> For the complete documentation index, see [llms.txt](https://ask.birdie.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ask.birdie.ai/core-concepts-and-entities/feedbacks.md).

# Feedbacks

## What Are Feedbacks?

A feedback is any piece of customer input that Birdie ingests and analyzes, such as an app review, an NPS response or a support ticket. Feedbacks enable many of Birdie's other features or entities such as areas, opportunities, dashboards, criteria, etc.

Depending on how the interaction happened, a feedback takes one of two forms:

* **Feedback**: A one-off review published by a user about a service, product, or transaction. E.g. App Store reviews, NPS, CSAT and CES responses.
* **Conversation**: An exchange of messages between a customer and your company, grouped as one record. E.g. Support tickets, complaints, social media threads, issues.

In the [ingestion schemas](https://ask.birdie.ai/core-concepts-and-entities/pages/5Z8b5BkvkdE2oYhrsKO8#id-2.-understanding-data-types), Feedbacks and Conversations are separate data types. Within the platform, both are feedbacks: a support ticket with 20 messages is counted, filtered and analyzed as one feedback, just like an NPS response.

Conversations can also be sent as audio recordings. Birdie transcribes the audio into a conversation before processing it.

## Why Conversations Are a Special Type of Feedback

A conversation is still one feedback, but its content is split into messages. Each message has its own text, posting date and author, and identifies whether the author is a customer, an agent or a bot.

Knowing who said what, and when, lets Birdie offer extra analysis for conversations:

* Conversation summary: Birdie summarizes the whole conversation using the order of its messages and information about each author. This summary is what lets the conversation be analyzed as a single feedback.
* [Conversation Features](/customer-intelligence/exploration-and-analysis/using-filters-in-birdie/conversation-features-fields.md): timing and volume metrics calculated from the messages, such as first response time, number of turns or total duration.
* [Frontline Intelligence](/frontline-intelligence/product-overview.md): evaluation of the service quality of your agents, based on the messages they wrote.
* [Ticket satisfaction](#ticket-satisfaction-survey): a survey answered at the end of a conversation (e.g., a CSAT survey after a ticket is closed) enables the [TCSAT and TNPS metrics](/getting-started/metrics-glossary.md#tickets).

## How Birdie Treats Feedbacks

Once ingested, feedbacks and conversations go through the same pipeline and are available in the same places:

* Enrichment: PII is anonymized and the text is classified with [Signal, Sentiment and Intentions](/core-concepts-and-entities/signal-sentiments-and-intentions.md). For conversations, the text comes from their messages.
* Classification: feedbacks are grouped and classified into [areas](/core-concepts-and-entities/areas.md), [opportunities](/core-concepts-and-entities/understanding-and-creating-opportunities-in-birdie.md), [criteria](/frontline-intelligence/criteria.md), etc.
* Analysis: feedbacks are available in [Exploration](/customer-intelligence/exploration-and-analysis.md), [filters](/customer-intelligence/exploration-and-analysis/using-filters-in-birdie.md), [Segments](/core-concepts-and-entities/segments.md), [Dashboards](/customer-intelligence/dashboards-and-reporting/custom-dashboards.md) and [Skye](/core-concepts-and-entities/understanding-and-using-skye-in-birdie.md), and are counted by the same [metrics](/getting-started/metrics-glossary.md). A conversation counts as one feedback, regardless of its number of messages.
* Export: Feedbacks can be exported in the same `feedbacks` file, and conversation messages in a separate `messages` file. See [Data Schema Definitions for Birdie Export](/integrations-and-data-ingestion/data-schema-definitions-for-birdie-export.md).

## What a Feedback Contains

Every feedback, regardless of its form, is described by a small set of common fields:

* An identifier, unique for each feedback (and for each message of a conversation). Sending a feedback with an existing identifier updates it.
* The posting date, used by every time-based chart, filter and metric.
* The text, either the comment or the messages of a conversation. Without text, Birdie can't detect Signal, Sentiment, Intentions or Opportunities.
* A source, a label that groups feedbacks by origin or purpose (e.g., "Support Ticket" or "NPS").

{% hint style="info" %}
Text is theoretically optional when you have ratings (e.g NPS surveys with no comment), but not having text really limits the analytical possibilities for the feedback.
{% endhint %}

Other fields are complementary and optional, e.g:

* A rating, when there is one (e.g., a star rating or an NPS score). Without ratings, rating based metrics for the feedback aren't available.
* An optional account, linking the feedback to an [Account](/core-concepts-and-entities/accounts.md).
* [Custom fields](https://ask.birdie.ai/core-concepts-and-entities/pages/5Z8b5BkvkdE2oYhrsKO8#id-4.-custom-fields) for attributes specific to your organization (e.g., plan type or region).

{% hint style="info" %}
Birdie supports a lot of fields, that each help power different functionalities. For an exhaustive list, please check the [ingestion schemas](https://ask.birdie.ai/core-concepts-and-entities/pages/5Z8b5BkvkdE2oYhrsKO8#id-3.-data-schemas) documentation.
{% endhint %}

## Sending Feedbacks to Birdie

Feedbacks can be sent through any of Birdie's [integration paths](/integrations-and-data-ingestion/birdies-integration-paths.md): [native connectors](/integrations-and-data-ingestion/how-to-integrate-with.md) such as Zendesk or Qualtrics, data lake and data warehouse connectors, or the [Ingestion API](/integrations-and-data-ingestion/ingestion-api.md).

See [Data Schema Definitions for Birdie Ingestion](https://ask.birdie.ai/core-concepts-and-entities/pages/5Z8b5BkvkdE2oYhrsKO8#id-3.-data-schemas) for the full definition of each type of feedback and conversation.

## More details on optional functionalities

Birdie contains many optional functionalities for feedbacks:

### Source

This field is sometimes requested with data uploads to allow the client to decide how they want to segment their data. When a custom source (also some times called a source alias) is provided, Birdie will store the data with a separate data type. With a customized source, we can configure many functionality behaviours specifically for that source, including:

* Editable user facing label within the application.
* Weight / relevance for [impact score](/admin-and-settings/organization.md#impact-score-weights-and-calculation).
* Custom ordering in visualizations / lists.
* Primary and secondary metric displayed in the [explore pages header](/customer-intelligence/exploration-and-analysis/indicators-in-the-header.md).
* Weight / relevance of transactional survey responses.
* [Cost per ticket](/admin-and-settings/organization.md#potential-saving-metric-premise).

### Ticket satisfaction survey

Support ticket conversations can include the answer to a survey sent at the end of the ticket, such as a CSAT survey after the ticket is closed. The answer is sent as optional fields on one of the conversation's messages, so the score stays linked to the ticket it rates. It enables the [TCSAT and TNPS metrics](/getting-started/metrics-glossary.md#tickets). Each ticket should have a single survey response: only one of its messages should carry a rating and survey type, and that type should be either CSAT or NPS, not both. Otherwise, the survey is counted more than once and the TCSAT and TNPS scores are skewed.

To send it, see the survey fields in the [support ticket schema](/integrations-and-data-ingestion/data-schema-definitions-for-birdie-ingestion.md#conversations-support-tickets).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://ask.birdie.ai/core-concepts-and-entities/feedbacks.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.