The useful metric
Track five things: how many scraped contacts survive filtering, how many messages actually arrive, how many replies are genuinely interested, how many turn into real conversations, and whether your accounts stay healthy. Messages sent is not on the list. It moves every day, it feels like progress, and it can go up while every one of those five gets worse.
Send volume is the most watched number in outreach because it is the most responsive. Do more work, watch it rise. It is satisfying in a way that the numbers that matter usually are not.
It is also the number most likely to mislead you, because it can climb while your targeting decays, your replies go unanswered, and an account drifts towards a limit. A month of high volume and low quality looks busy right up until it looks like a problem.
These five are less exciting and considerably more useful.
1. How much of your scrape survives filtering
Out of everyone you scraped, how many were actually worth contacting?
This is the health check on your sources. If a group used to yield 20% usable contacts and now yields 5%, you have worked through the good part of that community and are scraping the remainder: the people you already decided were not a fit.
It is also the earliest warning you get. Source quality decays before reply rates do, so watching this number saves you from running a campaign into a list that was never going to work.
Track it per group, not overall. The blended figure hides exactly the comparison you need, which is one source against another.
2. How many messages actually arrive
Sent and delivered are not the same number, and the gap is diagnostic.
Failures happen for ordinary reasons: deleted accounts, privacy settings that block messages from strangers, stale contact data. A modest failure rate is normal and not worth worrying about.
A rising one is worth investigating immediately, because it usually means one of two things: your list has aged badly, or the account doing the sending has a problem. Both are cheaper to catch here than three days later.
3. Interested replies, counted separately
Total replies is the number that lies most convincingly.
A vague message gets more replies than a precise one, because vague messages are easy to answer. If most of that extra volume is "who is this?", the campaign with the lower reply rate was the better campaign.
So split them. Interested, not now, not a fit, confused. Judge changes to your message or your targeting on the interested column, and treat a rise in the confused column as a warning rather than engagement.
| What went up | What it probably means |
|---|---|
| Interested replies | The targeting and the message are both working |
| Total replies, but not interested ones | Your message got vaguer, not better |
| Confused replies | Wrong people, or a message that does not explain itself |
| Opt-outs | Stop and look at the list before sending the rest |
4. Conversations that went somewhere
The number closest to money, and the slowest to arrive.
However you define a real opportunity in your business (a call booked, a proposal sent, a trial started), count it and trace it back to the group the contact came from. That link is the single most valuable thing your reporting can produce, because it tells you where to spend next month.
It arrives weeks after the campaign, which is exactly why the earlier numbers matter. They are what you steer by in the meantime.
5. Account health, which is not a performance metric
It belongs on the list for a different reason: it is the one that ends the programme.
Limits, warnings, unusual errors, sends failing in bulk. None of this measures how well outreach is working, and all of it determines whether you can keep doing it next week.
Check it before you look at anything else. A great reply rate from an account that is halfway to being restricted is not a result, it is a countdown, and results read against a damaged account are misleading anyway.
Never trade this one for the others. More volume can improve every metric above while quietly ruining this one, and it is the only one you cannot recover by trying harder next month.
How we checked this guide
The order runs from the earliest signal to the slowest: source quality moves before reply quality, which moves before revenue. That is deliberate, because the early numbers are the ones you can still act on while a campaign is running.
Use your data
Take your last campaign and write down these five. Most people find one has been quietly deteriorating for a while behind a healthy-looking send count.
Fewer numbers, each attached to a decision, beats a dashboard nobody acts on.
Trace the whole path: TeleBoost links lists, campaigns, replies, tickets and teams, so a conversation can be traced back to the group it started in without reconstructing anything.
Keep learning