Every sales leader has heard some version of this post-mortem. The deal looked healthy for four months. The champion answered quickly, sounded enthusiastic, and seemed senior enough. Then she took a job elsewhere, got pulled into a reorganization, or quietly lost an internal argument that nobody on the selling side knew was happening. The opportunity did not die loudly. It went quiet, slipped a quarter, and was eventually marked closed-lost with "no decision" in the reason field.

The standard fix is "multithread": get more people from the buying organization into the deal. The data behind that advice is real, but it is routinely misread. Counting contacts is easy, and it is also one of the least informative things you can measure about a buying relationship. A deal with nine names in the CRM can be as fragile as a deal with one, and a contact added carelessly can make consensus harder, not easier.

This article looks at what the evidence actually shows about multithreading, where the popular statistics overreach, and what a thread has to carry before it protects a deal.

A working definition: multithreading is building independent, two-way relationships with several people in the buying organization, so that the deal's progress does not depend on one person's availability, memory, or influence.

The important word is independent. A person copied on a recap email is not a thread. A name pulled from a webinar list is not a thread. A thread exists when someone has engaged in a real exchange about the decision and would still take your call if your champion disappeared tomorrow.

Multithreading is also not "going around" the champion. Done badly, it reads as distrust, and it spends the champion's political capital without asking. Done well, it is something the champion wants, because it spreads the personal risk of sponsoring a purchase across more people and gives them allies when the deal meets resistance.

The most cited data comes from conversation intelligence vendors, and the largest recent dataset is from Gong Labs. Analyzing 1.8 million new business deals closed in 2024, Gong reported that 77% of deals were multi-threaded, and that the deals that closed successfully had twice as many buyer contacts as those that did not [1]. Large strategic deals won by reps included an average of 17 contacts, and Gong found buying group size "directly correlated" with the dollar amount of the opportunity. Its headline claim is that multithreading boosts win rates by an average of 130% in deals over $50K [1].

Gong also measured the selling side: selling teams on closed-won deals were 67% larger than on lost deals, with an average of 6.7 members of the sales team involved by the end of discovery and scoping [1]. That second finding matters, and we will come back to it.

This is vendor data from customers of one platform, and Gong does not publish its method in the summary. It is still the largest public dataset on the question, and the direction is consistent with other sources.

The most useful academic work on this question predates the word "multithreading." In a 2008 Journal of Marketing study, Robert Palmatier modeled customer value in business relationships using three relational drivers: relationship quality (the caliber of relational ties), contact density (the number of relational ties), and contact authority (the decision-making capability of relational contacts) [2]. Using dyadic data across 446 business-to-business exchanges, he found that the value generated from interfirm relationships comes not only from the quality of ties (trust, commitment, norms) but also from the number and decision-making capability of contacts, and from the interactions among those drivers [2].

Two moderator findings are directly relevant to sellers. Increasing contact density paid off most for sellers whose customers had high employee turnover. Building relationships with key decision makers generated the highest returns among customers that were more difficult to access [2]. In plain terms: breadth is insurance against people leaving, and authority is what you need when access is scarce. They are different assets, and you cannot substitute one for the other by adding more names.

Neeli Bendapudi and Robert Leone studied the problem from the vendor side in a 2002 Journal of Marketing paper: what happens to a business customer's loyalty when the vendor's key contact employee leaves [3]. Their qualitative studies drew on 72 focus group respondents, 47 depth interviews, and open-ended questionnaires from 83 managers across 34 companies. Customers were "quite sensitive to the loss of a key contact employee," and the authors proposed that "the more points of contact there are between the customer and the firm, the less critical any single employee relationship seems to be" [3].

The logic transfers directly to the buyer side. A relationship anchored to one person is exposed to that person's career. Bendapudi and Leone also recorded the cost of overdoing it: some customers "felt overwhelmed by having to deal with multiple people from the vendor's side," and others wanted one accountable person high enough to make key decisions [3]. More contacts is not automatically better for the people on the other end.

Deals that progress accumulate contacts. A deal that reaches security review, legal, and procurement will have more names attached to it by the time it closes, because those functions only show up in deals that are moving. A lost deal that died after two calls never had the chance to add them. So "won deals have twice as many contacts" is partly a description of how far won deals traveled.

Gong's claim that multithreading "boosts" win rates is phrased causally, but the public summary does not show an analysis that controls for stage or for when contacts were added [1]. The finding is consistent with multithreading helping. It does not prove that adding a fifth contact to a stalled deal will revive it. The more defensible reading is the one Gong itself emphasizes: thread "early and often," before the deal's progress has done the threading for you.

Other vendor numbers are weaker still. UserGems reported, from an analysis of over 5,000 opportunities across multiple B2B SaaS companies, that multithreading improved win rates "by five times, which is 480% higher" than single-contact deals [4]. Five times is a 400% increase, not 480%, and the post does not define what counts as multithreaded or report any controls. Treat figures like this as directional at best.

The strongest caution comes from Gartner. In a survey of 632 B2B buyers conducted in August and September 2024, Gartner found that 74% of B2B buyer teams demonstrated "unhealthy conflict" during the buying decision, defined as members having conflicting objectives, disagreeing on the best course of action, or being overruled by external decision-makers [5]. Buying groups that reached consensus were 2.5 times more likely to report that their deal was high-quality. Gartner described buying groups ranging from five to 16 people across as many as four functions [5].

The finding that should change how sellers multithread is about content. Tailoring for buying group relevance had a positive impact on consensus of 20%. Content focused on individual-level relevance had a 59% negative impact on buying group consensus, because it reinforces each stakeholder's own perspective and makes them less likely to embrace a shared direction [5].

Now consider the default multithreading playbook: find the CFO and pitch cost savings, find the CIO and pitch integration, find the end users and pitch ease of use. That is individual-level tailoring by design. Each thread carries a different argument, and the seller has built, with good intentions, a buying group whose members are now each holding a different reason to buy. When they meet without you, they discover they are not agreeing on the same thing.

Jeff Johnson's 2023 qualitative study in the International Journal of Research in Marketing examined how B2B salespeople handle buying center dissenters: the one member who reacts negatively while others respond well [6]. He described in-meeting mitigation and post-meeting remediation strategies, shaped by situational factors. The relevant point for multithreading is simple. More threads mean you will meet dissent earlier. That is a feature. A dissenter you meet in month two is a problem to work on; a dissenter you meet in procurement is a veto.

If contact count is a weak signal, what should a seller track instead? The research points to a short test. For each person in the buying organization, ask four questions. A thread that passes all four is load-bearing: it would hold weight if the champion's thread snapped.

1. Continuity: would this person keep the deal alive without the champion? Do they know the problem, the decision timeline, and the next step without being briefed by the champion? If the answer is no, they are a contact of your champion's, not a thread of yours. This is Palmatier's density effect in practice: breadth protects you when people move [2].

2. Authority: what can this person do to the decision? Tag each thread with a role in the decision: decides, funds, blocks, uses, or informs. Ten threads tagged "informs" are less protective than one tagged "decides." This is contact authority, and Palmatier's results suggest it matters most where access is hardest [2].

3. Alignment: is this person hearing the same core case as everyone else? Every thread should carry one shared problem statement and one shared definition of success, with role-specific evidence attached to it, not a different argument. Gartner's results on group versus individual relevance make this the difference between threads that build consensus and threads that fragment it [5].

4. Exchange: has the thread been two-way recently? In the last 30 days, has this person asked a question, corrected something, or committed to something? Opens and cc's do not count. A thread that only receives is decoration.

As a working heuristic (not a researched threshold), a deal is meaningfully multithreaded when it has at least two load-bearing threads outside the champion, and at least one of them is tagged "decides" or "funds." A deal with eight contacts and zero load-bearing threads is single-threaded with extra names.

Recruit the champion into the threading. The cleanest route to new threads is through the champion, framed as protecting them, not checking on them. "If you are out for two weeks during legal review, who should I keep informed so this does not stall on your side?" That question invites introductions without implying distrust, and it surfaces who actually matters.

Thread at discovery, not at procurement. Threads added late tend to be introductions to the functions that can block you. Threads added early let you learn how each function sees the problem while the problem is still being defined, which is when it is easiest to shape a shared case.

Write one case, then attach lenses. Put the problem, the cost of inaction, and the decision criteria in a single document the whole group can see. Give the CFO the financial view and the IT lead the integration view as attachments to that case, not as separate pitches. When stakeholders meet without you, they should be arguing about the same document.

Match threads on your side, but keep one owner. Gong's finding that won deals had larger selling teams [1] suggests peer threads help: your sales engineer with their architect, your executive with theirs. Bendapudi and Leone's caveat applies in reverse [3]: buyers can feel overwhelmed by too many vendor contacts, and many want one accountable person. Add peers, but make sure the buyer always knows who owns the deal.

Plan for the dissenter. When a new thread pushes back, resist routing around them. Find out what they see that the champion does not. Johnson's work suggests the response happens in two places, in the meeting and after it [6], so plan both.

Change the question in pipeline reviews. Replace "how many contacts are on this deal?" with "how many load-bearing threads does this deal have, and which decision role is missing?" The second question exposes risk the first one hides.

Multithreading has costs. Every thread takes time to build and maintain, and in a small deal decided by one or two people, forcing six threads adds friction for the buyer and delay for you. Gong's strongest win-rate finding applies to deals above $50K [1], which is a reasonable signal that the payoff scales with deal size and buying group complexity.

Some organizations treat contact outside the designated sponsor as a breach of etiquette, and some champions guard access for legitimate reasons. In those accounts, the champion-led route is not just preferable; it is the only one that will not cost you the deal.

Finally, the evidence base is uneven. The largest datasets are vendor analyses with limited disclosure of method. The peer-reviewed work is older and either cross-sectional (Palmatier) or qualitative (Bendapudi and Leone, Johnson). Gartner's survey is recent and specific, but it measures buyer-reported deal quality, not win rates. The findings point in the same direction, which is encouraging, but none of them proves that a given thread will save a given deal.

The practical conclusion is narrower than the slogan. Multithreading works when each thread is independent of the champion, connected to real decision authority, carrying the same case as every other thread, and active in both directions. More names in the CRM is the byproduct, not the goal.