Split the Plan Across Risk Tiers

A plan built purely from ranked opportunity scores converges on the same safe middle, because every scoring input comes from what is already measurable. Portfolio thinking fixes the split before the ranking runs, so a known share of capacity goes to work that might fail and failure is budgeted rather than treated as a mistake. Apply this once you have two quarters of delivery history, because before that you cannot say what the safe tier actually returns.

O ficheiro da competência

---
name: content-bet-portfolio-mix
description: Use when every item on the content plan is a safe obvious topic and the programme produces steady small gains but nothing that changes the trajectory.
---

# Split the Plan Across Risk Tiers

A plan built purely from ranked opportunity scores converges on the same safe middle, because every scoring input comes from what is already measurable. Portfolio thinking fixes the split before the ranking runs, so a known share of capacity goes to work that might fail and failure is budgeted rather than treated as a mistake. Apply this once you have two quarters of delivery history, because before that you cannot say what the safe tier actually returns.

## O que precisa primeiro

- the last two quarters of published pages with clicks and conversions measured at 90 days per page
- committed capacity for the coming quarter in finished pages, not planned pages
- a written statement of which topics sit next to revenue and which are adjacent
- the stakeholder tolerance for a page that returns nothing, stated out loud before you start

## Método

1. Score last quarter's output in outcome bands rather than as an average: how many pages beat their forecast, matched it, and returned nothing. Averages hide that a handful of pages usually carry the entire programme.
2. Set the split before you open the backlog. Sixty percent proven, thirty adjacent, ten speculative is a common starting point, but the exact numbers matter less than fixing them first, because ranking the backlog always eats the speculative share.
3. Define each tier by evidence, not by feeling. Proven means this format has ranked for you on a comparable cluster. Adjacent means the topic is core but the format is untested, or the format is proven on a topic that is new. Speculative means neither.
4. Fill the proven tier first and cap it hard. Overflow goes to next quarter rather than into the adjacent budget, otherwise the cap is decorative.
5. Give each tier its own success test and horizon. Judge proven work at 90 days on clicks. Judge speculative work at 6 months on whether it produced something reusable such as a dataset, a format or a link source, even where traffic never arrived.
6. Agree in writing what a failed speculative piece costs before it is published. Unagreed failure gets relitigated as incompetence and the tier quietly disappears next quarter.
7. Rebalance once a quarter from the outcome bands, promoting a format to the tier above only after two independent wins. A single win is often a weak SERP rather than a repeatable method.

## O que isto produz

A one page quarterly allocation showing pages committed per tier, the success test and horizon for each, and the agreed cost of a failed bet.

## Onde isto corre mal

- ranking the backlog before setting the split, which produces a plan of nothing but safe work because safe work scores best on the metrics you already collect
- judging a speculative piece against the same 90 day click target as a proven one, so every experiment is logged as a failure and the tier is cut next planning round
- calling a format proven off one win, then scaling something that only worked because that particular SERP was weak

---

Da biblioteca de competências da QuQi - https://www.quqi.io/pt/skills/content-bet-portfolio-mix

O que precisa primeiro

• the last two quarters of published pages with clicks and conversions measured at 90 days per page
• committed capacity for the coming quarter in finished pages, not planned pages
• a written statement of which topics sit next to revenue and which are adjacent
• the stakeholder tolerance for a page that returns nothing, stated out loud before you start

Método

1. 01 Score last quarter's output in outcome bands rather than as an average: how many pages beat their forecast, matched it, and returned nothing. Averages hide that a handful of pages usually carry the entire programme.
2. 02 Set the split before you open the backlog. Sixty percent proven, thirty adjacent, ten speculative is a common starting point, but the exact numbers matter less than fixing them first, because ranking the backlog always eats the speculative share.
3. 03 Define each tier by evidence, not by feeling. Proven means this format has ranked for you on a comparable cluster. Adjacent means the topic is core but the format is untested, or the format is proven on a topic that is new. Speculative means neither.
4. 04 Fill the proven tier first and cap it hard. Overflow goes to next quarter rather than into the adjacent budget, otherwise the cap is decorative.
5. 05 Give each tier its own success test and horizon. Judge proven work at 90 days on clicks. Judge speculative work at 6 months on whether it produced something reusable such as a dataset, a format or a link source, even where traffic never arrived.
6. 06 Agree in writing what a failed speculative piece costs before it is published. Unagreed failure gets relitigated as incompetence and the tier quietly disappears next quarter.
7. 07 Rebalance once a quarter from the outcome bands, promoting a format to the tier above only after two independent wins. A single win is often a weak SERP rather than a repeatable method.

O que isto produz

A one page quarterly allocation showing pages committed per tier, the success test and horizon for each, and the agreed cost of a failed bet.

Onde isto corre mal

• ranking the backlog before setting the split, which produces a plan of nothing but safe work because safe work scores best on the metrics you already collect
• judging a speculative piece against the same 90 day click target as a proven one, so every experiment is logged as a failure and the tier is cut next planning round
• calling a format proven off one win, then scaling something that only worked because that particular SERP was weak

Use esta skill na sua própria IA

O ficheiro é markdown simples, com o nome e o gatilho no frontmatter. Quando um assistente consegue carregar skills sozinho, é esse frontmatter que lê para decidir que esta se aplica.

Mais em Estratégia de conteúdo