How to use this resource
Ask what you expected, what happened, which source produced the good conversations, what people objected to, whether the account came through it healthy, and what one thing you will change. Fifteen minutes, and the last question is the only one that matters. A review that ends without a change was entertainment.
The usual campaign review takes about ninety seconds. Someone looks at the reply rate, says it was roughly what we got last time, and everyone moves on to the next campaign.
The problem is not laziness, it is that reply rate genuinely does not tell you what to do differently. It is one number covering targeting, message, timing, account health and follow-up, and when it moves you cannot tell which of those moved it.
These six questions separate them out. It is worth doing properly after every campaign that mattered, and it takes about a quarter of an hour.
What to prepare
01Six questions, in order.
02Why you write down the prediction first.
03The question people avoid asking.
04How to stop reviews producing five changes at once.
1. What did you expect, and what happened?
Write the expectation down before you look at the results, or better, before the campaign ran.
This sounds like a formality and it is the most valuable line in the review. Without a written prediction, whatever happened will feel roughly like what you anticipated. Memory adjusts itself to the outcome, reliably and unconsciously.
With a number written down beforehand, you find out something real: not just whether the campaign worked, but whether you understand your own market. Being consistently wrong in the same direction is far more useful to know than any single campaign result.
2. Which source actually produced the good conversations?
Not replies. Conversations that went somewhere.
Break the results down by the group each contact came from. In most campaigns the difference between sources is much larger than the difference between message versions, and it is the finding you can act on immediately: scrape more like the good one, stop bothering with the other.
If every source performed about the same, that is informative too. It usually means your targeting is doing less work than you thought, and the results are being driven by the offer instead.
3. What did people push back on?
Group the negatives by reason, because they are not the same thing.
Silence is the hardest to read, which is why the confused replies are so valuable. They are the only people who tell you directly that something did not make sense.
| What they said | What it points at |
|---|---|
| "Wrong person / not my department" | Targeting |
| "We already use something" | Positioning, or the timing of your approach |
| "Who is this?" | The message failed to explain itself, or the list was wrong |
| Nothing at all | Either the wrong people, or a message nobody could answer easily |
4. How did the account come out of it?
The question people skip because the campaign is already over.
Check whether any account picked up limits, warnings or unusual errors during the run. A campaign that hit its numbers while quietly damaging an account has not succeeded, it has borrowed against the next one.
It is also the only way to learn where your real sending ceiling is. If two campaigns at a given pace went fine and the third caused problems, that is a boundary worth knowing before you plan the next one.
Note the pace you actually used, not the one you configured. They diverge more often than people expect.
5. Did anyone get left waiting?
The uncomfortable question, and worth asking honestly.
Count the replies that never got an answer, and how long the slowest ones waited. This is where campaigns quietly damage a business: not by annoying uninterested people, but by attracting interested ones and then ignoring them.
If the number is anything other than zero, the fix is upstream. You sent more than you could handle, and the next campaign should be smaller rather than better organised.
6. What is the one thing you will change?
One. This is the discipline that makes the whole exercise worth doing.
A review that produces five changes teaches you nothing, because when the next campaign performs differently you cannot attribute it to anything. Pick the single change with the largest expected effect, usually a source or an audience decision rather than a wording tweak, and leave everything else alone.
Write it down with the date. The next review starts by checking whether it did what you expected, which is how a series of campaigns turns into actual knowledge rather than a sequence of impressions.
Keep the reviews together. Six months of one-page reviews is the most valuable document your outreach produces, and nobody ever regrets having written them.
How we checked this guide
The order is deliberate: questions about who you contacted come before questions about what you said, because targeting effects are larger and easier to detect at outreach volumes. Keep the review about the decisions rather than the person who launched it. The moment reviews start assigning blame, people stop reporting the weak signals that make them useful.
Put it to work
Run this on your last campaign, including the prediction question. Write down what you think happened before you open the numbers.
The point of a review is not to grade the campaign. It is to make the next one a slightly better bet.
Everything the review needs, already recorded
TeleBoost keeps campaign settings, contacts, source groups, conversations and account health in one workspace, so a review starts from the data rather than from a reconstruction.
Keep learning