41,300 cold emails sent from 20 pre-warmed Microsoft 365 inboxes over 90 days produced a 6.8% reply rate, 94.2% primary inbox placement, and a cost per qualified lead of $11.40. Those are the headline numbers from a case study we ran with a 4-person lead generation agency between March and June 2026. This post publishes the full week-by-week data — including the week the campaign nearly died, the fresh-vs-pre-warmed split we didn't plan, and the exact math behind the cost per lead. No rounded-up vanity metrics. The raw numbers, what moved them, and how to replicate the setup.
💡 TL;DR
A 4-person lead gen agency sent 41,300 cold emails from 20 pre-warmed Microsoft 365 inboxes over 90 days. Results: 94.2% inbox placement, 6.8% reply rate, 2.9% positive reply rate, 119 qualified leads, and $11.40 cost per lead on infrastructure. The 5 fresh inboxes they ran alongside for comparison hit 61% placement and got 2 accounts suspended by week 4. Spam complaint rate stayed at 0.05% — well inside the 0.1% ceiling Microsoft now enforces. Total inbox cost: $99.80/month at $4.99/inbox from Litemail.
The 90-Day Scorecard
Before the story, here is the full scorecard. Every metric below comes from the agency's Smartlead dashboard, Microsoft 365 message trace logs, and Google Postmaster Tools for the Gmail-hosted portion of the list. Nothing is self-reported by us.
Metric | Result | 2026 Benchmark | Verdict |
|---|---|---|---|
Total emails sent | 41,300 | — | 90 days, 20 inboxes |
Primary inbox placement | 94.2% | ~83% average | +11 pts above average |
Open rate | 57.8% | 36–44% | Well above benchmark |
Reply rate | 6.8% | 4.2% | 62% above benchmark |
Positive reply rate | 2.9% | 1.5–2% | Strong |
Bounce rate | 1.4% | Under 2% target | Pass |
Spam complaint rate | 0.05% | Under 0.1% required | Pass with margin |
Qualified leads booked | 119 | — | $11.40 CPL (infra) |
Inboxes suspended | 0 of 20 pre-warmed | 2 of 5 fresh | The gap that matters |
For context on where these benchmarks come from, see our cold email open rate benchmarks for 2026 and the wider Microsoft 365 cold email reply rate data we've published from other campaigns.
Who Ran This Test — and Why They Picked MS365
The agency is a 4-person outbound shop serving B2B SaaS clients — mostly HR tech and fintech. They book meetings on commission, so deliverability isn't an abstract metric for them. It's rent.
They chose Microsoft 365 cold email infrastructure for one reason: their target lists skewed heavily toward companies running Outlook and Exchange Online. Roughly 68% of their prospects were on Microsoft-hosted mail. There's a persistent belief that Google Workspace is always the better sending platform. That's the first piece of common advice this case study contradicts.
In our testing at Litemail, MS365 senders reaching MS365 recipients consistently show a 4 to 7 point placement advantage over cross-provider sends. Same-ecosystem trust is real. If your list is Outlook-heavy, sending from Microsoft infrastructure is the correct call — not the compromise.
The Infrastructure Stack
Here's the full setup, so nothing is hidden:
20 pre-warmed Microsoft 365 inboxes from Litemail at $4.99/inbox/month — $99.80/month total. Each arrived with 4 to 12 weeks of genuine warm-up history and full admin access.
10 aged domains (2 inboxes per domain), all secondary to the agency's main brand. Their registrar and naming approach followed our pre-warmed domains setup.
SPF, DKIM, and DMARC pre-configured on every inbox before delivery. The agency verified all three with MXToolbox on day one — every record passed. If you're doing this manually, our SPF, DKIM, DMARC auto-setup guide covers the exact records.
Smartlead as the sending platform, connected via Microsoft OAuth. Connection took under 2 minutes per inbox.
Volume cap: 25 emails per inbox per day, ramping to 35 by week 5. Never higher.
And 5 fresh Microsoft 365 inboxes bought directly from Microsoft, warmed with a warm-up tool. Those were the control group. More on how that went later — it's not pretty.
Week-by-Week: Where the Numbers Actually Moved
Aggregate numbers hide the story. The week-by-week data shows exactly when things worked, when they wobbled, and what fixed them.
Period | Sends | Placement | Reply Rate | What Happened |
|---|---|---|---|---|
Week 1–2 | 4,900 | 95.1% | 7.2% | Postmaster showed Good/High within 48 hrs. Clean start. |
Week 3 | 3,400 | 81.6% | 3.9% | Placement crash. Bounce spike to 4.1%. See next section. |
Week 4–5 | 6,200 | 91.8% | 5.7% | Recovery after list fix. Volume ramp to 35/day resumed. |
Week 6–9 | 13,800 | 94.9% | 7.1% | Steady state. Best-performing stretch of the test. |
Week 10–13 | 13,000 | 95.3% | 7.4% | Placement improved as domain reputation compounded. |
Notice something? Placement in weeks 10–13 was higher than weeks 1–2. Pre-warmed history gets you a strong start, but reputation keeps compounding if your sending behaviour stays clean. Fresh inboxes never get this compounding effect in the first quarter — they spend it climbing out of Unknown reputation.
One more number worth flagging: open rates on Outlook-hosted recipients ran 61.2% versus 52.4% on Gmail-hosted recipients. Same copy, same list quality. That's the same-ecosystem advantage showing up in the data. Subject line rendering also differs between the two — our Outlook subject line display guide covers why the agency shortened subjects to 38 characters for the Outlook segment.
Week 3 Almost Killed the Campaign
Here's the part most case studies edit out. In week 3, placement dropped from 95% to 81.6% in five days. Reply rate nearly halved. The agency's first instinct was to blame the inboxes.
It wasn't the inboxes.
A client had handed over a 3,800-contact list scraped eight months earlier and never verified. The agency loaded it without running verification. Bounce rate on that segment hit 4.1% — double the 2% ceiling that Microsoft's filtering treats as a sender quality signal. Hard bounces at that rate tell Exchange Online you're sending to addresses you have no relationship with. Placement followed the bounces down.
🚩 The Fix — and the 6-Day Recovery
The agency paused the dirty segment, ran the full list through verification, removed 611 invalid addresses, and cut sending volume by 40% for six days while reputation recovered. By day 7 placement was back above 91%. Total cost of skipping verification on one list: roughly 3,400 impressions sent into spam and an estimated 8–10 lost meetings. Verification on that list would have cost about $15. Our email verification tools comparison covers which tools catch what.
The lesson isn't "verify your lists" — everyone says that. The lesson is that a bounce spike degrades placement within days on Microsoft 365, and recovery requires an active volume cut, not just cleaning the list and carrying on. Most teams we've worked with do the cleaning but skip the volume reduction, then wonder why placement stays stuck for three weeks. If you're already deep in a placement hole, our deliverability recovery guide and Outlook troubleshooting fixes cover the full protocol.
Fresh vs Pre-Warmed: The Split Test We Didn't Plan
The 5 fresh Microsoft 365 inboxes were supposed to be a cost-saving experiment. They became the most useful data in the study.
Metric (90 days) | 20 Pre-Warmed Inboxes | 5 Fresh Inboxes |
|---|---|---|
Time to campaign-ready | 24 hours | 6+ weeks warming |
Primary inbox placement | 94.2% | 61.3% average |
Reply rate | 6.8% | 2.1% |
Accounts suspended | 0 of 20 | 2 of 5 by week 4 |
Postmaster reputation | Good/High in 48 hrs | Unknown for 5+ weeks |
True monthly cost per working inbox | $4.99 | $14.20 (incl. warm-up tool + losses) |
The suspensions are the number to sit with. Microsoft's 2026 outbound policies flag new tenants that show cold-email-shaped behaviour early in account life. Two fresh accounts got restricted in week 4 despite conservative volume. The warm-up tool subscription, the six weeks of waiting, and the two dead accounts pushed the real cost of "cheap" fresh inboxes to nearly triple the pre-warmed price. We ran a bigger version of this comparison in our fresh vs pre-warmed MS365 field test — same pattern, larger sample.
Fair warning though: pre-warmed inboxes are not a licence to blast. The agency held every inbox at or under 35 sends per day. Warm-up history buys you trust. Volume discipline keeps it. Our safe daily volume guide covers the exact caps by inbox age.
Cost Per Lead: The Math Nobody Publishes
Reply rates are nice. Cost per lead is what the agency's clients pay for. Here's the full calculation, with nothing hidden in "miscellaneous."
Line Item | 90-Day Cost |
|---|---|
20 pre-warmed MS365 inboxes ($4.99/mo each) | $299.40 |
10 domains (annual cost, prorated) | $36.00 |
Smartlead subscription (prorated share) | $94.00 |
List verification | $61.00 |
Lead data (Apollo export credits) | $867.00 |
Total | $1,357.40 → 119 qualified leads → $11.40/lead |
Strip out lead data and the pure infrastructure cost per lead is $4.12. For comparison, the agency's paid LinkedIn campaigns for the same clients ran $67 to $140 per qualified lead in the same quarter. Cold email at this placement level isn't just cheaper — it's a different economic category. According to Backlinko's analysis of 12 million outreach emails, the average cold email reply rate sits at 8.5% for emails that reach the inbox. Placement is the multiplier on everything else you spend.
You might be thinking — but what about compliance risk in that math? Reasonable question. The agency ran B2B-only lists with unsubscribe links and physical addresses on every send, which kept them inside CAN-SPAM and GDPR legitimate-interest territory. The rules differ enough by region that it's worth 15 minutes with our CAN-SPAM and GDPR cold email guide before you copy this playbook into the EU.
Replicating These Microsoft 365 Results — 7 Steps
This setup isn't complicated. It's disciplined. Here is the exact sequence to reproduce the case study numbers with your own Microsoft 365 cold email campaign.
Buy secondary domains — never send from your main domain. 2 inboxes per domain, maximum. The agency used 10 domains for 20 inboxes.
Get pre-warmed MS365 inboxes with verifiable history. Postmaster-verified reputation within 48 hours is the test that can't be faked. Litemail inboxes ship with 4 to 12 weeks of genuine warm-up and full admin access — plans start at $4.99/inbox on the pre-warmed inbox page.
Verify SPF, DKIM, and DMARC before the first send. All three, on every domain. One failing record cancels the warm-up history you paid for.
Verify every list, every time. Target under 2% bounce rate. The week-3 crash in this study came from skipping this once.
Cap volume at 25/day per inbox, ramp to 35 by week 5. Not 50. Not 100. The suspension-free record came from this cap.
Match sender ecosystem to recipient ecosystem where you can. Outlook-heavy list? Send from Microsoft 365. Gmail-heavy? Consider a Google Workspace inbox mix at a 60/40 split.
Watch bounce and complaint rates weekly. Bounces under 2%, complaints under 0.1% — ideally under 0.05% like this campaign. Cross either line and cut volume by 40% for a week before it becomes a reputation problem.
If you're building this for multiple clients rather than one campaign, the full agency-scale version — domain pools, inbox rotation, client reporting — is in our cold email infrastructure overview.
What the Agency Would Do Differently
We asked the agency's founder what they'd change on round two. Three honest answers, none of them flattering to conventional wisdom.
✂️1. Skip the fresh-inbox experiment entirely
The 5 fresh inboxes consumed six weeks of attention, a warm-up tool subscription, and produced 2 suspensions and a 2.1% reply rate. Total value delivered: one useful comparison table. The founder's words: "We paid $70 to learn what a $5 inbox already solved."
📉2. Ignore open rates earlier
Here's the advice this study contradicts directly: "optimise your open rate first." With Apple Mail Privacy Protection and Outlook's link scanning inflating opens, the metric is too noisy to steer by. The agency wasted week 2 A/B testing subject lines against phantom opens. Reply rate and placement rate are the only two numbers they steer by now.
🗓️3. Start the second inbox batch at week 6, not week 12
Scaling from 20 to 40 inboxes at the end of the study meant a flat month while new capacity came online. Ordering the second batch at week 6 — once the first batch proved stable — would have pulled roughly 30 extra leads into the quarter. Capacity planning math is covered in how many pre-warmed inboxes you need.
Use Instantly or Smartlead? Keep Your Stack — Fix the Inboxes
Every result in this case study came from the inbox layer, not the sending tool. The agency used Smartlead, but the same infrastructure connects via OAuth to Instantly, Lemlist, Saleshandy, and Apollo in under 3 minutes per inbox. If you're currently renting platform-locked accounts, this is the swap that changes your numbers without changing your workflow — we broke down the rental trade-offs in our Instantly review.
Replicate this case study's infrastructure: pre-warmed Microsoft 365 inboxes from $4.99/inbox — full admin access, dedicated US and EU IPs, automated DNS, Postmaster-verified within 48 hours. Your inboxes survive any platform decision you make later. Get Pre-Warmed MS365 Inboxes →
The Bottom Line
20 pre-warmed Microsoft 365 inboxes produced 94.2% placement and a 6.8% reply rate across 41,300 sends — with zero suspensions in 90 days.
Fresh MS365 inboxes in the same campaign managed 61.3% placement, a 2.1% reply rate, and lost 2 of 5 accounts to suspension by week 4.
Infrastructure cost per qualified lead came to $4.12 — total cost per lead including data was $11.40, versus $67+ on paid channels.
One unverified list crashed placement by 13 points in five days. Keep bounces under 2% and complaints under 0.1%, and cut volume 40% for a week when you breach either.
Match sender ecosystem to your list: Outlook-heavy recipients opened 61.2% of MS365-sent emails versus 52.4% on Gmail-hosted addresses.
Hold volume at 25–35 sends per inbox per day. The zero-suspension record came from discipline, not luck.
Frequently Asked Questions
What reply rate is realistic for Microsoft 365 cold email in 2026?
With pre-warmed inboxes, verified lists, and volume held under 35 sends per day, a 5 to 7% reply rate is realistic — this case study hit 6.8% across 41,300 sends. On fresh inboxes with Unknown reputation, expect 2 to 3% at best, because a large share of your emails never reach the primary inbox.
Is Microsoft 365 or Google Workspace better for cold email?
It depends on your list. If most of your prospects are on Outlook or Exchange Online, Microsoft 365 cold email typically shows a 4 to 7 point placement advantage from same-ecosystem trust. For Gmail-heavy lists, Google Workspace wins. Most agencies run a 60/40 mix so every recipient segment gets a same-ecosystem sender.
How many emails per day can one Microsoft 365 inbox send safely?
For cold outreach, 25 to 35 per day per inbox is the safe range for pre-warmed accounts — this study ran 25/day for four weeks, then 35/day, with zero suspensions. Microsoft's technical limits are far higher, but cold-email-shaped behaviour at high volume triggers restrictions fast, especially on newer tenants.
Why did the fresh inboxes get suspended but the pre-warmed ones didn't?
Microsoft flags new tenants that show outbound-heavy behaviour early in account life. Fresh accounts have no sending history to offset that pattern, so identical volume that's safe on an aged account triggers restriction on a new one. The pre-warmed inboxes in this study carried 4 to 12 weeks of genuine send-and-reply history, which is what kept all 20 alive through 90 days.
What caused the week 3 placement crash, and how long did recovery take?
An unverified 3,800-contact list pushed bounce rate to 4.1% — double the 2% ceiling. Microsoft's filtering treats hard bounces as a sender quality signal, and placement dropped 13 points in five days. Recovery took six days: remove invalid addresses, cut volume 40%, resume ramping. Cleaning the list without cutting volume typically stretches recovery to two or three weeks.
Where do the pre-warmed MS365 inboxes in this case study come from?
Litemail. Each inbox costs $4.99/month and ships with 4 to 12 weeks of genuine warm-up history, SPF, DKIM, and DMARC pre-configured, dedicated US and EU IPs, and full Microsoft 365 admin access — verified Good or High in Postmaster Tools within 48 hours. No minimum order, delivered within 24 hours, and they connect via OAuth to Smartlead, Instantly, Lemlist, and every major platform.
What's a good cost per lead for B2B cold email?
This campaign delivered qualified leads at $11.40 all-in, or $4.12 counting infrastructure only. Anything under $25 per qualified B2B lead is strong in 2026 — paid social and LinkedIn typically run $67 to $140+ for comparable lead quality. Placement rate is the biggest lever: the same spend at 61% placement instead of 94% roughly triples your cost per lead.
Do I need to verify my email list even if it's from a reputable data provider?
Yes — every list, every time. B2B contact data decays at roughly 2 to 3% per month as people change jobs, so even a clean export goes stale within a quarter. Verification costs a few dollars per thousand contacts. The unverified list in this study cost an estimated 8 to 10 lost meetings in a single week.
Run the Same Playbook — Pre-Warmed MS365 Inboxes from $4.99
Every number in this case study started with the inbox layer: 4 to 12 weeks of genuine warm-up history, SPF/DKIM/DMARC configured before delivery, dedicated US and EU IPs, and full Microsoft 365 admin access you own outright. Postmaster-verified within 48 hours. Delivered in 24. Works with Smartlead, Instantly, Lemlist, Saleshandy, and Apollo via OAuth.
Get Pre-Warmed Inboxes from $4.99 →
No minimum order · MS365 and Google Workspace available · Automated DNS · US and EU IPs included
About Litemail — Litemail provides pre-warmed Google Workspace and Microsoft 365 inboxes for cold email outreach. From $4.99/inbox with automated DNS setup, dedicated US and EU IPs, 4 to 12 weeks of genuine warm-up history, and full admin access. View pre-warmed inbox plans →
Related reading: Microsoft 365 Cold Email Reply Rate Data 2026 · Fresh vs Pre-Warmed MS365 — Field Test 2026 · Troubleshooting Microsoft 365 Cold Email Inboxes for B2B Sales · Pre-Warmed MS365 Inboxes for Lead Gen Agencies · Best Pre-Warmed Inbox Providers 2026 (Ranked) · Microsoft 365 Cold Email for Startups · Litemail — Pre-Warmed Inboxes, Plans and Pricing
