Virtual MindsThe outbound funnel template
Sent, accepted, replied, booked. Those four numbers are the whole funnel — this template defines each, gives you the ratio to watch between them, and names the wrong conclusion people draw from each.
- Time to use
- 30 minutes to set up
- What you end up with
- A funnel you can hold a provider to, including the question that exposes a vanity metric.
The four stages
| Stage | What it counts | What a drop here means |
|---|---|---|
| Sent | Messages that actually left, per channel | Sending infrastructure or list hygiene — not messaging |
| Accepted | Delivered and not bounced, spam-filed or rejected | Deliverability: domain reputation, warm-up, list quality |
| Replied | A human answered, positive or negative | Targeting or message — the two are separable, see below |
| Booked | A meeting that went on a calendar | Offer or qualification, not volume |
Count negative replies as replies. A campaign with no negative replies is usually not reaching anyone.
Separating targeting from message
A weak reply rate has two causes that look identical in a dashboard. Split them: hold the message constant and change the list, then hold the list constant and change the opener. Whichever move shifts the number is the cause. Running both changes at once is the most common way teams learn nothing from a month of sending.
The question that exposes a vanity metric
Ask any provider: "Show me sent, accepted, replied and booked for last month, per channel."
If the answer is a single "leads delivered" figure, that is not a funnel — it is a list. A list has no accepted stage, so it cannot tell you whether anyone received anything. This is the same reporting we hold ourselves to; see how we run outbound on our own system.
What to leave out
Open rate. Pixel-based opens are inflated by privacy proxies and prefetching to the point of being noise, and optimising for them pushes you toward subject lines that earn an open and no reply. Replies are the first honest signal.