Apollo Email Marketing Setup, Limits and Deliverability

Can you actually use Apollo to send email?

Yes. In Apollo, email marketing means outbound sales sequences. Newsletters belong in a different tool. You build a sequence of outbound emails, enroll prospects, and Apollo sends each step from a mailbox you have connected to it.

Apollo does not send from servers of its own. Its help page on sending limits says emails go out through your connected mailbox (Apollo Knowledge Base, sending limits). The hard ceilings are mailbox caps set by Google or Microsoft, and every reputation hit lands on your domain.

Access is the cheap part. Apollo's pricing page says email campaigns come with every Apollo account, and the Basic plan includes email sequencing (Apollo pricing). The hard part is the setup around the sequence: domains, DNS records, warmup, how many mailboxes you run, and how clean the list is.

One disclosure before the steps: my team runs B2B outreach on email plus LinkedIn, and Apollo is just one of the databases we pull lists from. I have no stake in whether you keep it.

How do you set up an Apollo email sequence step by step?

The order matters more than the steps. Authenticate first, warm up second, add volume last. Every time.

1. Buy separate sending domains. Keep cold sequences away from the domain your invoices, contracts and support replies use. Add domains as your mailbox count grows, so no single domain carries the whole program.
2. Publish SPF, DKIM and DMARC on every domain. I put these records in place before a single message leaves, at any volume. They live in your DNS. Apollo does not manage them.
3. Connect the mailboxes. On paid Apollo plans you can link Gmail, Microsoft mailboxes or other providers.
4. Turn on warmup and wait it out. No real prospect hears from a new mailbox until warmup is done.
5. Build the list from more than one source, then verify it. I build targeted lists from RocketReach, ZoomInfo, UpLead, Apollo and Sales Navigator, never from a single export or a scrape. Every address then goes through a separate verification tool before it enters a sequence.
6. Write the steps. Keep the emails short and specific, and ask for a conversation; the copy and timing of each step are in how to build a cold email sequence. In my experience no copy rescues a weak list or a weak sender, which is why writing comes sixth.
7. Set daily and hourly limits for each mailbox using the table in the next section.
8. Enroll a small first batch. Read the replies and the bounces before you add anyone else.
9. Raise volume last, and do it by adding mailboxes.

How many emails per day can Apollo send from one mailbox?

Apollo recommends 50 to 100 emails per mailbox per day. The provider caps are far higher, and neither is the number to aim for.

My cold email setups usually spread across 10 to 20 domains and 20 to 40 mailboxes, and bigger companies get more of both. When a campaign works, I scale it by adding accounts, each one carrying a modest load. I never get more volume by pushing one mailbox harder.

Here is the per-mailbox load on one of our campaigns, whatever tool does the sending. A tech startup in the UAE sent about 1,250 emails a day across 50 email accounts. That is about 25 per mailbox, a quarter of Apollo's top recommendation and a tiny share of the provider caps. It brought 684 demos in 4 months. A sequence that fails at 200 sends will not start working at 20,000. Test small, then add accounts.

All the caps, side by side:

The Workspace figures come from Google's own limits page (Google Workspace Admin Help, sending limits). Apollo says these caps can change without notice.

Apollo's limits do not reset at midnight. They run on a rolling window, and emails over the limit stay scheduled and retry on their own. Google's caps also apply over a rolling 24 hours, and a user who hits one is blocked from sending for as long as 24 hours.

Apollo's counter does not count everything. Its Sent vs Limit counter includes sequence, manual, one-off and Chrome extension sends. Warmup traffic and email sent straight from the inbox outside Apollo do not count there. Google's cap covers the whole account, so the counter in Apollo can show headroom the mailbox no longer has.

Apollo's rule for raising the number is to start each mailbox at 50 a day and go higher only after reaching at least a 5% reply rate with a strong domain reputation. The page does not say where these thresholds come from.

Compare that with real campaigns. Replies usually land near 1%; at thousands of emails daily they often drop toward 0.5%, because larger volume lowers the average.

On those numbers, most sequences never reach the 5% bar Apollo sets for going past 50 a day.

Platform-wide data is kinder but makes the same point. Instantly's 2026 benchmark report puts the average cold email reply rate at 3.43%, with the top quartile at 5.5% and elite campaigns above 10%. So even there, a 5% reply rate is top-quartile territory. Only highly specific, heavily personalized campaigns clear it, and in my campaigns those can reach 15% to 17%.

To size a program, divide the daily target by the load per mailbox. At 25 emails a day each, 1,250 a day needs 50 mailboxes. At Apollo's starting 50 it needs 25, and at the top recommended 100 only 13, with no margin for a bad week.

Growing past that point is mostly a decision about mailbox count, which is the core of scaling cold email without breaking what already works.

Does Apollo still have email warmup?

Yes, through a third party. Apollo's current help article describes a working warmup feature run through Warmbox.ai. It also says Apollo does not control or manage that service and takes no responsibility for it (Apollo Knowledge Base, email warmup). So warmup is available in Apollo today, but another company runs it.

Apollo's recommended settings:

Apollo also says to let warmup finish before sending any automated or manual email from that mailbox. That matches my own floor: every new account gets 10 to 14 days at minimum. Apollo's two weeks sits at the top of that range. I never raise volume while warmup is still running.

A brand new domain has its own clock too, so read how long a new sending domain needs before cold outreach before you buy domains on Monday and plan to send on Friday.

Warmup also costs credits. According to Apollo, each seat on a paid plan typically covers warmup for one mailbox, and each additional linked mailbox costs 200 credits per 30 days of billing. So mailbox count drives cost as well as safety.

• 25 mailboxes on 2 seats leaves 23 extra mailboxes -> 4,600 credits every 30 days.
• 50 mailboxes on 4 seats leaves 46 extra mailboxes -> 9,200 credits every 30 days.

Warmup emails are system-generated test messages with "wbx" in the subject line. Filter them into a folder if they clutter the inbox. Apollo advises against mass-deleting them or flagging them as junk, because that works against the signal warmup is trying to build.

What do Gmail and Outlook require from Apollo senders?

Both providers judge your domain. Microsoft states this in its high-volume sender requirements: authentication belongs to your domain, whoever does the sending. For an Apollo user, that means your SPF, DKIM and DMARC records sit in your own DNS. Nobody at Apollo can fix them for you.

Google's side is spread over two official pages (Gmail Help, sender guidelines and Gmail Help, sender guidelines FAQ), and Microsoft's rule is in its high-volume sender announcement. Side by side:

Aligned, in that table, means the domain that passes SPF or DKIM is the same domain shown in your From line. Both rules have a narrow scope. Google's bulk guidelines cover mail to personal Gmail accounts only. Microsoft's rule covers consumer Outlook.com addresses (outlook.com, hotmail.com, live.com) only. Many B2B prospects sit on business mailboxes outside both rules. I still put the full bulk standard on every domain from day one, because it is written down and specific, and it is a one-time DNS job.

One line in Google's guidelines closes a common shortcut. All mail from the same primary domain counts toward the bulk threshold, so cold sequences sent from a subdomain of your main domain still count against your main domain. That is why step 1 says to use separate domains.

Unsubscribe rules matter for sequences too. Google asks for one-click unsubscribe only on promotional and marketing mail, excludes transactional mail, and lets recipients decide which is which. Treat a cold sequence as promotional.

On the Microsoft side, watch SPF. Microsoft warns that an SPF record needing more than 10 DNS lookups can fail. Each sending service you add to SPF can add lookups, so recheck the record every time your stack changes.

Where does Apollo-based sending break deliverability?

The failures I see most often are setup choices made before the first send:

1. Sending from the main company domain. Once that domain is flagged, invoices, proposals and support replies start landing in spam next to your outreach.

2. Too few mailboxes, pushed toward the top of Apollo's range. Two mailboxes at 100 a day look efficient on paper, but one bad week can burn both. AWS says big swings in volume and cadence can hurt sender reputation long term. A 10x jump in a week can trigger spam filters and temporary blocks, per MailTester. I would rather spread the same 200 emails over 8 mailboxes at 25 each.

3. Warmup cut short. Anything shorter than the two weeks above means real prospects hear from a mailbox with no sending history.

4. Raw database exports enrolled without an independent check. A database's own "verified" label is a starting point. A bounce spike can undo weeks of warmup. Infobip's deliverability docs say high bounce rates may get a sender throttled or blocked. Gmail's sender guidelines say to cut volume once messages start bouncing.

5. A list from a single source. One database means one set of blind spots and one set of stale records.

6. Scaling a sequence that never worked small. More mailboxes multiply a bad sequence as fast as a good one.

This is also where an agency setup differs from a tool setup. Every client campaign we manage uses dedicated sending IPs kept apart even between domains, and a single client may sit on 10 to 20 IPs. We handle domains, SPF, DKIM and DMARC, warmup and monitoring as one job. When you run Apollo yourself, each of those jobs is yours.

Should you keep sending from Apollo or move sending elsewhere?

Keep Apollo for data and light sequencing. Move the sending once your mailbox count, domain spread or monitoring needs outgrow what you can watch from inside Apollo. Moving the sending means Apollo stays your list source while a separate setup owns the domains, mailboxes, warmup and monitoring; the tool trade-offs are in Apollo alternatives for cold email.

That last row is how I run a lot of my B2B outreach. For a financial advisory firm in Canada, we built the lists from Sales Navigator and Apollo, opened on LinkedIn and followed up with warm emails: 421 booked sales calls in 12 months. On campaigns built this way, LinkedIn opens the conversation and email follows it.

Plan the volume with real conversion rates. In my campaigns, meetings usually come from 10% to 50% of replies. Replies run near 1% or lower at volume, so booked meetings are a small fraction of sends. If that is not enough for your targets, add mailboxes. Raising the per-mailbox limit will not help.

If you are rebuilding the stack around Apollo, start with the tools cold email needs, job by job.

Before you touch a limit in Apollo this week, write down mailboxes connected, emails per mailbox per day, and days each mailbox has been warming. If you run two mailboxes, add mailboxes before you add volume.

Running sequences from Apollo and watching bounces climb? Book a call and send me your sending setup. I will tell you where I would look first.