small advances, surfaces hidden approvers early, and gives the buying group a low anxiety path to a large decision.

Protect the price at quarter end. Buyers have learned that seller deadlines produce discounts, and a seller who suddenly applies calendar pressure teaches the buyer to wait. If you offer a concession for a signature date, trade it explicitly for something (a reference, a case study, a multi year term), consistent with the trading rules in the negotiation article (number 13).

Treat silence after a verbal yes as fear, not rudeness. The JOLT data suggests the most likely explanation for a buyer who committed and then went quiet is not that they stopped caring, but that acting feels riskier than stalling. The productive response is a safety move, not an urgency move: a recommendation, a smaller first step, or a direct conversation about what would make this decision feel safe.

Closing, on the evidence, is not a moment of persuasion. It is the last mile of a risk removal process that started at discovery. Sellers who understand that close more, ask less awkwardly, and lose far fewer deals to the most expensive competitor of all: no decision.