Results, with the denominators and the caveats
By Jānis Plūme, Founder, AllboundPros and Outbound Pros · 2026-08-06
Quick answer
Every number below comes from campaigns run inside the Outbound Pros group, with its denominator, sample size, period and class stated. Client names are withheld by default and appear only with written sign off. Our working benchmark for a sequence worth keeping is 0.5% to 1% positive on sends, measured per sequence after warm up. Our portfolio wide figure sits near 0.05% of sends and is derived rather than directly measured, which is stated on every appearance. Our best recorded positive reply ratios are in the 40% range, and those two families of number are not comparable, which is the most important sentence on this page.
Most agency results pages are a highlight reel with the denominators removed. This one is built the other way around, because a number you cannot check is not evidence and because the reader this site is for can do fractions.
What exactly do these numbers measure?
Three definitions are fixed in advance and applied everywhere on this site, and they are the reason two figures here that look comparable are not.
A positive reply is a prospect explicitly asking to talk, asking for materials, or referring us to the right person inside their company. "Not now, try Q3" is recorded separately and is not counted as positive. The definition is set before a campaign starts and does not move afterwards.
A positive reply rate divides positive replies by emails sent. This is the planning number, because it is the only one that converts directly into required volume.
A positive reply ratio divides positive replies by replies received. This is the iteration number, because it isolates copy and targeting quality from deliverability and volume. It is not a planning number and it is never placed next to a per send benchmark.
What are the per send rates?
| Figure | What it is | Sample and period | Class |
|---|---|---|---|
| Roughly 0.05% | Portfolio wide positive reply rate across all active sending | Rolling, all campaigns live at a given moment, including warm up and pre kill | Derived, see below |
| 0.5% to 1% workable, 1%+ strong | Our working benchmark for a sequence worth keeping | Per sequence, measured after warm up | Measured |
| Under 0.5% kill, 0.5 to 1% iterate, 1%+ scale, 2%+ pour | The decision thresholds we act on | Per sequence, per segment, over a window set before launch | Policy |
What are the positive reply ratios?
| Figure | What it is | Sample and period | Class |
|---|---|---|---|
| 15% rising to 45% | Positive reply ratio across three optimization cycles, one client in fact checking software | Three sequential cycles of targeting and copy change on the same infrastructure | Measured |
| 40%, sister variant 33.33% | Two variants of one sequence, one agency client, UK and EU segments | Same list, same window, two copy variants | Measured |
| 28.79%, from 19 positive replies | One sequence, with the absolute count published alongside the ratio | Single sequence | Measured |
What is the volume and scale behind them?
These belong to the group's execution reporting, not to this site, and they are stated here once, as group context, with the detail available from the group's outbound arm.
| Figure | What it is | Sample and period | Class |
|---|---|---|---|
| 44,649 emails, 377 replies, 0.84% | Weekly sending volume and reply rate on the largest account | One week | Measured, group reporting |
| 616 | Positive replies on record across the group | Cumulative, as of 2026-07-17 | Measured, group reporting |
| 36 active B2B clients, 1500+ campaigns | Active book of business | As of 2026-07-20 | Measured, group reporting |
| Since 2024, official Salesforge Expert Partner | Operating history and partnership status | Current | Verifiable externally |
Group sending infrastructure, mailbox and domain counts are the parent's figures, not ours, and they are described where the parent describes them. We do not restate them here, because a number republished on four domains stops being original data and starts being a footprint.
What does each number not measure?
This section is longer than the tables on purpose, and the answer for each figure is at least as useful as the figure.
The roughly 0.05% portfolio figure is derived, and that is the first thing to know about it. Two segments in the group's reporting carry published multiples of the fleet baseline, and dividing either segment's rate by its multiple lands near 0.05% of sends. Those segment measurements belong to our sibling property MultichannelPros and are published with their denominators there. What the figure describes is the mean across everything sending at a given moment, including sequences still in warm up, sequences in their first angle test, and sequences days away from being killed, which is why it is lower than any campaign we would show you as a success. We publish it because it is the correct order of magnitude to build a planning model on, and because an agency whose portfolio mean equals its case study is showing you a selected subset. It does not measure what a well targeted campaign achieves, it should not be used to judge any individual sequence, and it should not be quoted without the word derived attached. If it can later be pulled directly from group reporting with a source file and a period, it moves to measured and this page will say so.
The 0.5% to 1% working benchmark is a target applied to an individual sequence after warm up, neither an average nor a promise. It does not measure anything at the account or fleet level, and a client whose first sequence lands below it is not failing, they are at the start of the iteration loop the thresholds exist to govern.
The kill and scale thresholds are decision rules, not results. They tell you what we do at a given measured rate. They do not measure what your campaign will do, and applying them before warm up completes or on a sample too small to be stable is the most common way they get misused.
The positive reply ratios, 15% to 45%, 40%, 33.33% and 28.79%, all divide positives by replies received. Every one of them describes copy and targeting quality on a specific sequence. None of them measures volume, pipeline or revenue, and none of them can be compared to any figure that divides by sends. The 28.79% row publishes its absolute count of 19 positive replies specifically because a percentage without a count invites you to imagine a larger campaign than the one that happened. We would rather you saw the 19.
The 44,649 email week is a volume and deliverability proof from the group's execution reporting. The 0.84% is a reply rate, which counts every reply including rejections and out of office responses. It does not tell you how many of those 377 replies were positive, that count has not been pulled from the record, and we are not going to infer it. In particular, the 19 positive replies elsewhere on this page belong to the 28.79% sequence and are a different dataset entirely.
The 616 positive replies is a cumulative absolute count across many clients since 2024. It is not a per client expectation and it is not a rate. It tells you the group has operated at some scale for some time, which is all a cumulative count can ever tell you.
Client identity. Every figure here is anonymized by default. Client names appear only with written sign off, so descriptors like one client in fact checking software are deliberate rather than evasive. Where a named case study exists, it will say the name.
What is not on this page. No revenue figures, because we do not close deals for clients and attributing revenue to our activity would require assumptions we cannot verify. No inbox placement claim, because placement varies by mailbox provider and window and any single figure would be misleading. No customer testimonials, star ratings or review scores, and no rating markup of any kind, because we have not collected written approved testimonials and inventing or implying them is both illegal under the FTC rule on consumer reviews and testimonials and dishonest. No pricing.
Why do you refuse to compare a ratio to an industry average?
Because the two fractions divide by different things, so placing them side by side produces an arithmetic error printed as a result. The standard results claim in this industry takes a positive reply ratio, a number in the high single digits or the teens, and prints it next to an industry average that divides by sends. It reads as an advantage of ten to thirty times. It is an arithmetic error, not a result, because the first figure divides by replies received and the second divides by emails sent, and a comparison between two different denominators is not a comparison.
That comparison does not appear on this site, in any variant. Our own earlier marketing made it, which is part of why this property exists. The full explanation, with real numbers shown in both forms, is on the page about which fraction a quoted rate is using. If a competitor's results page shows you a number in the high single digits or teens as a positive reply rate, ask which fraction it is. The answer is nearly always the ratio.
What do you do with these numbers?
Put them into your own model. The calculator ships with the derived baseline as the conservative column and the working benchmark as the optimistic one, so you can see the spread before you commit to either. To see how these figures were produced and what the execution actually involves, start with the done for you audit. The methodology rules behind every figure here are set out on the about page.
How to cite this page. Cite the figure with its denominator and its class, for example: 28.79% positive reply ratio from 19 positive replies on one sequence, positives divided by replies received (AllboundPros, part of the Outbound Pros group, 2026). A rate quoted without its condition is not a fact about the channel, it is a fact about one campaign, and we would rather be cited accurately than cited often.
Frequently asked questions
Why are your published rates lower than other agencies'?
Because ours divide by sends and theirs frequently divide by replies. Converted to the same fraction the numbers are far closer than the marketing suggests. Our positive reply ratios reach 40% and 45% on real campaigns, which sits comfortably alongside anything published in this category. We simply refuse to print that number without saying what it divides by.
Why publish a fleet baseline that makes you look worse?
Because it is the number a client should plan with, and because the gap between a fleet mean and a case study is the single most useful piece of information an agency can give a prospect. A portfolio contains failures by design. Finding the sequence that hits 1% requires running the ones that hit 0.1%, and an agency showing you only the survivors is hiding the shape of the distribution you are buying into.
Can I see a named case study?
Only where a client has signed off in writing, and we will not push a client for that permission to close a deal. What we will do on a call is walk through anonymized campaign structure, real reporting, and the actual thresholds and decisions on a comparable account. That is more useful than a logo wall and it is what a serious buyer should be asking for anyway.
Do you guarantee results?
No. We do not guarantee revenue, a specific number of meetings, or specific reply rates, because those depend on the offer, the market and your ability to close. We guarantee execution: properly warmed infrastructure, verified leads, tested copy, consistent sending, full reporting and transparency. Any agency guaranteeing a meeting count is either pricing the risk into the retainer or planning to count meetings generously.
How current are these numbers?
The dated figures carry their date, the largest being 616 positive replies as of 2026-07-17 and 36 clients as of 2026-07-20. Rates and infrastructure counts are current at the last update at the top of this page. When a figure goes stale we change the number instead of the date.
Where do the numbers actually come from?
Campaign level reporting inside the group's own systems, across 36 active clients and 1500+ campaigns, reviewed against the raw sending platform data instead of against a summary dashboard. The full execution methodology, infrastructure, sending, warm up and reporting, is described by the team that produces these figures on the parent site.
Last updated: 2026-08-06
Run the arithmetic
on your own pipeline
Revenue target in, required pipeline, meetings, replies and sends out, split by channel mix.
Free. No signup, no email capture.