Apollo.io has become the dominant outbound platform for B2B sales teams. Data, sequencing, calling, analytics, and CRM integration in one tool — and a pricing model that makes it accessible to businesses that can’t afford Outreach or Salesloft. For most B2B businesses at the growth stage, it’s a serious platform with serious capability.
The problem is that most implementations are wrong. Not accidentally — systematically. The defaults in Apollo are set up to get you sending quickly. They’re not set up to protect your deliverability, build a sustainable outbound motion, or produce accurate performance data. If you follow the onboarding flow and send sequences the next day, you’re almost certainly doing it wrong.
This is what a proper setup looks like.
Before you touch Apollo: the foundation
There are two things that need to be in place before any configuration in Apollo makes sense.
A clearly defined ICP. Not a rough description — an operational definition. The specific job titles, company size ranges, industries, geographies, and if possible, firmographic signals (funding stage, headcount trajectory, tech stack) that define the accounts and contacts you’re targeting. If you can’t translate your ICP into a set of Apollo search filters that reliably produces a list of the right people, the ICP isn’t defined tightly enough.
A clear value proposition tied to a specific problem. Your sequences will say something. What they say needs to be specific to a problem your ICP has, not a generic description of your capability. The problem with most cold outreach isn’t the platform — it’s that the message doesn’t land because it’s not about anything real to the recipient. Fix this before building sequences.
With both of those in place, the technical setup becomes meaningful. Without them, you’re automating a bad message at scale.
Domain and mailbox setup
This is where most businesses make their biggest mistake, and it’s the one with the most lasting consequences.
Use sending domains, not your primary domain. If you’re running outbound at any volume, you should not be sending from your main domain (yourcompany.com). Create one or two sending domains that redirect to your main domain — something like mail.yourcompany.com or outbound.yourcompany.com. If deliverability goes wrong and the domain gets flagged, it hasn’t destroyed your main domain’s reputation.
Set up DNS records properly. Every sending domain needs properly configured SPF, DKIM, and DMARC records. This is non-negotiable for inbox placement. Apollo provides guidance on this but it needs to actually be done — many implementations skip it and don’t notice the deliverability impact until months later.
Warm up mailboxes before sending. New mailboxes need to be warmed before you run sequences. Use a warmup tool (Instantly, Warmup Inbox, or similar) to build sending reputation over four to six weeks. During warmup, the mailbox is sending and receiving warm emails that signal to providers that it’s a legitimate, active inbox. Don’t send cold sequences from an unwarmed mailbox.
Limit mailboxes per domain. No more than two to three mailboxes per sending domain. Spreading sending across multiple domains also spreads risk — if one domain has a deliverability problem, the others keep running.
Sending limits
Volume limits are one of the most misunderstood aspects of outbound infrastructure. Higher volume does not mean better results — it means higher risk.
Per-mailbox sending limits. During warmup: 30–50 emails per day per mailbox. Post-warmup: 80–100 emails per day per mailbox. Anything above this significantly increases the risk of being flagged.
Sequence step intervals. Apollo lets you set the gap between sequence steps. Don’t compress it. A sequence step every two to three days is reasonable. A sequence with steps every day looks like spam — because by volume standards, it is.
Total sequence length. Four to six steps maximum. If someone hasn’t responded after six well-timed touchpoints, adding more steps isn’t going to change that. It’s going to get you marked as spam more often.
Building a sequence that works
Sequences in Apollo are multi-step cadences that combine email steps, manual tasks, and call steps. The structure of the sequence matters more than most people realise.
Step 1: The cold email. Short. Specific to a real problem. Not a pitch — an invitation to determine whether there’s relevance. The structure that works: a reference to something specific about their situation (not a compliment), one sentence on who you help and with what problem, one sentence on a proof point or outcome, a low-friction ask. Under 150 words.
Step 2: A follow-up that adds something. Not “just following up” — that’s the most hated phrase in cold outreach. A short email that adds a relevant data point, a different angle on the problem, or a genuinely useful piece of information. Something that would make sense to receive even if you weren’t being sold to.
Step 3: A LinkedIn connection or touch. Outside Apollo, but part of the cadence. Add the connection on LinkedIn without a note, a day or two after step 2. Creates a second channel of visibility without another email.
Step 4: A different angle. By step 4, you’ve made one attempt and one follow-up. This step should approach from a different angle — a different business problem you solve, a different proof point, or a different format (a short question rather than a mini-pitch).
Step 5: The breakup email. Short, honest, low-pressure. “I’ll stop reaching out after this — if the timing’s ever different, I’m easy to find.” Breakup emails often get replies that earlier steps didn’t. People respond to closure.
ICP targeting and list building in Apollo
Apollo’s search functionality is genuinely powerful — and consistently underused. Most users search by job title and company size and stop there. The filters available go significantly further.
Intent signals. Apollo surfaces intent data that can be used to prioritise accounts showing research activity in your category. This isn’t perfect data, but it’s a meaningful signal for prioritising who to reach out to first.
Technology filters. Targeting accounts using specific tools — a competitor, an adjacent platform, a tech stack you integrate with — allows for far more specific messaging. “I see you’re using [tool]” is a much stronger opening than “I saw you’re in [industry].”
Growth signals. Filtering for companies with headcount growth in specific departments, or recent job postings in relevant roles, identifies accounts with active need or budget. A company posting five sales roles is not the same as a company with a static headcount.
Clay integration. For the most precise list building, Apollo is the execution layer and Clay is the data layer. Clay handles waterfall enrichment — pulling contact data across multiple providers — and can build trigger-based lists that feed into Apollo sequences automatically. The two tools together are significantly more capable than either alone.
CRM integration: what “properly” means
Apollo has a native HubSpot integration. Most businesses turn it on and assume it’s working. Whether it’s working well depends on the configuration.
Sync bi-directionally. Changes in HubSpot should reflect in Apollo and vice versa. If a deal is closed or a contact asks to be removed from outreach, that needs to be reflected across both systems automatically.
Map deal creation correctly. When a sequence reply generates a meeting, that should create a deal in HubSpot at the right stage, associated with the right contact and company, without manual intervention. If deal creation is happening manually or not at all, you’re losing attribution and creating data hygiene problems.
Track sequence activity in HubSpot. Email opens, clicks, replies, and meeting bookings from Apollo sequences should log in the contact’s activity timeline in HubSpot. Your sales team needs this context — it’s the difference between a rep knowing what a prospect has already seen and going in cold.
Analytics and what to actually track
Apollo has a reasonable analytics suite. The metrics that matter:
Open rate. A proxy for subject line quality and deliverability. Below 30% is a problem. Above 60% on cold email is suspicious — may be bot clicks inflating the number.
Reply rate. The real performance signal. For well-targeted cold sequences with good copy: 3–8%. Positive reply rate (genuine interest, not unsubscribes): 1–3%.
Bounce rate. Above 5% is a deliverability warning. Above 10% means your list quality needs immediate attention.
Meeting conversion. Replies that convert to a booked meeting. Track this by sequence to understand which messaging is actually driving commercial outcomes, not just replies.
Unsubscribe rate. High unsubscribe rates suggest targeting or messaging problems — you’re reaching people who aren’t relevant, or relevant people who find your approach off-putting.
Review these metrics at the sequence level, not just in aggregate. A high overall reply rate masking a terrible-performing sequence means you’re optimising based on misleading data.
The most common Apollo failure mode is a business that’s technically running sequences but has no idea whether they’re working, a CRM that isn’t receiving the activity data, and a deliverability problem quietly building in the background. Getting the infrastructure right at the start prevents all of this.
For the full picture of how Apollo fits into a connected outbound stack — with Clay for data enrichment and HubSpot for pipeline management — the outbound infrastructure guide covers how the three tools work together. And if you want to understand what a proper Apollo implementation looks like in practice, the Apollo partner page has the details.
Free resources
Four phases for removing yourself from day-to-day sales without revenue dipping.
Free download22-point pre-raise checklist and a 90-day post-raise build plan.
Free downloadScore your revenue function across 5 areas. Find the gaps before you hire.
Free download