> For the complete documentation index, see [llms.txt](https://help.bizmachine.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.bizmachine.com/en/ai-prioritization/ai-prioritization-overview.md).

# AI Prioritization: overview of features and options

An AI agent evaluates the companies in your segment against your use case and gives each one a value from 0 to 100.

*Updated: 22 October 2026*

Do you have hundreds or thousands of companies in a segment and don't know where to start? Or do you offer several products to different types of customers and don't know what to offer to whom first? AI Prioritization does this work for you. You describe in plain language what you sell and to whom, and an AI agent gives each company a value from 0 to 100, based on how well it matches your use case. For each company, you also see why it got this value.

### Why AI Prioritization, and not just an AI assistant?

"I'll export a list from Prospector, upload it to my AI assistant and it will rank the companies for me." For a few companies, that really is enough. But a month later you want to rank them again, and your colleague has their own chat. With a hundred companies, the assistant also starts to mix up facts about different companies. AI Prioritization differs from Claude, ChatGPT, Copilot or another AI assistant in three ways: where the data comes from, what happens with the result, and how good the evaluation itself is.

|                                  | Your own AI assistant                                                                                                                                                                                                                                                                                                                                                                    | AI Prioritization in BizMachine                                                                                                                                                                                                                                                                                                                                                                                                                                             |
| -------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Where the data comes from**    | The web and the model's memory (unless you connect the assistant through the MCP connector). What it finds depends on what the company publishes. Financials, job ads or owners have to be looked up separately, with no guarantee they are current. Any other data you have to supply yourself, for example by uploading a file to the chat or through your own API integration.        | The same range of data for every company, where available in the company's country: business description, financials, full job ad texts, 90 days of signals *(as of October 2026)*, new hires, owners, technologies, vehicle fleet. The data is already in BizMachine, so there is nothing to export or upload.                                                                                                                                                             |
| **What happens with the result** | It stays in the chat. A month later you ask again and may get a different answer. Each colleague has their own chat. You pay your AI provider for every evaluation.                                                                                                                                                                                                                      | The value, the written reasons and the date of the calculation stay on the company in your BizMachine workspace: a column in the table, a filter, sharing with the whole organization. You can export the results to Excel, read them through the API, or let an assistant read them through the MCP connector. Need fresh data? Run the prioritization again on the same use case. The evaluation runs within your license and the limits of your BizMachine subscription. |
| **How good the evaluation is**   | One conversation holds all the companies at once. The longer the list, the more the assistant mixes up which facts belong to which company. To do it properly, you have to evaluate one company at a time and know which data matters in your industry. You also have to make sure data from different companies does not get mixed up. At that point you are building your own AI tool. | A team of AI agents built for this one job works on your segment. First, AI Prioritization sets an evaluation strategy from your use case and our know-how. Then each company gets its own agent. A thousand companies means a thousand separate evaluations, all on the same scale.                                                                                                                                                                                        |

All we need from you is a good description of what you sell. Our know-how does the rest. We keep a library that maps which company data matters for which industry and country, and which data would only lead the AI to guess. We combine your description with our library to prepare the evaluation strategy. For each company, you will see why it got its value.

AI Prioritization does not replace your assistant. It gives the assistant a starting point, so the assistant can skip the raw data and focus on what it does well. Rank the segment with AI Prioritization, then let your assistant, connected through the MCP connector, prepare call notes or a pre-meeting summary for the top twenty companies. You take care of your sales process, and we take care of the data and the evaluation logic. See [Through the MCP connector](#through-the-mcp-connector).

### How AI Prioritization helps you

Filters tell you which companies meet your criteria. But the result is often a list where all the companies look similar at first sight. Your team then calls them in order of revenue, in alphabetical order, or by intuition.

AI Prioritization helps you work through the companies in your segment in the right order. Your team starts with the companies that have the most potential for your use case, and for each one it has a specific reason why. This specific reason is what you need for the first call and for a meeting.

The agent also works with data that you cannot filter by in Prospector: it reads the full text of job ads, the company's business description, the ownership structure and a summary of company signals from the last three months *(as of October 2026)*. Filtering selects the segment. AI Prioritization tells you which companies in it to work on first.

### When to use AI Prioritization

You can run AI Prioritization on one company, on companies under a label or on a whole segment. Your choice depends on what you need:

1. **One company.** You receive an inquiry, or you want to check one specific company before a call. Click **Run AI Prioritization** on the company profile. You can also tick the company in the results table and click **Run AI Prioritization** in the **1 selected** bar.
2. **Companies under a label.** You want to see where your existing customers still have room for more business, or you have a list of companies from a conference. Save the companies under a label and run AI Prioritization on the label.
3. **A segment.** You want to find out which companies in the market to focus on. You may also be preparing a campaign and looking for companies to contact. Create a segment with filters and run AI Prioritization on the segment.

AI Prioritization also helps in these situations:

**Comparing several products.** Do you sell more than one product or service? Create a separate use case for each one. Each company then has a separate value for each product, and you see at once what to offer each company first.

**A campaign for a specific promotion.** Do you have a special price, a new product or a campaign for one market? Write a separate use case for it and run AI Prioritization on an existing segment. For example: "We have good prices for transport to Poland. We are looking for companies that export to Poland regularly." You then see which companies in the segment fit this one campaign best.

**Preparing for a meeting.** Click the value of any company to open its detail. You see the reasons, split into individual dimensions such as industry fit or financial health. Instead of a general "looks good", you know what to offer the company and why.

### How to run AI Prioritization

#### 1. Prepare a list of companies

Create a segment with filters, or save companies under a label. You run AI Prioritization on a segment, a label or selected companies.

#### 2. Click **Run AI Prioritization**

You'll find the button in the bar above the results, next to the **Export** button. For one company, you can also run AI Prioritization from the company profile.

![Run AI Prioritization button in the bar above the results, next to Export](https://247275268-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYkl7SJ14Nk0vzhVx7upl%2Fuploads%2Fgit-blob-bd9d91c8f80afa0cb9a77e6207dada0015c1252f%2Fai-prioritization-overview-tlacitko.png?alt=media)

#### 3. Choose or create a use case

Choose an existing use case under **Select Use Case**, or create a new one. A new use case has a **Name** and a **Value Proposition**: a description of your product and your ideal customer.

When you create a new use case, also set its **Visibility**. The default is **Only me**, so only you can see the use case. If you want to share it with your whole team, switch the visibility to the option for everyone in your organization.

You can write the use case yourself, or have an AI assistant write it for you through the MCP connector. The assistant works in your BizMachine workspace, sees your segments, labels and existing use cases, and suggests the new use case based on them. You'll find both ways, with examples, in [How to write a use case for AI Prioritization](https://help.bizmachine.com/en/ai-prioritization/ai-prioritization-tips.md).

![Run AI Prioritization dialog with the Name, Value Proposition and Visibility fields and the Options list](https://247275268-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYkl7SJ14Nk0vzhVx7upl%2Fuploads%2Fgit-blob-271bb24de0d9e7fcdbc1811140ee1be1c18369dd%2Fai-prioritization-overview-vyber-nabidky.png?alt=media)

#### 4. Choose the scope

Under **Options**, choose how many companies the agent evaluates:

* **First Results** to run a quick test and improve your use case
* **All Results** for the whole segment
* **Selected Results** if you have already ticked specific rows
* **Custom Count** for any number you choose

#### 5. Run it and wait

Click **Run AI Prioritization**. Depending on the number of companies, the evaluation takes from a few minutes to a few tens of minutes. You can follow the running job in the [Task Center](https://help.bizmachine.com/en/lists-and-export/task-center.md). In the meantime, you can keep working in Prospector.

#### 6. Look at the results

When the evaluation is finished, a new column with the value appears in the table. Its header is **AI:** followed by the name of your use case.

![AI Prioritization results: a new column with the value for each company](https://247275268-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYkl7SJ14Nk0vzhVx7upl%2Fuploads%2Fgit-blob-8eb2b4a5e7124e92d3f143c99df55b14f23d72e4%2Fai-prioritization-overview-vysledky.png?alt=media)

### How to read the results

Every company gets a value from 0 to 100. The higher the number, the better the company matches your use case.

Click a company's value to open its detail. You see:

* **The overall value** and a chart that shows it
* **Reasons for each dimension**, for example industry fit, technical need, financial health or growth dynamics
* **A written explanation** of why the company got this value

![Value detail for a company: the overall value and the reasons for each dimension](https://247275268-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYkl7SJ14Nk0vzhVx7upl%2Fuploads%2Fgit-blob-388aeae118b03e4d597ac556d579687b1fb374d7%2Fai-prioritization-overview-detail-skore.png?alt=media)

Each dimension has its own description. You see exactly what the agent took into account, and where the company is strong or weak.

### Filtering by AI Prioritization

When the evaluation is finished, your use case appears as a filter under **AI Prioritization** in the **Add filter** menu.

![Add filter menu with the AI Prioritization group open and the name of a use case as a filter](https://247275268-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYkl7SJ14Nk0vzhVx7upl%2Fuploads%2Fgit-blob-ed1ef190d31210f57451a252cec91aaeb44b2f2b%2Fai-prioritization-overview-filtrovani.png?alt=media)

You can:

* **Sort companies** from the highest to the lowest value, or the other way round
* **Filter by a value range**, for example only companies with a high value
* **Combine the AI Prioritization filter with other filters** and build a focused list for a specific sales activity

### What data the agent works with

To evaluate a company, the agent uses the following data, where it is available in the company's country:

* Business description, custom categories and keywords
* Financial indicators (revenue, profit, number of employees and details from the financial statements)
* Number of branches, e-shops and premises, and their ratings
* Open job positions from the last 180 days *(as of October 2026)*
* The full text of the last 20 job ads *(as of October 2026)*
* Company signals from the last 90 days *(as of October 2026)*
* New people and roles in the team from LinkedIn profiles, last 180 days *(as of October 2026)*
* Ownership structure and related companies
* Branches by region
* Technologies in selected categories (CRM, ERP, e-commerce and others)
* A monthly summary of vehicle registrations by make and vehicle type

The agent always chooses which of this data to use based on your use case. It compares that data with the description of your use case and evaluates how well the company matches it.

{% hint style="info" %}
Want more accurate results? Read [How to write a use case for AI Prioritization](https://help.bizmachine.com/en/ai-prioritization/ai-prioritization-tips.md).
{% endhint %}

### How to get data out of AI Prioritization

#### Export to Excel

Click **Export** above the results table and choose **Export AI prioritization**.

![Export menu with the options Export company data and Export AI prioritization](https://247275268-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FYkl7SJ14Nk0vzhVx7upl%2Fuploads%2Fgit-blob-ce4d3a08d731bd271ba70c2107b8fbf642406946%2Fai-prioritization-overview-export-menu.png?alt=media)

The export is a separate file with the prioritization results only. It has one row for each evaluated company, with:

* company identification (unique ID, Company ID, name)
* the value and its category
* the written reasons for the value
* a breakdown of the factors that affected the value, and their weights
* the date of the calculation
* a link back to the company profile in Prospector

If you also need company attributes (revenue, address, contacts), use **Export company data** as well. Both options are in the same **Export** menu.

#### Through the API

You can get the values of individual companies through the `indicators` API endpoint. For details, see [Overview of API endpoints](https://help.bizmachine.com/en/api-and-integrations/api-endpoints.md).

#### Through the MCP connector

If you have connected BizMachine to Claude, ChatGPT, Copilot or another AI assistant through the MCP connector, the assistant sees the values from your AI Prioritization directly. You don't need to export anything. Just ask:

> *"In the segment \[segment or label name], pick the 20 companies with the highest AI Prioritization value. For each one, give the value and the main reason for it."*

The assistant can summarize the results, compare the companies with each other or prepare notes for a meeting. This is useful when you want to keep working with the results as text, not in a table. You still have to run the evaluation itself in Prospector: the connector only reads the values. To connect it, see the [MCP connector](https://help.bizmachine.com/en/mcp-connector/mcp-connector.md) section.

### Related articles

* [How to write a use case for AI Prioritization](https://help.bizmachine.com/en/ai-prioritization/ai-prioritization-tips.md)
* [How credits and limits work in Prospector](https://help.bizmachine.com/en/lists-and-export/credits-and-limits.md)
* [How to create a segment in Prospector](https://help.bizmachine.com/en/search-and-filtering/creating-segments.md)
