> ## Documentation Index
> Fetch the complete documentation index at: https://docs.kime.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Campaign Builder

> Find the prompts where paid placement is worth buying, and turn them into OpenAI ad campaign proposals.

Campaign Builder helps you build an OpenAI ad campaign from the prompts where you already show up, or don't. It compares your **organic strength** on each tracked prompt with your competitors', sorts every prompt into one of four target areas, and generates campaign proposals for the prompts you select. The proposals land in [Ads Manager](/features/ads-manager) for review.

You find it in the sidebar under **Ads > Campaign builder**.

<Note>
  Campaign Builder is in beta. It only creates proposals. Nothing is sent to OpenAI Ads or spends money until someone on your team accepts a proposal in Ads Manager.
</Note>

## Key concepts

### Organic strength

Every brand gets an **organic strength** score from 0 to 100 on each prompt. It combines how often the brand is mentioned with how prominently it is placed:

**Organic strength = 0.6 × [visibility](/metrics/visibility) + 0.4 × [placement](/metrics/placement)**

Placement rewards being named earlier in the answer.

The **gap** on a prompt is your competitors' organic strength minus yours. A positive number means they are ahead of you, a negative number means you are ahead. The bigger the positive number, the more there is to win back.

### The four target areas

Each prompt is plotted with your organic strength on one axis and your competitors' on the other. The dashed lines at 50 split strong from weak, which gives four areas:

<CardGroup cols={2}>
  <Card title="Catch up" icon="person-running">
    Competitors are strong, you are weak. They show up well and you barely do. Ads get you into the answer while you work on organic presence.
  </Card>

  <Card title="Open" icon="flag">
    Both you and competitors are weak. Nobody has claimed the topic yet, so it is often the cheapest place to win attention.
  </Card>

  <Card title="Compete" icon="users">
    Both are strong. The answer is crowded, and ads help you stand out from brands that are also present.
  </Card>

  <Card title="Defend" icon="shield">
    You are strong and competitors are weak. Watch these prompts and protect them if competitors start to advertise.
  </Card>
</CardGroup>

## What you see

### Summary

The top of the page shows four numbers for the prompts that match your filters:

| Number | What it means |
| - | - |
| **Prompts in scope** | How many prompts match your filters. |
| **Your visibility** | Your average visibility across those prompts. |
| **Competitor visibility** | Your competitors' average visibility across the same prompts. |
| **Average gap** | How far competitors lead you on average. A positive number means competitors lead. |

A bar below shows your share of the **measured visibility** against your competitors'.

### Where to play

The **Where to play** tab shows where your opportunities are.

* **Target areas** chart: each bubble is a group of prompts placed by your organic strength (horizontal) and competitors' organic strength (vertical). The number on a bubble is how many prompts it holds. Prompts are grouped into 10-point cells so bubbles never overlap.
* **Coverage by category**: a table showing how many prompts in each category fall into Catch up, Open, Compete and Defend, plus the total.

Use **Slice by** to switch the chart between **Prompts**, **Categories**, **Countries** and **Tags**.

### What to run

The **What to run** tab helps you choose which prompts to advertise on, and shows where you trail.

* **You vs competitors per prompt**: two dots on one track for each prompt, showing your organic strength and your competitors'. Use **Rank by** (for example **Biggest gap**) to choose which prompts to surface.
* **Visibility by category**: your raw mention rate against your competitors' for each category. It is simpler than organic strength because it ignores placement.

<Tip>
  Hover any underlined word in Campaign Builder for a definition. Click **How this works** to reopen the in-app tutorial.
</Tip>

### Prompt panel

The panel on the right lists every prompt in your filter, grouped by target area. Each prompt shows its country, category and gap (for example **+42**). Use the checkboxes to choose which prompts to include, **All** or **Clear** to select or clear everything, and the sort menu (for example **Biggest gap**) to put the largest opportunities on top.

## Filters and settings

* **Date range**: pick custom dates, or use the **7 days**, **14 days** and **30 days** presets. Longer periods give more stable scores.
* **Target**: show all target areas or only one, for example only Catch up prompts.
* **Competitor set**: choose who to compare against. **Top 5 + your brand** (your five most visible competitors, plus you) is the default. You can also pick **All competitors**, **Your brand only**, or a single competitor.
* **Countries**, **Categories**, **Tags** and **Prompts**: limit the analysis to a market, product line or specific prompts.
* **Category priorities**: opens **Adjust ads priorities**, where you set how important each category is to you for ads recommendations (Medium by default). You can also reach it from Categories and Tags in [Workspace settings](/administration/workspace-settings).
* **How this works**: a short in-app explanation of the scoring and target areas.

## How to set it up

<Steps>
  <Step title="Track the prompts you want to advertise on">
    Campaign Builder works on the prompts in your workspace. Add the questions your customers ask before buying in [Prompts](/getting-started/setting-up-prompts), and organize them with categories and tags.
  </Step>

  <Step title="Add your competitors">
    The target areas depend on who you compare against. Add your main competitors in [Competitors](/getting-started/adding-competitors).
  </Step>

  <Step title="Add your markets">
    Proposals are built per country. Add every market you want to advertise in under [Markets & languages](/getting-started/localizing-prompts).
  </Step>

  <Step title="Set category priorities">
    Open **Category priorities** and set how important each category is to you for ads recommendations.
  </Step>

  <Step title="Connect Ads Manager">
    To launch proposals, connect your OpenAI Ads account in [Ads Manager](/features/ads-manager#how-to-set-it-up). You can generate and review proposals before connecting.
  </Step>
</Steps>

## How to use it

<Steps>
  <Step title="Choose your period and filters">
    Start with **30 days** for stable scores. Narrow by country, category or tag if you want a campaign for one market or product line.
  </Step>

  <Step title="Read the summary">
    Check **Your visibility**, **Competitor visibility** and **Average gap** to see how much ground there is to win.
  </Step>

  <Step title="Find where to play">
    On **Where to play**, look at how prompts spread across the four target areas. Use **Coverage by category** to see which categories have the most Catch up and Open prompts.
  </Step>

  <Step title="Select prompts">
    Every prompt in scope starts ticked. Click a bubble, bar or cell to add its prompts, shift-click to remove them, or tick and untick prompts in the panel on the right. Sort by **Biggest gap** to find the largest opportunities first.
  </Step>

  <Step title="Check where you trail">
    Switch to **What to run** to compare your organic strength with competitors' prompt by prompt.
  </Step>

  <Step title="Generate proposals">
    Click **Generate proposals**. KIME drafts campaigns, ad groups and ads from the selected prompts.
  </Step>

  <Step title="Review in Ads Manager">
    Open [Ads Manager](/features/ads-manager), go to **Campaigns** and switch to **Proposed**. Review each proposal, then **Accept** or **Reject** it.
  </Step>
</Steps>

## Get the most value

<AccordionGroup>
  <Accordion title="Build focused campaigns, not one big one" icon="bullseye">
    Generate proposals per target area or per category, for example one run for Catch up prompts and another for Open prompts. Focused campaigns get sharper ads and are easier to measure than one campaign covering every prompt.
  </Accordion>

  <Accordion title="Start with Open and Catch up" icon="flag">
    Open prompts are unclaimed and usually cheap to win. Catch up prompts are where competitors take customers from you today. Together they are normally where ad spend has the biggest effect.
  </Accordion>

  <Accordion title="Do not over-invest in Defend" icon="shield">
    You already win Defend prompts organically. Only add ads there if competitors start advertising on them, which you can check in Ads Library.
  </Accordion>

  <Accordion title="Fix the organic gap at the same time" icon="list-check">
    Ads buy you a place in the answer today. Use [Actions](/features/actions) to work on the content and sources that improve your organic strength, so you can reduce spend over time.
  </Accordion>

  <Accordion title="Re-run regularly" icon="rotate">
    Organic strength changes as AI answers change. Re-run Campaign Builder every few weeks and compare: prompts moving from Catch up to Compete or Defend show your work is paying off.
  </Accordion>

  <Accordion title="Compare paid and organic on the same prompts" icon="scale-balanced">
    After a campaign has run, compare what you bought with what you earned on the same prompts to see whether paid placement also lifts your organic visibility.
  </Accordion>
</AccordionGroup>

## Who it is for

* **Performance marketers** deciding where ChatGPT ad spend will have the most effect.
* **Brand and marketing leads** who want to close visibility gaps against competitors quickly.
* **Agencies** building data-backed paid AI search plans for clients before any budget is committed.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.