Skip to content Last reviewed: August 2026 | Next review: February 2027
Every pipeline conversation I have joined in the last three years has followed the same shape. The number is at risk. Coverage is reported. Coverage looks acceptable. The quarter still misses.
Then somebody says generate more pipeline, and the machine starts up. More outbound, more campaigns, a partner push, an event brought forward.
I have had this conversation with chief revenue officers, chief operating officers, heads of revenue operations and heads of marketing, in businesses from a few hundred people to global technology firms. What has struck me over twenty years is how little the geography matters. The same conversation happens in London, in Frankfurt, in Boston, in Singapore. Different market maturity, different comp structures, different cultural norms around how directly people will tell you the forecast is wrong. Same underlying problem, and it has sharpened every year I have been doing this.
The uncomfortable part is that the coverage number was never acceptable in the first place. It was measured against a rule of thumb that stopped being true a decade ago.
“Stage names are the label. The sales process is everything underneath.”
Single-year benchmark swings are mostly noise. The long series is not.
The Bridge Group has run the same biennial study of the Account Executive role since 2006, now in its tenth edition, most recently across 158 B2B companies. Same firm, same method, ten rounds. That makes it the most reliable series available on this question, and it is worth reading four of its lines together.
| Metric | Trend | Source |
| Share of AEs at quota | 74% (2012) → 66% (2022) → 51% (2024) → 48% (2026) | The Bridge Group, 2026 |
| Time to full productivity | 5.7 months (2023) → 6.2 months (2026), highest in study history | The Bridge Group, 2026 |
| Experience required at hire | 2.7 years (2022) → 3.7 years (2026) | The Bridge Group, 2026 |
| Quota to OTE ratio | 4.2x (2024) → 4.6x (2026) | The Bridge Group, 2026 |
Quotas rising against earnings. Ramp lengthening. Companies buying in more experience to compensate for it. And a smaller share of people clearing the bar each time the study runs.
None of that is a coaching problem.
Pipeline coverage is not a benchmark, it is a calculation. Required coverage equals one divided by your win rate. Three times coverage assumes you win a third of what you qualify, which is where the rule came from, and almost nobody in enterprise technology operates there now.
Win rate benchmarks currently sit well below that, though they move around depending on how the denominator is defined. Winning by Design flagged enterprise win rates falling into the 17 to 20 percent range back in 2023. Ebsta and Pavilion reported an average of 19 percent in their 2025 GTM Benchmarks, alongside 78 percent of sellers missing quota.
One caution on that source, and it matters more than the figure. Ebsta and Pavilion reported 29 percent the year before. A ten point single-year swing in a market-wide win rate is almost certainly a change in what was counted rather than a change in the market. Use the direction. Do not put a ten point market swing in front of a board.
Take a team carrying a £10m quarterly new ACV quota at a 19 percent win rate on qualified opportunities.
That team is £17.6m short and reporting green. It will find the gap in week ten of the quarter, roughly nine weeks after anything could have been done about it.
Unqualified pipeline is not neutral. It eats the scarcest thing you have, which is qualified selling hours.
A ramped AE can only run so many real opportunities at once. In the enterprise technology teams I have worked with, that ceiling sits somewhere between eight and fifteen depending on cycle length and complexity. Every slot filled by something that was never real is a slot unavailable to something that was.
Then the sequence takes over. Coverage looks thin, the team is told to generate more, standards loosen to make the number, quality falls, win rate falls, required coverage rises again.
I will be honest that I cannot cleanly separate how much of the market-wide win rate compression is genuine buyer caution and how much of it is organisations doing this to themselves. The published data does not distinguish the two, and I have not found a way to isolate it inside a single business either. What I am confident of is that the second part exists and that it is the part you can act on.
Plays are the visible part, so start there. A play is not a deck, and a library of plays is not a system.
Play taxonomy. Two types, and confusing them is why libraries bloat.
Anchor plays are tied to your ICP and core value proposition, with a twelve month life. In the businesses I have seen sustain this well it is three to five, no more. They carry most of the sourced pipeline and justify real investment in content, training and coaching
Tactical plays cover competitive displacement, product launch, event follow-through, install base expansion. Time-boxed at six to twelve weeks with an expiry date set at launch. Tactical plays that never get retired are how you end up with the library nobody opens
Three layers, built in that order.
Umbrella message, then ICP, then buying persona. Content is derived from this, not written alongside it. FocusVision research cited by Highspot puts buyers at roughly thirteen content pieces consumed before first contact with a seller. If your economic buyer and your technical evaluator get the same narrative across those thirteen touches, you have one layer rather than three
The internal kit. Entry trigger, target account list, qualification bar, discovery questions, objection handling, defined next step. This is what sellers actually consume and it is usually the thinnest part of the build
Every play names the route that carries it: direct, marketing-sourced, SDR-sourced or partner. Plays without a named route default to the AE, which is the most expensive channel you own
Each play carries a named owner, an entry trigger, an activity standard, an exit definition and a sourced pipeline target. Without those you cannot tell a play that failed from a play that was never really run, which means every post-mortem turns into an argument about effort.
Every play carries a source identifier from first touch through to closed. If you cannot attribute sourced pipeline to a specific play then you cannot evidence which ones work, and if you cannot do that you will never retire the ones that do not. Sales operations owns this. It is not an analytics nicety, it is the mechanism that lets the system improve instead of just accumulate.
Enablement runs across all of it. Play development, launch training, coaching to the execution standard, and deal reviews conducted against play criteria rather than a rep’s narrative. A play launched without a coaching plan is effectively dead within a quarter, however well it landed on the day.
Sales, marketing and partner should not be holding separate pipeline plans. One plan, one sourced pipeline target, split by route with each route carrying its own number and its own owner. Where a partner network exists it is a route with a target, not a goodwill relationship.
One more thing on cadence, because it is the most common structural error I see. The demand generation cadence is not the forecast call. The forecast call looks backwards at in-quarter deals. The demand generation cadence looks forwards at creation: sourced pipeline against target, by play, by route, with marketing and partner in the room. Run them as one meeting and creation loses to closing every single week, because closing is louder and closer. Alongside both, a win room for named must-win pursuits. Standing, cross-functional, convened by deal significance rather than by exception.
None of this assembles itself.
Discovery gets treated as a stage to pass through. It is the last point in the cycle where you can still change the outcome cheaply. By proposal the deal has been shaped and the requirements written, and often not by you.
Start with what is actually killing your deals. Your most common competitor is not another vendor. Matthew Dixon and Ted McKenna analysed more than 2.5 million recorded sales conversations for The JOLT Effect and found that 40 to 60 percent of deals are lost to no decision. The more useful finding is the split inside that: 56 percent of those losses came from buyer indecision, a fear of getting it wrong, and 44 percent from genuine preference for the status quo. Those need opposite responses. Status quo losses need a harder case for change. Indecision losses need risk taken off the table, and the research found that piling on urgency makes indecisive buyers worse most of the time.
Then the committee. The published figures differ because they count different things: Gartner puts complex purchases at six to ten decision makers, 6sense’s 2025 Buyer Experience Report puts the average buying group at 10.1 across 4,510 respondents, and Forrester’s State of Business Buying 2026 counts 22 people in a typical decision, being 13 internal stakeholders and 9 external influencers. Take the range rather than any one number. They all say the same thing about single-threaded deals.
And then the part that most qualification exercises skip. A rep saying budget exists is not budget verification. A champion saying they can get it signed is not access to the economic buyer.
MEDDIC and its extensions MEDDICC and MEDDPICC came out of PTC in the 1990s and remain the default in enterprise technology. Around them sit BANT and its descendants CHAMP and ANUM, HubSpot’s GPCTBA/C&I, RAIN Group’s FAINT, and SPICED from Winning by Design. They are all serviceable. Organisations spend months choosing between them and get nothing back, and the reason is almost never the framework. It is that the framework ends up as CRM fields rather than stage exit criteria, and nobody inspects it.
The distinction worth holding onto is between seller activity criteria and buyer verification criteria. Demo delivered, proposal sent, meeting held: entirely within your control and therefore worthless as a predictor. Economic buyer has confirmed the business case, evaluation criteria documented and shared, procurement path and timeline known: that is a different kind of evidence. A deal should not advance a stage on seller activity alone. Something has to have changed on the buyer’s side, and it has to be in the record.
Worth saying plainly, because it is where a lot of budget goes to die.
A qualification framework is a test applied at a gate. A sales methodology is an operating system for running an opportunity: Challenger, SPIN, Sandler, Solution Selling, Value Selling, Miller Heiman Strategic Selling, Command of the Message®, Gap Selling, Target Account Selling.
Most of what a methodology gives you has nothing to grip on until an opportunity exists. Stakeholder mapping, competitive positioning, value quantification, mutual action plans, negotiation structure, all of it presumes an active deal with scope, a committee and a timeline. So when a business answers a pipeline creation shortfall by rolling out a methodology, it is answering a demand problem with a conversion instrument. The training will be well received. The pipeline gap will not move.
The boundary is not absolute. The diagnostic half of discovery genuinely is reusable, and SPIN and Gap Selling are built for exactly that moment. Some methodologies also carry a message-generation component that feeds play design, and Challenger’s commercial teaching insight is the obvious example, along with the value framework in Force Management’s Command of the Message®. Both are designed to disrupt a buyer’s status quo, which is demand creation work. Those are legitimate inputs into the customer-facing layer of a play. Inputs, not a substitute for the system.
Outside that, methodology earns its return after qualification, inside an active sales motion. Judge it on win rate, cycle length and average deal value. Do not judge it on pipeline created, and do not buy it to fix that.
“Our win rate is low because we are outgunned on product and price, not because of qualification.” Sometimes true and worth testing first. The test is your no-decision rate, and JOLT gives you something to test against. If 40 to 60 percent of your losses are to no decision then you are in the normal range and the problem is qualification and business case, which is entirely yours. If most losses are to named competitors, that is positioning and product. Most organisations do not separate the two in loss reporting, which is why the argument never resolves.
“We are a growth business. We need volume now. Qualification discipline is premature optimisation.” The strongest objection on this list and there are markets where it holds. In genuinely high-velocity motions with short cycles and high win rates, coverage really is the binding constraint. The distinction is cycle length. Past roughly ninety days, with committees past a handful of people, the arithmetic inverts, because a wasted opportunity slot now costs you a quarter rather than a week.
“We have MEDDPICC. It is in the CRM. This is an adoption problem.” Half right. Adoption is a management issue. But a framework captured as optional CRM fields has been designed as documentation rather than as a gate, and management pressure does not convert one into the other. If a deal can reach a late stage with those fields blank, the system permits it.
“We hit our sourced pipeline number. The conversion problem is downstream of marketing.” Usually accurate and usually beside the point. If marketing is measured on sourced volume and sales on closed revenue, the two are optimising against different definitions of a good opportunity, and both can hit target while the business misses. The fix is not better collaboration. It is one sourced pipeline definition, agreed qualification criteria at the handover, and one plan with one number split by route.
Every function can be individually correct and the system can still fail. That is what makes this a design problem rather than a performance problem.
This is RISE: infrastructure and execution. Process, data, execution rhythm and technology, in that order, because that is the dependency chain.
RISE is the layer everything else runs on. PACE cannot work without it, because a play with no traceability cannot be evaluated and an academy with no adoption data cannot be improved. PERFORM cannot be planned without it, because capacity models, coverage design and quota setting all depend on data that means the same thing everywhere it is measured.
It is also the least visible of the three, which is why it is chronically underfunded. Nobody puts a data model on a board slide. The consequences of not having one show up on every board slide, attributed to something else.
The Diagnostic starts here more often than clients expect. A pipeline problem or a forecast problem is frequently a process problem that has been visible for years in a form nobody recognised.
If your systems are working and your numbers still do not reconcile, that is the conversation to have. Get in touch.
Browse more perspectives from our team of operators and advisors.
The best first step is a simple conversation about where your sales organisation stands, and what to fix first. No pressure, no pitch.
A practitioner-led sales transformation practice anchored on the COMPOSE Sales Model®.
Copyright © Compose Sales Partners Limited. COMPOSE Sales Partners, the Compose Sales Partners logos, COMPOSE Sales Model is a trademark and service mark of COMPOSE Sales Partners Limited. PERFORM, PACE, and RISE are proprietary frameworks of COMPOSE Sales Partners. Other brands and product names are the trademarks of their respective companies.
Optimized by Seraphinite Accelerator
