The method_
How we got 20–35%.
20–35% of your achievable revenue, up to 1 in 3 dollars, is already sitting in your pipeline. Capturing that revenue is a 25–54% top-line lift, without buying a single new lead. An OAM is built to recover that revenue.
Parlay analysis of research from Challenger (The JOLT Effect), Belkins, Conversica, Velocify and Invoca.
In short: the leak stats are third-party research, cited one by one below. The 20–35% range and the 25–54% lift are Parlay's arithmetic on that research, shown step by step.
[01]The leaks_
Four leaks, each with its source.
Sources checked against the original publishers on November 17, 2026. Where a publisher's report is no longer online, we say so.
Leak 01
No decision
The deal is qualified and the buyer wants it, then nothing happens. It never shows up as a loss to a competitor, so it rarely gets a post-mortem. This is the largest of the four leaks.
40–60% of qualified deals end in no decision, not a loss to a competitor.
Matthew Dixon and Ted McKenna, The JOLT Effect (2022); “Stop Losing Sales to Customer Indecision,” Harvard Business Review, June 2022 · 2022 · 2.5M+ recorded sales calls
56% of no-decision losses come from buyers who wanted to move forward but couldn't commit. Only 44% chose the status quo.
Matthew Dixon and Ted McKenna, The JOLT Effect (2022); “Stop Losing Sales to Customer Indecision,” Harvard Business Review, June 2022 · 2022 · 2.5M+ recorded sales calls
Leak 02
Slow response
A buyer who raised a hand and heard nothing back moves on. The opportunity never becomes the deal it could have been.
About 1 in 4 mid-market and enterprise companies never responded to an inbound lead, up from 5% in 2020.
Conversica, 2023 Sales Effectiveness Report (secret-shopper audit) · 2023 · 100 mid-market and enterprise companies · Linked summary; Conversica's original report is no longer online.
In 2012, the average company took nearly 47 hours to answer a web lead.
InsideSales.com, ResponseAudit research, AA-ISP 2012 · 2012 · Secret-shopper web-form tests
Leak 03
Early quit
Most replies arrive after the first touch, but the average lead gets fewer than two calls. The deal is dropped while the buyer is still deciding.
58.6% of replies come at follow-up steps 2–6.
Belkins, “Sales follow-up statistics in B2B: Belkins’ 2026 study” · 2026 · 7.5M emails
Sales teams averaged 1.45 call attempts and 1.56 emails per lead.
Velocify secret-shopper study (2014), as reported by Cirrus Insight · 2014 · Linked summary; Velocify's original report is no longer online.
Leak 04
Missed calls
A buyer's call that never reaches a person is a conversation that never happened. Most callers don't leave a message, so nobody knows to call back.
Only 56% of calls to businesses reach a person.
Invoca, Lead Conversion Benchmarks Report 2026 · 2026 · 70M+ calls across 10 industries
About 7 in 10 callers who reach voicemail hang up without leaving a message.
Moneypenny caller research (2016) · 2016 · Cited from Moneypenny's published caller research.
[02]The range_
How the leaks add up to 20–35%.
- 1
Start with the deals that stall
40–60% of qualified deals end in no decision
Out of every 10 qualified deals, 4 to 6 end in no decision rather than a loss to a competitor (Challenger, The JOLT Effect, from 2.5M+ recorded sales calls).
- 2
Keep only the buyers who wanted to buy
40% × 56% = 22.4% · 60% × 56% = 33.6%
56% of no-decision losses are buyers who wanted to move forward but couldn't commit. The other 44% chose the status quo, and we leave them out. That puts 22–34% of qualified pipeline with buyers who wanted to buy and didn't.
- 3
Round to a range: 20–35%
22.4–33.6% → 20–35%
We treat that share as the revenue still in reach, and round it out to 20–35% of achievable revenue: up to 1 in 3 dollars. Two assumptions turn deals into dollars: stalled deals are about the size of the deals you win, and the pipeline those buyers sit in is revenue you could have won.
- 4
Why the other three leaks matter
Slow response (1 in 4 companies never answered an inbound lead), early quit (58.6% of replies come at follow-ups 2–6, yet teams averaged 1.45 calls) and missed calls (only 56% of calls reach a person) are how willing buyers stall and slip away. They are why the deals in step 2 go quiet, so we don't add them on top: counting them again would double count the same deals.
[03]The lift_
Why 20–35% untapped is a 25–54% lift.
If a share u of achievable revenue goes uncaptured, today's revenue is (1 − u) of what's achievable. Capturing all of the untapped share multiplies today's revenue by 1 ÷ (1 − u), so the lift on today's revenue is u ÷ (1 − u).
| Untapped share u | u ÷ (1 − u) | Lift |
|---|---|---|
| 20% | 0.20 ÷ 0.80 = 0.250 | 25% |
| 35% | 0.35 ÷ 0.65 = 0.538 | 54% |
The lift is on today's revenue, and it assumes the whole untapped share is captured. Capturing part of the untapped share gives a proportionally smaller lift.
[04]The calculator_
What the calculator does with your numbers.
- 1. Today = pipeline × win rate
- 2. Achievable = today ÷ (1 − u), for u = 20% and u = 35%
- 3. Untapped = achievable − today
- 4. Lift = untapped ÷ today
- 5. Untapped is capped at pipeline × no-decision rate
Worked example: a $10,000,000 pipeline at a 25% win rate produces $2,500,000 today. At u = 20%, achievable revenue is $3,125,000 and $625,000 is untapped. At 35%, achievable is $3,846,154 and $1,346,154 is untapped. With 50% of deals ending in no decision ($5,000,000), that means 12.5%–26.9% of the stalled deals would have to close.
The no-decision rate defaults to 50%, the middle of the 40–60% JOLT range; change it to your own. The rate also works as a ceiling: the calculator never shows more untapped revenue than your whole no-decision pipeline.
[05]Read this_
What the 20–35% figure is, and isn't.
- The figure is Parlay's analysis of third-party research: the arithmetic above, with every input cited and every assumption stated.
- The figure isn't a measurement of your pipeline, a figure published by any one of these sources, or a guarantee of results with Parlay.
- The sources differ. The studies come from different years (2012 to 2026), industries and sales motions, and several are vendors studying their own data or customers. Where we link a summary because the original report is no longer online, the citation says so.
- Your numbers will differ. The Pipeline Leak Audit looks at where your own pipeline leaks.
[06]Your pipeline_
Now run your numbers.
Put your own pipeline, win rate and no-decision rate through the same math.
