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

# Prepare your index structure

> Learn how to prepare your index structure for Advanced Personalization.

To optimize personalization, you should configure your existing index's structure to align with the requirements of the Advanced Personalization feature.
A well-prepared index structure enhances this feature's ability to deliver personalized search experiences for your website or app.

<Callout icon="credit-card" color="#c084fc">
  This feature isn't available on every plan.
  Refer to your [pricing plan](https://www.algolia.com/pricing) to see if it's included.
</Callout>

## Use categorical attributes

**Categorical attributes** are attributes with a fixed number of possible string values.
Incorporating such attributes into your index structure categorizes your data into well-defined, smaller buckets.
For example:

```json JSON icon=braces theme={"system"}
[
  {
    "objectID": "ID01",
    "title": "Chair",
    "color": "red",
    "categories": ["Furniture", "Outdoors"]
  },
  {
    "objectID": "ID02",
    "title": "Table",
    "color": "green",
    "categories": ["Furniture", "Outdoors"]
  },
  {
    "objectID": "ID03",
    "title": "Water bottle",
    "color": "blue",
    "categories": ["Outdoors"]
  },
  {
    "objectID": "ID04",
    "title": "Book",
    "color": "red",
    "categories": ["Books"]
  },
  {
    "objectID": "ID05",
    "title": "Headphones",
    "color": "green",
    "categories": ["Electronics"]
  },
  {
    "objectID": "ID06",
    "title": "Phone",
    "color": "blue",
    "categories": ["Electronics"]
  },
  {
    "objectID": "ID07",
    "title": "Television",
    "color": "red",
    "categories": ["Electronics"]
  },
  {
    "objectID": "ID08",
    "title": "Can",
    "color": "green",
    "categories": ["Cooking & Dining", "Kitchenware"]
  },
  {
    "objectID": "ID09",
    "title": "Bowl",
    "color": "blue",
    "categories": ["Cooking & Dining", "Kitchenware"]
  }
]
```

### Good categorical attributes

An attribute like `color` with finite values like red, green, and blue is a good categorical attribute because it organizes your index into three distinct buckets.
Other good categorical attributes include `gender`, `brand`, and `categories`.

### Bad categorical attributes

Attributes that are unique for each record, such as `objectID` and `title`, are poor choices because they don't offer any grouping into buckets.
Other poor categorical attributes include `description`, `sku`, and `price`.

## Avoid nesting attributes

Aim to [organize data within records](/doc/guides/sending-and-managing-data/prepare-your-data) with a flat structure.
Reserve deeper nesting for [hierarchical facets](/doc/guides/managing-results/refine-results/faceting#hierarchical-facets).

Consider the following example of several nested keys:

```json JSON icon=braces theme={"system"}
{
  "key1": {
    "key2": {
      "key3": "value"
    }
  }
}
```

To make it easier for Advanced Personalization to process the index,
[simplify the attribute-value pair](/doc/guides/sending-and-managing-data/prepare-your-data#simplify-your-records):

```json JSON icon=braces theme={"system"}
{
  "key1_key2_key3": "value"
}
```

To optimize performance, Advanced Personalization limits the number of key-value pairs for nested attributes to 50.
Adhering to this limit ensures efficient processing of your nested attributes.
If you need to exceed these constraints, contact the [Algolia support](https://support.algolia.com/hc/en-us/requests/new) team.

## Avoid mixing attribute types

Maintaining a consistent type for attributes is an important step in preparing your index structure.

```json JSON icon=braces theme={"system"}
[{ "color": "red" }, { "color": ["red", "blue"] }, { "color": 250000000 }]
```

Using this index as is would lead to unexpected results.
This is because the attribute `color` can be **a string**, **an array**, or **an integer**.

A structure like this often indicates an underlying issue with your data.
Ensure a consistent type for your attributes.

```json JSON icon=braces theme={"system"}
[{ "color": "red" }, { "color": "green" }, { "color": "blue" }]
```

## Avoid mixing attributes from different domains

When preparing your index structure,
ensure that attributes are relevant to a single domain.

For example, an index for articles shouldn't contain attributes relevant to product information and vice versa.

Language is a common domain that could lead to a mix of attributes within your index.

```json JSON icon=braces theme={"system"}
[
  {
    "objectID": "MH001",
    "color": "red",
    "title": "Bag"
  },
  {
    "objectID": "MH002",
    "color": "rouge",
    "title": "Sac"
  },
  {
    "objectID": "MH001",
    "color": "blue",
    "title": "Chair"
  }
]
```

The second record is doesn't belong in this index because it has French language attributes.
Review the index to ensure attributes from different domains aren't mixed.

```json JSON icon=braces theme={"system"}
[
  {
    "objectID": "MH001",
    "color": "red",
    "title": "Bag"
  },
  {
    "objectID": "MH002",
    "color": "red",
    "title": "Bag"
  },
  {
    "objectID": "MH001",
    "color": "blue",
    "title": "Chair"
  }
]
```

## How Advanced Personalization validates attributes

Advanced Personalization prioritizes attributes that directly improve personalization,
building user profiles on meaningful data.
It also filters out attributes based on user interactions.

* **An attribute-value pair must show significant user interaction.**
  It doesn't set a fixed benchmark for checks but instead looks for a significant number of interactions.
  If Advanced Personalization randomly picks users from last month for the `color:red` attribute,
  it expects some users to have interacted with red products.

* **Discard attributes with minimal relevance.**
  If an attribute applies to too few products,
  it's often filtered out due to a lack of user interactions.

* **Discard attributes with excessive diversity.**
  For example, Advanced Personalization might filter out an attribute like `brand` if it detects thousands of distinct brands across millions of products,  yet no individual brand garners enough user interactions to be considered "important".

* **Discard unusual attribute values.**
  While attributes with many unique values aren't discarded,
  Advanced Personalization doesn't process the unusual values, for example, `color:pink_with_brown_dots`.