Section 1
The five challenges at a glance
Five mid-funnel failure patterns account for most demo-to-close leakage in service firms. They are ordered roughly as deals experience them: momentum decays first, then structural under-coverage of the buying group shows up, then committee dynamics and indecision finish the job - all hidden by weak instrumentation. The table summarizes root causes, victims, and evidence; the sections that follow analyze each cluster against the underlying research.
Section 2
Where mid-funnel leakage starts: speed and momentum decay
The most rigorous data on sales momentum remains Oldroyd, McElheran and Elkington's study of 1.25 million leads across 42 companies (Harvard Business Review, 2011): firms that attempted contact within an hour of an inquiry were nearly seven times as likely to have a meaningful conversation with a decision maker as those who waited even an hour longer, and more than sixty times as likely as those waiting a day. The study measured first response, but the decay mechanism - attention and intent perish quickly - applies across the funnel. A demo that ends without a scheduled next step, a proposal delivered and followed up five days later, a stakeholder question answered after the committee already met: each is the same leak. For founder-led firms the root cause is structural, not motivational. Salesforce's State of Sales research (2023) found sellers spend under 30% of their time actually selling, and a founder splitting attention between delivery and sales is usually well below that. The practical implications are mechanical: never end a demo without the next meeting booked; send the recap with decision-relevant materials the same day; treat any deal silent for more than five business days as formally at risk with a prescribed re-engagement play. Velocity in the mid-funnel is mostly the absence of self-inflicted pauses, and pauses are a process problem with a process fix.
Section 3
The committee problem: single-threading and unmanaged conflict
Gartner's buying-journey research (2019) established the structural facts: typical B2B purchases involve 6-10 decision makers, buyers spend only about 17% of the journey meeting suppliers, and when comparing options they give any single seller perhaps 5-6% of their time. A founder running a great demo for one champion has therefore reached a fraction of the people who will decide - and the more recent finding makes it worse: a 2025 Gartner survey found 74% of buying teams demonstrate unhealthy conflict during the decision process, while groups that reach consensus are 2.5 times more likely to call their purchase high-quality. The deal a small firm loses in week six was often structurally lost in week two, when nobody mapped who else would weigh in. Vendor research points the same direction, with the usual caveat: Gong's analysis of deals on its own platform (vendor research, not peer-reviewed) associates multithreaded deals - multiple engaged stakeholders on both sides - with substantially higher win rates than single-threaded ones. The composite playbook: in or immediately after the demo, ask directly who else will be involved in evaluating, approving, and living with the decision; create stakeholder-specific artifacts (a one-page economic case for finance, an implementation outline for operations); and equip the champion to sell internally, since most of the persuasion happens in the 83% of the journey where the seller is not in the room (Gartner, 2019).
Section 4
Indecision, not competition, kills the late funnel
The most important mid-funnel research of the past decade is Dixon and McKenna's analysis of more than 2.5 million recorded sales conversations (The JOLT Effect; Harvard Business Review, 2022): 40-60% of qualified deals end in no decision - and critically, the majority of those buyers had already concluded the status quo was inadequate. They did not stall because the seller failed to prove value; they stalled from indecision - fear of choosing wrong, of overpaying, of botching implementation. That distinction inverts the standard rescue play. When sellers respond to stalling by re-pitching ROI and dialing up fear of missing out, the research found it often makes things worse, because the buyer's problem is personal risk, not business value. The effective behaviors Dixon and McKenna identified map to a JOLT pattern: judge the level of indecision honestly (and qualify it like budget), offer a recommendation rather than an open menu, limit the exploration phase by taking options off the table, and take risk off the buyer with phased starts, pilots, and explicit safety nets. For service firms this is a natural advantage: engagements can be sliced into a diagnostic or pilot phase with bounded cost, success criteria, and an exit - converting one large frightening decision into a small reversible one. Firms should also record no-decision as its own outcome in the CRM; mixing it with competitive losses hides the funnel's largest leak (Dixon & McKenna, 2022).
Section 5
Innovative solutions
Several practices now visible among high-performing small firms attack the gap directly. Mutual action plans: a shared one-page timeline from demo to go-live, co-owned with the champion, listing every step, owner, and date - making stalls visible the moment a date slips and giving the committee a consensus object to align around, a direct response to Gartner's buying-group conflict findings (Gartner, 2025). Same-day decision kits: instead of a recap email, sellers ship a short asset pack - economic one-pager, implementation outline, reference contacts - built for forwarding, because most internal selling happens without the seller present (Gartner, 2019). Indecision scoring: deal reviews add an explicit indecision grade alongside fit and intent, with playbooks keyed to it - recommendation-led proposals for option-overwhelmed buyers, pilots and guarantees for risk-frozen ones (Dixon & McKenna, 2022). AI-assisted velocity instrumentation: even two-person teams can now auto-log meetings, flag deals aging past stage norms, and draft same-hour follow-ups, directly attacking the response decay documented by Oldroyd et al. (2011) without adding admin burden - which matters because Salesforce (2023) found admin already crowds selling below 30% of seller time. And proposal-as-decision-document design: replacing twelve-page proposals with a one-page recommended option plus appendix, on the evidence that more options deepen indecision rather than resolve it.
Section 6
Solution framework
The framework treats the mid-funnel as four sequential disciplines. Discipline one - momentum SLAs: same-day recaps, next meeting always booked before the current one ends, 24-hour answers to stakeholder questions, five-day silence triggers a prescribed re-engagement play (Oldroyd et al., 2011). Discipline two - coverage: every deal past demo must have a stakeholder map naming the 6-10 likely participants, with at least three engaged contacts before proposal; single-threaded deals are flagged as structurally at risk regardless of how warm the champion is (Gartner, 2019). Discipline three - consensus engineering: a mutual action plan on every qualified deal, stakeholder-specific artifacts, and explicit facilitation of the internal debate - offering to join the committee meeting, pre-answering the conflict points Gartner (2025) shows derail 74% of buying teams. Discipline four - de-risking: classify stalled deals by indecision type and respond with recommendation, option-narrowing, or risk transfer (pilot, phased scope, exit clause) rather than value re-pitching (Dixon & McKenna, 2022). Underneath all four sits instrumentation: clear stage definitions, stage-aging visibility, and no-decision tracked as its own loss reason. A founder who can see where deals sit, how long they have sat, and why they died has converted the fog between demo and close into a managed system - which is the actual difference between founder-led selling and repeatable GTM.
Section 7
Evidence-based action plan
Week 1: instrument the gap. Define stages from demo to close in plain language, backfill the last 20 stalled deals, and label each loss honestly - competitor, disqualified, or no decision. Expect the no-decision share to land in the 40-60% range the conversation research found (Dixon & McKenna, 2022); that number is your agenda. Week 2: install momentum SLAs - same-day recap, next-step-before-hangup, five-day silence alarms - and automate what you can, given that selling time is already scarce (Salesforce, 2023; Oldroyd et al., 2011). Weeks 3-4: retrofit stakeholder maps onto every live deal; for any single-threaded opportunity, use the champion to open two more relationships this month (Gartner, 2019). Month 2: build the standard kit - mutual action plan template, economic one-pager, implementation outline, pilot offer with bounded cost and exit. Rewrite your proposal as a recommendation document with one primary option. Month 3: run stalled-deal reviews keyed to indecision type and apply the matching play - recommend, narrow, or de-risk. Quarterly: track three numbers - median days demo-to-close, percentage of deals with 3+ engaged stakeholders, and no-decision rate. If velocity improves while no-decision falls, the gap is closing; if only velocity improves, you are likely just disqualifying earlier, which is also a win. For adjacent evidence in this pillar, see [Sales Compensation for Small Firms: What Incentive Research Says About Comp Design, Quotas, and Failure Modes](/blog/growth-sales-compensation-small-firms) and [CRM Adoption That Sticks: The Honest Evidence on Failure Rates and the Minimal-Viable-CRM Approach](/blog/growth-minimal-viable-crm-adoption).