Tracking & Improving: Metrics That Matter

Outbound email without measurement is guessing measurement without changing anything is busywork

This lesson sets up a simple weekly review routine of 20 minutes once a week that tells you exactly what to change to improve results.

The Only Metrics That Actually Matter

Reply Rate (Target: 5 to 10%)

Reply rate equals total replies divided by total delivered multiplied by 100.

This is the primary metric for a cold email campaign. It tells you whether the combination of list quality, signal, and copy is working.

Under 2% means something is wrong. Either the list is wrong, the signal is too generic, or the email is too long and salesy.

2 to 5% is average and workable but there is room to improve one variable.

5 to 10% is good. This is where precision targeting and signal-based personalisation gets you.

Above 10% is excellent. You have found the right niche, signal, and copy combination. Scale it.

Positive Reply Rate (Target: 2 to 4%)

Not every reply is positive. Measure them separately.

Positive replies are from people who are interested, asking questions, wanting to know more, or booking a call. Neutral replies are out of office responses, "wrong person try X" messages, or requests to forward to someone else. Negative replies are not interested, unsubscribe requests, or angry responses.

A 7% reply rate with a 1% positive rate means your email is starting conversations but the offer or signal is not landing with the right people. A 3% reply rate with a 2% positive rate is more valuable.

Bounce Rate (Keep Under 2%)

Bounce rate equals bounced emails divided by emails sent multiplied by 100.

A bounce happens when the email cannot be delivered because the address does not exist which is a hard bounce, or the inbox is full which is a soft bounce.

Hard bounces above 2% tell email providers your list quality is poor which damages your sender reputation. If your bounce rate is high, run the list through verification again before the next campaign.

Spam Complaint Rate (Keep Under 0.3%)

Google enforces this for senders above 5,000 emails per day. At the volumes in this course you will not trigger this threshold numerically but monitoring for complaints tells you whether your emails are being marked as spam rather than replied to.

What Open Rate Tells You (and Does Not)

Open rate has been unreliable since Apple's Mail Privacy Protection launched in 2021. Apple pre-loads emails in the background registering "opens" even when the person never looked at the email.

Do not optimise for open rate. Use it as a directional signal only. A very low open rate under 15% on a healthy list suggests the subject line is not working. A very high open rate over 60% with a low reply rate suggests the email body is not delivering on what the subject line implied.

Rely on reply rate for actual decisions.

The Weekly Review Routine (20 Minutes)

Every Monday morning run through this sequence.

Check numbers (5 minutes)

In Instantly or your own tracking sheet, note how many emails were sent, how many delivered, how many replied, how many were positive replies, and what the bounce rate is.

Read the replies (5 minutes)

Read every reply from the week. Look for patterns. What language do people use when they say yes? That is the signal language that works so use it in future hooks. What do people say when they say no? That tells you about the mismatch between who you are targeting and who actually wants this. Note any replies that were not yes or no but opened a different conversation.

Change one thing (5 minutes)

Only change one variable per review cycle.

If the reply rate is under 2%, change the hook by making the signal more specific. If still under 2%, question the list and ask whether these are the right people. If people are opening but not replying, change the bridge because the connection to the problem is not landing. If the ask is too big or too vague, make it smaller and clearer. If positive replies are low, reconsider the offer or the angle for this segment. If bounce rate is high, re-verify the list before the next send.

Plan next week's outreach (5 minutes)

Who are the next 20 to 30 prospects? Which campaign continues and which needs to change? What is the one variable being tested this week?

The Claude Review Prompt

After 2 weeks of running a campaign, run this:

Using @EMAIL.md — help me review and improve my outbound email campaign.

Campaign: [NAME]
Segment targeted: [INDUSTRY, ROLE, COMPANY SIZE]
Emails sent: [NUMBER]

Reply rate: [%]

Positive reply rate: [%]

Bounce rate: [%]

Replies received (paraphrase or paste):

[LIST THE REPLIES — positive and negative]

Current Email 1 body:

[PASTE EMAIL 1]

Tell me:

1. What the reply patterns suggest about the list quality
2. What the replies suggest about whether the signal and hook is landing
3. The one specific change I should make to the email first
4. Whether the segment or offer angle needs reconsidering
5. Whether the follow-up sequence needs any adjustment based on where in

the sequence most replies are coming from

When to Kill a Campaign

Kill a campaign and stop sending from it if after 50 or more emails sent the bounce rate exceeds 5%, the reply rate stays below 1% after testing 2 different hooks, you get 3 or more spam complaints, or the positive replies reveal you are targeting the wrong people entirely.

Do not kill a campaign after 10 emails. The sample size is too small to draw conclusions. Wait for at least 50 sends before making a definitive judgment.

Building Compound Improvement

The campaigns that work best are not necessarily the ones with the perfect first email. They are the ones that ran long enough to iterate.

Each week of data makes the next campaign slightly more accurate. You develop better understanding of which signals resonate, better language in the hook borrowed from replies, and better list targeting refined based on who replied positively.

After 3 months of consistent outbound your positive reply rate will typically be 2x what it was in the first month. Not because the market changed but because you understood it better.

Adding to EMAIL.md

## Campaign Tracking
Campaign: [NAME]
Segment: [WHO YOU TARGETED]
Dates: [START TO END]
Emails sent: [NUMBER]

Reply rate: [%]

Positive reply rate: [%]

Bounce rate: [%]

What worked: [YOUR NOTES]
What to change next time: [YOUR NOTES]

Your Action Right Now

Set up a simple tracking spreadsheet with these columns: Date, Campaign, Sent, Delivered, Replied, Positive Replies, Bounce Rate, and Notes. Update it every Monday.

If your first campaign is already running, check the numbers now and run the Claude review prompt with whatever data you have.