The 70/20/10 framework for a marketing budget that actually grows

Most marketing budgets are set by one of two methods: last year plus a bit, or whatever is left after everything else. Both lead to the same place, a plan that repeats what was comfortable and never finds the next channel.

The 70/20/10 model is a way to protect the core while deliberately funding the search for what comes next.

The split

70% on what’s proven. Channels and campaigns with a track record of returning more than they cost. This is the engine. The job here is efficiency: better creative, tighter targeting, lower cost per result, not reinvention.

20% on what’s promising. Things that have shown early signal but aren’t yet reliable. A channel you tested last quarter that looked good on a small budget. A new audience. A content format that outperformed once. This money buys evidence.

10% on what’s unknown. Genuine experiments. A platform you’ve never used, a wildly different creative approach, a partnership. Most of this will not work, which is the point: a small, capped budget for finding out.

Why the ratio matters

Without the 10%, nothing new ever gets tried, and the 70% slowly decays as channels saturate and costs rise. Without the 70%, you’re betting the quarter on ideas. The ratio isn’t sacred; a startup might run 50/30/20 and an established brand 80/15/5. What matters is that the three buckets exist and that money moves between them on evidence.

How to run it

1. Label every line item in the current budget as proven, promising or unknown. Be honest. “We’ve always done it” is not the same as “it works.”
2. Set promotion rules. What result moves something from unknown to promising, and from promising to proven? Write the thresholds down before the test starts.
3. Review quarterly, not annually. Channels change faster than fiscal years.
4. Kill things. The 70% bucket should lose a line item every so often. If it never does, the labels have stopped being honest.

Use the calculator above with your own number. If you’d like help deciding what belongs in each bucket, that’s a conversation we have with most new clients in the first week. Book one.