Setting up your first cold-email pack.

Before you open the pack

Twenty templates do nothing if your domain is flagged before the first send. Before you touch any template, spend thirty minutes on infrastructure. Check your SPF record at mxtoolbox.com. Make sure DKIM is signed and passing. Add a DMARC policy of at least p=none with a reporting address so you can see failures. This is not optional prep work. It is the reason your emails land in primary instead of promotions.

If you are using a domain you have been sending from for more than ninety days with a clean bounce history, you are probably fine. If you registered this domain in the last sixty days, warm it up first. The industry standard warm-up is: 20 sends on day one, 40 on day two, doubling every two days until you reach your target volume. Tools like Instantly and Smartlead have built-in warm-up sequences that automate this.

On the sender side, four-mailbox rotation is what professionals actually use. One primary, three aliases across two subdomains. Spreading volume across mailboxes keeps any single sender below 50 outbound sends per day, which is the threshold where Google starts watching you more closely. Below that number, inbox placement rates typically sit above 85 percent. Above 100 sends per day from a single mailbox on a young domain, that number collapses fast.

Pick the three templates that match your motion

The pack has twenty templates. Most of them are not for you right now. The right three depend entirely on how you sell and at what volume.

If you are doing founder-led outbound to 50 named accounts a week, you want the short-form personalized opener (template 1), the case-study hook (template 7), and the breakup email (template 13). That sequence covers the full arc from cold introduction to last-chance follow-up.

If you are an AE running a 7-step automated sequence through Apollo or Smartlead, the mix shifts. You want the persona-led opener (template 3), the problem-agitation body (template 9), the social-proof bump (template 12), and the meeting-link nudge (template 17). Four steps, not three, because your sequence has a longer runway.

Read the description of each template once. Match them to the steps you already have in your sequence. Do not use a template just because it looks good in isolation.

Replace placeholders with real data

Every template ships with placeholders: [FIRST_NAME], [OFFER], [SPECIFIC_SIGNAL], [COMPANY], [MUTUAL_CONNECTION]. Your first thirty minutes with the pack is just filling these in for your first batch of targets.

[SPECIFIC_SIGNAL]is the one that separates replies from ignores. It should be something you actually observed about this specific recipient. They just raised a round. They posted about a hiring pain on LinkedIn last Tuesday. Their company opened a new vertical that matches exactly what you sell. A generic signal like “I noticed you are in the SaaS space” does not count. If you cannot find a real signal for a prospect, skip them and find one you can.

For the LinkedIn URL placeholder, always verify the link resolves before you send. A broken profile link in a personalized cold email is worse than no personalization at all. It signals that you automated everything without checking.

Send a five-send test before you scale

Before you load 200 prospects into a sequence, send the first email manually to five targets. Real targets, not your own test accounts. Watch what happens over 48 hours.

If you get one reply out of five, that is a 20 percent reply rate. That is the floor you need before scaling. Most well-tuned cold email sequences to a warm ICP land between 20 and 35 percent on the first email. If your five-send test comes back at zero replies, the problem is not volume. The template is wrong for your audience, the signal was generic, or the offer is not landing. Scaling a zero-reply template to 200 sends just gets you 200 non-replies and possibly a spam complaint or two.

Fix the template at five sends, not at two hundred. This is the single highest-leverage thing in this guide.

What to track in week one

Four numbers. That is it.

Reply rate per step. Not overall reply rate. Per step. If step one gets 18 percent and step three gets 2 percent, your follow-up is killing your sequence. Rewrite that step before adding more prospects.

Inbox placement.Check Google Postmaster Tools every other day in the first week. If your domain reputation drops from “high” to “medium,” stop sending and investigate before the algorithm flags you entirely. This is the early warning system. Most operators skip it and find out too late.

Unsubscribe rate. Above 1 percent means the audience is wrong, not the template. If people are actively opting out, you are reaching people who were never going to buy. Tighten your ICP filter before week two.

Send-from-mobile replies.These are short, lowercase, often just “interested, call me.” They are the highest-signal replies you will get. Someone replying from their phone was not at their desk, they were in a meeting or commuting, and your email still made them stop. Count these separately. A sequence that generates send-from-mobile replies is working.

After seven days you will know whether the pack is working for your motion. If reply rate is above 20 percent and inbox placement is holding at high, scale up. If not, come back to step two and pick different templates.