Sales Funnel：卖方的衡量工具，而非买方的决策路径
A sales funnel is the seller's count of how many prospects reach each stage of a sales process, from first contact to a closed deal. It measures the seller's own process, so it shows where prospects stop moving forward, not how buyers actually decide.
That distinction sounds small. It isn't. Most funnel mistakes start when a team reads its stage report as a map of the buyer's thinking, then spends a quarter fixing the wrong step.
Sales Funnel, Marketing Funnel, and Conversion Funnel
Three funnel terms overlap, and they're easy to blur. The marketing funnel describes the wider path from first brand discovery through to loyalty. The conversion funnel focuses on the steps a visitor takes toward one specific action, usually on a website or in an app.
A sales funnel is narrower in one way and stricter in another. It starts where a seller can count a named prospect and ends at a won or lost deal, plus any expansion revenue the seller tracks afterward. Its job is accounting. Which prospects moved, which stalled, and how long each step took.
The Stages a Sales Funnel Counts
There's no single official stage list. Teams name stages differently, and that's fine, as long as every stage has an entry rule nobody can argue about.
From lead to qualified lead
The funnel opens when a contact becomes a lead the seller can track, usually as a result of lead generation work. The next gate is qualification. Does this prospect fit, and is there a real need behind the inquiry? A lead that passes becomes a qualified lead, and many teams split that gate into a marketing-qualified step and a sales qualified lead step. A rising lead count with a flat qualified count tells a very different story from the reverse.
From opportunity to closed deal
Once sales accepts a qualified lead and starts an active conversation, most teams call it an opportunity. From there it closes as won, closes as lost, or just sits. That third outcome gets far less attention than it deserves. A report that shows only wins and losses hides the deals that are quietly aging, and those are often where the real leak is.
In e-commerce the same logic compresses into a few clicks: product view, cart, checkout, order. The stages look different. The counting job is identical.
After the first sale
Some funnels stop at the first order. That's a choice, not a rule. A peer-reviewed study of sales funnel structures on B2B digital platforms, published in Industrial Marketing Management, reports that in its model a 1% increase in up-selling efforts leads to a 0.70% increase in sales conversions. The abstract doesn't say where the data came from, so treat it as a finding from one study setting. The lesson for funnel design: activity after the sale can feed back into conversion, so a funnel that ends at the first invoice may miss part of what drives it.
None of these four gaps shows up as a stage on a funnel report, but each one changes how stage counts should be read.
| Gap | What the evidence says | What it means for the funnel | Source (label) |
|---|---|---|---|
| Buyers research before they reach you | 67% of 646 B2B buyers prefer a buying experience without a sales rep. In a separate release from the same survey period (645 buyers), 70% prefer a completely digital, self-service experience, and buyers used an average of seven information sources. These answer different questions and should not be merged. | Much of the customer journey happens before the first trackable stage, so the funnel records a late arrival as the beginning. | Gartner, analyst surveys fielded August through September 2025; respondent geography not stated (REPORTED) |
| The real decision happens inside the buying group | 74% of buyer teams showed unhealthy conflict (632 B2B buyers). Buying groups that reached consensus were 2.5 times more likely to report a high-quality deal. In the 2025 survey, buyers with high decision confidence were twice as likely to report one. | A deal stuck at one stage may be a stakeholder disagreement, not a pricing problem. A stage count can't tell which. | Gartner, analyst surveys fielded August through September 2024 and 2025; respondent geography not stated (REPORTED) |
| Most buyers aren't in the market today | If business buyers replace a provider roughly once every five years, only about 20% are in the market in a given year, or about 5% in a given quarter. The resulting 95% figure is a rule of thumb, not a precise law. | A quiet quarter may reflect market timing, not a broken stage. | Dawes, J., Ehrenberg-Bass Institute for Marketing Science, May 2021; heuristic calculation (REPORTED) |
| Stages don't move independently | Research on sales funnels often looks at stages in isolation. This study's model looks at how online sentiment and sales strategies affect stages over time. In its model, a 1% increase in up-selling efforts leads to a 0.70% increase in sales conversions. | A change at one stage, such as a tighter qualification rule, can move the close rate even though nobody touched closing. | de Almeida et al., Industrial Marketing Management, Vol. 132; peer-reviewed; data geography not stated (VERIFIED, abstract) |
What a Sales Funnel Can't See
Here's where the seller's view and the buyer's reality start to split. Four gaps show up in the research, and none of them appears as a stage on a funnel report.
Buyers research before they reach you
In a Gartner sales survey of 646 B2B buyers, fielded August through September 2025, 67% said they prefer a buying experience without a sales rep. A separate Gartner release from the same survey period, based on 645 buyers, reported that 70% prefer a completely digital, self-service buying experience, and that buyers used an average of seven information sources during a recent purchase. These are analyst survey figures. Neither release states where respondents were located, and the 67% and 70% answer different questions, so they shouldn't be merged into one number.
The implication is plain. Much of the customer journey happens before the first trackable stage. By the time a prospect shows up as a lead, they may already have narrowed the field, and the funnel records that arrival as the beginning.
The real decision happens inside the buying group
B2B purchases rarely have a single decision-maker. In a Gartner survey of 632 B2B buyers conducted August through September 2024, 74% of buyer teams showed unhealthy conflict during the decision process, and buying groups that reached consensus were 2.5 times more likely to report a high-quality deal. The 2025 survey behind the 67% figure adds a related finding: buyers with high decision confidence were twice as likely to report a high-quality deal as buyers with low confidence.
None of this is visible in a stage count. A deal that has been sitting at the proposal stage for six weeks might be a pricing problem. Or it might be three stakeholders who still don't agree on what problem they're solving. The funnel can't tell you which one it is.
Most buyers aren't in the market today
John Dawes of the Ehrenberg-Bass Institute for Marketing Science, writing in May 2021, worked through a simple calculation. If business buyers replace a provider roughly once every five years, only about 20% of them are in the market in a given year, or about 5% in a given quarter. Dawes is clear that the resulting 95% figure is a rule of thumb rather than a precise law. Still, it changes how you read a thin funnel. A quiet quarter may reflect market timing, not a broken stage.
Stages don't move independently
Funnel reports tend to treat each stage as its own box. The Industrial Marketing Management study mentioned earlier notes that research on sales funnels often looks at stages in isolation, and its own model examines how online sentiment and sales strategies affect stages over time. In practice, a change at one stage can show up somewhere else. Tighten the qualification rule, for instance, and the close rate can move even though nobody touched the closing process.
These are the reasons US online shoppers gave in Baymard Institute's survey. The 42% "just browsing" share covers all respondents, the other five reasons cover the shoppers who remained, and the page does not state the survey year or sample size.
| Reason for abandoning the cart | Share of shoppers | Respondent base | Source (label) |
|---|---|---|---|
| Just browsing or not ready to buy | 42% | All surveyed shoppers; no checkout fix will change this group | Baymard Institute, survey of US online shoppers (REPORTED) |
| Extra costs were too high | 40% | Remaining shoppers, after excluding "just browsing" | Baymard Institute, survey of US online shoppers (REPORTED) |
| Delivery was too slow | 20% | Remaining shoppers, after excluding "just browsing" | Baymard Institute, survey of US online shoppers (REPORTED) |
| Did not trust the site with card details | 19% | Remaining shoppers, after excluding "just browsing" | Baymard Institute, survey of US online shoppers (REPORTED) |
| Required to create an account | 18% | Remaining shoppers, after excluding "just browsing" | Baymard Institute, survey of US online shoppers (REPORTED) |
| Checkout process too long or complicated | 17% | Remaining shoppers, after excluding "just browsing" | Baymard Institute, survey of US online shoppers (REPORTED) |
Measuring a Funnel Without a Borrowed Benchmark
Every funnel team eventually asks the same question. Is our conversion rate good?
Why no universal conversion rate holds up
Research reviewed for this entry found no government statistics office or peer-reviewed study that publishes a universal sales funnel conversion benchmark. The averages encountered during that research came from vendors and agency blogs, mixed industries and deal sizes, and rarely defined their stages the same way. A lead in one company's system is a qualified opportunity in another's.
The more useful comparison is internal. Your own funnel, this quarter against last quarter, with the same stage definitions on both sides.
Stage conversion, time in stage, and cohorts
Three measures do most of the work. Stage conversion is the share of prospects at one stage who reach the next. Time in stage is how long they take to get there. Cohort counting follows a group that entered during the same period all the way through, instead of comparing this month's new leads with this month's closed deals.
That last point catches a lot of teams out. In a long sales cycle, today's wins and today's new leads belong to different people. Cohorts fix the mismatch, at the cost of waiting longer for an answer.
The checkout stage in e-commerce
For online stores, the bottom of the funnel is the cart, and the scale of loss there is large. Baymard Institute, a commercial UX research institute, puts the average documented cart abandonment rate at 70.22%, an average across 50 studies, on a page last updated September 22, 2025.
Baymard's survey of US online shoppers adds a useful split. 42% abandoned a cart because they were just browsing or not ready to buy, which no checkout fix will change. Among the remaining shoppers, the leading reasons were extra costs that were too high (40%), delivery that was too slow (20%), not trusting the site with card details (19%), being required to create an account (18%), and a checkout process that was too long or complicated (17%). The page doesn't state the survey year or sample size.
Baymard also estimates that the average large e-commerce site could raise its conversion rate by 35.26% through better checkout design alone. That estimate rests on roughly ten years of checkout usability testing on large US and European sites, so read it as a figure for large sites, not a promise for any single store.
Automating a Stage, and When It Doesn't Pay Off
Automated lead nurturing is the obvious tool for moving prospects between early stages without a salesperson. The evidence on whether it raises conversion is mixed.
A Journal of Marketing study by Habel, Hartmann, Wiseman, Ahearne, and Vaid (Vol. 89, No. 5, September 2025) combined one qualitative study with three quantitative ones. On average, automated lead nurturing raised the probability of lead conversion by 23 percentage points in Study 2 and by zero in Studies 1 and 3. Those are separate studies in different settings. They shouldn't be averaged into one headline figure.
The authors tie the difference to context: the length of the sales cycle, the expected sales volume, and whether the lead already has a relationship with the seller. The American Marketing Association's summary of the study links the stronger result to relatively simple products with shorter sales cycles, and it recommends judging automation by lead-to-meeting conversion, sales cycle speed, and revenue rather than by email opens and clicks. The mechanism the abstract describes is information: both sides know more before they talk, which improves the conversation that follows.
So automate a stage where prospects lack information and a human touch would be expensive, then test it in your own funnel.
Common Sales Funnel Mistakes
Most of these trace back to one root cause, treating the seller's stages as the buyer's path.
Reading stage order as decision order. A prospect who books a demo may already have chosen two finalists, while the funnel records the demo as early interest.
Benchmarking against a borrowed average. Without shared stage definitions, someone else's conversion rate isn't comparable to yours.
Calling survey differences a trend. Analyst surveys can change questions and samples between waves, so don't string separate waves into a trend line.
Fixing one stage in isolation. Stages affect each other, so a change in qualification rules or post-sale activity can move numbers further along the funnel.
Ignoring stalled deals. Wins and losses get reported. Deals that sit untouched often don't, and they can hide the clearest signal you have about friction inside the buying group.
Frequently Asked Questions
What are the stages of a sales funnel?
There is no official list. A common practitioner sequence runs from lead to qualified lead, then to opportunity, then to a won or lost deal, with expansion such as up-selling sometimes tracked after the first sale. In e-commerce the stages are usually product view, cart, checkout, and order.
What is a good sales funnel conversion rate?
Research for this entry found no universal benchmark from official statistics or peer-reviewed studies. The most reliable comparison is your own funnel over time, using the same stage definitions and counting cohorts rather than snapshots.
Why do online shoppers abandon checkout?
In Baymard Institute's survey of US online shoppers, 42% were just browsing. Among the rest, the top reasons were extra costs (40%), slow delivery (20%), distrust of the site with card details (19%), forced account creation (18%), and a long or complicated checkout (17%).
Does automated lead nurturing increase conversions?
Sometimes. In a 2025 Journal of Marketing study, it raised lead conversion probability by 23 percentage points in one study and by zero in two others. The effect depended on sales cycle length, expected sales volume, and the lead's existing relationship with the seller.
Why do qualified deals stall in a B2B sales funnel?
Often the cause sits outside the funnel. A Gartner survey fielded in 2024 found unhealthy conflict in 74% of buyer teams, and only part of any market is buying at a given moment, so a stalled deal can reflect group disagreement or timing rather than a weak stage.
Is a sales funnel still useful if buyers research on their own?
Yes, as a measuring tool. A Gartner survey fielded in 2025 found that 70% of B2B buyers prefer a completely digital, self-service experience and that buyers used an average of seven information sources. The funnel can't track that research, but it shows where prospects stall once they reach you.
Related Terms
- Marketing Funnel
- Conversion Funnel
- Lead Nurturing
- Sales Qualified Lead (SQL)
- Customer Acquisition
Sources
Gartner, sales survey press release, March 9, 2026. Survey of 646 B2B buyers, August through September 2025; respondent geography not stated. Analyst survey. REPORTED. https://www.gartner.com/en/newsroom/press-releases/2026-03-09-gartner-sales-survey-finds-67-percent-of-b2b-buyers-prefer-a-rep-free-experience.
Gartner, survey press release, May 20, 2026. Survey of 645 B2B buyers, August through September 2025; respondent geography not stated. Analyst survey. REPORTED. https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-survey-finds-sixty-nine-percent-of-b-two-b-buyers-turn-to-sales-reps-to-validate-ai-generated-insights.
Gartner, sales survey press release, May 7, 2025. Survey of 632 B2B buyers, August through September 2024; respondent geography not stated. Analyst survey. REPORTED. https://www.gartner.com/en/newsroom/press-releases/2025-05-07-gartner-sales-survey-finds-74-percent-of-b2b-buyer-teams-demonstrate-unhealthy-conflict-during-the-decision-process.
Dawes, J., Ehrenberg-Bass Institute for Marketing Science, article on advertising effectiveness and the 95-5 rule for B2B buyers, May 2021. Heuristic calculation; the author states it is not a precise rule. REPORTED. https://marketingscience.info/news-and-insights/advertising-effectiveness-and-the-95-5-rule-most-b2b-buyers-are-not-in-the-market-right-now.
de Almeida, M. I. S., et al., "Sales funnel structures in B2B digital platforms: dynamic effects of online sentiment and sales strategies on marketing performance," Industrial Marketing Management, Vol. 132, pp. 135-149. DOI 10.1016/j.indmarman.2025.12.005. Peer-reviewed journal; data geography not stated in the abstract. VERIFIED (abstract). https://repositorio.bc.ufg.br/items/0d730296-0b28-4231-b441-6da81e029e16.
Baymard Institute, cart abandonment rate statistics page, last updated September 22, 2025. Commercial UX research institute; average across 50 studies; abandonment reasons from a survey of US online shoppers, survey year and sample size not stated. REPORTED. https://baymard.com/lists/cart-abandonment-rate.
Habel, J., Hartmann, N. N., Wiseman, P., Ahearne, M. J., and Vaid, S., "Sales Pipeline Technology: Automated Lead Nurturing," Journal of Marketing, Vol. 89, No. 5, pp. 66-87, September 2025. DOI 10.1177/00222429251321417. Peer-reviewed journal; abstract via McMaster Experts. VERIFIED (abstract). https://experts.mcmaster.ca/scholarly-works/3596577.
American Marketing Association, summary of the Habel et al. automated lead nurturing study, March 25, 2025. Publisher summary of a peer-reviewed study. REPORTED. https://www.ama.org/2025/03/25/does-automated-lead-nurturing-really-work-a-new-study-challenges-the-hype/.
