Outbound Math for Agencies: How Many Cold Emails Does It Take to Get a Client?

Outbound math for agencies is a chain of multiplications: emails sent, times reply rate, times the share of replies that are interested, times the share that become held meetings, times your close rate. Under the assumptions in AGP's projection, 50,000 emails a month produces about 4 new clients, or roughly 12,500 emails per client. This guide shows the formula, AGP's published numbers, worked tables and how to plug in your own rates, as part of the AGP growth machine.

Nerdy Joe is owned by Agency Growth Partners (AGP), so this is our own system and we are not neutral about it. The numbers below are either AGP's published results (attributed), AGP's projection (labeled), or our own illustrative arithmetic (labeled). None of them is a guarantee.

This is stage 11 of 12 in our series. The previous stage covered the agency CRM, dashboard and closing, and the next covers white label outbound for agencies.

The outbound formula

Every outbound forecast is the same five numbers multiplied together. Write it down once and reuse it for every campaign.

Replies             = Emails sent x Reply rate
Interested replies  = Replies x Interested rate
Meetings held       = Interested replies x Meeting rate
New clients         = Meetings held x Close rate
New revenue         = New clients x Deal value

Emails per client   = 1 / (Reply rate x Interested rate x Meeting rate x Close rate)
Emails needed       = Target new clients x Emails per client

Definitions matter, so fix them before you calculate:

AGP's projection at 50,000 emails a month

AGP publishes a projection for an engine sending 50,000 emails a month. It is a projection, not a result, and it rests on these assumptions:

Two things are implied by that projection, by our arithmetic. First, about 25 to 30 meetings from 40 interested prospects means roughly 62% to 75% of interested prospects become meetings. Second, about 4 clients from 25 to 30 meetings means a close rate of roughly 13% to 16%. If your close rate is far below that, the projection does not apply to you yet.

"Once it builds" is doing real work in that table. Long-cycle follow-ups add meetings from prospects who said "not now" months earlier, so the meeting count climbs over time instead of arriving in month one.

What AGP has actually published

Projections are only useful next to real numbers. AGP reports these on its results page and site.

AGP's own outbound, 90 days

Our arithmetic on those published figures:

• Reply rate per email: about 0.89% (6,300 divided by 704,000).
• Reply rate per prospect: about 1.6% (6,300 divided by 397,000), higher because each prospect received about 1.8 emails on average.
• Positive share of replies: about 12.4% (779 divided by 6,300).
• Emails per positive reply: about 904.

AGP's results page headline block also lists 12 closed deals a month from the team's outbound work. We do not divide closed deals by positive replies here, because the two figures come from different blocks and periods on AGP's page.

A live campaign, last 30 days

AGP reports one live campaign on the Artemis dashboard over the last 30 days at a 0.98% reply rate, with 8.39% of replies interested, 85 leads in pipeline and 39 open conversations. Those rates sit close to the 1% and 8% assumptions in the projection, which is why we use them as a starting point below.

One client engine, ten months

AGP also reports one client engine, run by the AGP team, that produced 613 qualified leads over ten months, noting that the monthly totals moved up and down while the system kept producing. That is the honest shape of outbound: monthly numbers vary, and the trend is what you manage.

Worked tables

These tables are our illustrative arithmetic, not AGP results. Baseline assumptions: 8% interested rate, 65% meeting rate and 15% close rate, which reproduce AGP's projection of about 4 clients at 50,000 emails and 1% replies.

Same volume, different reply rates (50,000 emails a month)

Same rates, different volumes (1% reply rate)

At these baseline rates, emails per client works out to about 12,800 (1 divided by 0.01 x 0.08 x 0.65 x 0.15). That is close to the roughly 12,500 implied by AGP's projection.

How to plug in your own numbers

Use this sequence with real data from your last campaign, or with conservative guesses if you have none.

1. Set your target. How many new clients a month do you want from outbound? Start with a number your delivery team can absorb.
2. Find your close rate. Signed clients divided by meetings held over the last six to twelve months, from your CRM. If you have no history, assume a lower close rate than you hope for.
3. Estimate your meeting rate. Of the people who said they were interested, how many actually held a call? Count no-shows as misses.
4. Start with 1% replies and 8% interested. These match AGP's projection and sit close to its published 30-day campaign. Replace them with your own figures after a month of sending.
5. Calculate emails per client, then multiply by your target to get monthly volume.
6. Check capacity. Can your domains, mailboxes and list support that volume without damaging deliverability? If not, the volume is the constraint, not the math.

A worked example with your inputs

Say your agency signs clients at a hypothetical $5,000 monthly retainer (your number, not an AGP price), keeps them 12 months on average, closes 20% of held meetings and holds 60% of interested replies as meetings. You want 3 new clients a month.

Now change one input at a time and watch which lever moves the result most. For most agencies, close rate and interested rate are cheaper to improve than raw volume, because they depend on the offer and the list rather than on more infrastructure. Our guide to lead generation pricing models covers how deal value and payment terms change what a lead is worth.

Why volume at the top matters

It is tempting to send a few thousand emails, see one meeting and decide outbound "doesn't work." The math explains why that conclusion is usually premature.

Small numbers are mostly noise

At 10,000 emails a month and our baseline rates, you would expect less than one new client. One good or bad month tells you almost nothing. At 50,000 emails, the same rates give enough replies (about 500) to compare lists, angles and offers with some confidence.

Testing needs volume

Artemis runs A/B variants per step and campaigns self-optimize toward messages that earn positive replies. That optimization needs enough replies per variant to separate a real winner from luck. Low volume slows learning, which slows every rate in the formula.

Long-cycle follow-ups compound

Many agency buyers are not ready today. Inside Artemis, "Not now" replies get dated follow-ups timed to the prospect's next project, so a reply in January can become a meeting in April. More volume today means a larger pool of dated follow-ups maturing in later months. AGP's timeline describes this as follow-ups compounding from month two onward.

Volume only helps if it lands in the inbox

More emails into spam folders is just more damage. Google's sender guidelines tell senders to keep their spam rate below 0.1% and to prevent it from ever reaching 0.3% or higher. That is why volume has to be spread across dedicated sending domains and many mailboxes.

Inside Artemis, an engine can send up to 500,000 emails a month, but sends are spread across mailboxes under daily limits, the client's main domain is never used for cold email, and volume drops automatically when a mailbox starts landing in spam. Our guide to agency cold email sending infrastructure covers how that capacity is built.

Volume does not fix a bad list

If the reply rate is low because the list is wrong, doubling volume doubles the waste and the spam complaints. Fix targeting first with a sharp ideal client profile for outbound, verify every address, then scale.

Common outbound math mistakes

• Using open rates. Opens are unreliable. Build the model on replies.
• Counting every reply as a lead. Out of office and unsubscribe replies are not interest.
• Ignoring ramp time. Expect about 3 to 4 weeks to launch, then a build-up period before the funnel reflects steady state.
• Assuming the best month is normal. Use a rolling average, not your best month.
• Forgetting capacity on the back end. Twenty-six meetings a month is a lot of calls for a founder who also runs delivery.

Who this math is not for

The model assumes a repeatable offer sold to a definable market. It fits poorly for:

• Agencies with fewer than a few hundred possible buyers in total, where volume runs out fast.
• Agencies that cannot yet say who they sell to and what it costs.
• Teams that need a client this week. Outbound has a build and warm period before any math applies.

For those agencies, offer and positioning work comes first. For everyone else, the math is the planning tool, and the complete growth machine shows how each stage moves one of the rates.

Want this built for your agency?

If you want to run these numbers on your own offer, close rate and deal value, we can do it together and tell you honestly whether outbound math works for your agency. Book a call with David. Read how AGP's growth machine works first if you want the full picture.

Sources

• Agency Growth Partners, home: https://agencygrowth.partners/
• Agency Growth Partners, Results (90-day outbound, headline block, client engine): https://agencygrowth.partners/results
• Agency Growth Partners, Artemis (live campaign dashboard, sending capacity): https://agencygrowth.partners/artemis
• Agency Growth Partners, Services: https://agencygrowth.partners/services
• Google, Email sender guidelines FAQ (spam rate thresholds): https://support.google.com/mail/answer/14229414?hl=en