---
slug: budget-entscheiden
title: "Which campaign actually delivered?"
category: berichte
locale: en
updated: 2026-10-17
keywords: ["budget","effect","experiment","switchback trial","attribution","decision","campaign"]
questions: ["What do I move my budget on, if attribution does not measure effect?","Reallocate budget","Should I move budget from Google to Meta?","Which campaign to switch off","Effect instead of attribution","Measure incrementality","Is display worth it","Is my Meta advertising worth it?","What does the switchback trial tell me?"]
url: https://bench.metrunio.com/en/help/budget-entscheiden
---

# Which campaign actually delivered?

**What do I move my budget on, if attribution does not measure effect?**

The “Check budget” area sorts your campaigns into three levels: measured (an experiment found a difference), attributed (a model credits it a share) and not answerable (with a reason). Only the top level says anything about effect, and it always carries its interval.

Attribution answers the question “which share do I credit to which campaign”. It does not answer the question that comes before a budget decision: “would the conversion have happened without this campaign”. The “Check budget” area keeps the two apart rather than passing one off as the other.

| Level | Where it comes from | What you can do with it |
| --- | --- | --- |
| Measured | A completed experiment: two groups, one difference, one interval | Move budget, since this is the only level that evidences effect |
| Attributed | Attribution under the chosen model, together with ad spend | Form a suspicion and start an experiment |
| Not answerable | Costs or attributed conversions are missing | Supply the missing figure |

> **Note:** Sorted by level, not by revenue. What was measured is at the top, even if another campaign carries more. The level is the actual statement.

> **Note:** The next step depends on your plan. On Growth it is “start switchback trial”: you switch the campaign off for a while and back on, and comparing the periods tells you, with an interval, whether the conversions would have happened without it. On Start and Pro, “start experiment” carries a lock with “from Growth”. The “measured” level in this table comes from two campaign groups, so on Scale and the agency plan.

## Why it ends with “start experiment”

For an attributed row the next step is not “move budget” but “start experiment”. That is the difference from a media mix model: where other tools output a daily figure across all channels, we output the action that produces the figure in the first place. A number that comes from a regression over past weeks looks exactly like a measured one but is something else.

## When the interval contains zero

Then it says “no measurable difference”. That is a result, not a missing answer: the trial ran and found no difference beyond the noise. Nothing is extrapolated, and the campaign still stays on the top level: you now know more than before.

> **Caution:** Only experiments of the “two campaign groups” design appear here. A switchback trial (“on, off, on”) applies to the whole account and cannot be attributed to a single campaign. That would be an assignment again, and this level is meant not to be one. Its result is on the experiments page. For small shops it often ends with “no measurable difference”: that means “not shown”, not “does not work”.

## Also asked

- Reallocate budget
- Should I move budget from Google to Meta?
- Which campaign to switch off
- Effect instead of attribution
- Measure incrementality
- Is display worth it
- Is my Meta advertising worth it?
- What does the switchback trial tell me?

## Related

- https://bench.metrunio.com/en/help/experimente
- https://bench.metrunio.com/en/help/roas-und-kennzahlen
- https://bench.metrunio.com/en/help/attributionsmodelle
- https://bench.metrunio.com/en/help/werbekosten-importieren
- https://bench.metrunio.com/en/help/tarife