> For the complete documentation index, see [llms.txt](https://docs.dapta.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.dapta.ai/ai-text-agents/use-cases.md).

# What people use Text Agents for

Six things teams actually use Text Agents for on WhatsApp, what each one needs, and which pages to follow to build it.

Every agent starts from a **purpose** you pick when you create it. This page takes the six most common ones and shows what a working setup looks like: what the agent does, what you have to configure, and what to watch once it is live.

Pick the one closest to your business and build that first. You can always add the rest later — one agent can do several of these.

{% hint style="info" %}
**About the cost estimates.** Each case below carries a worked monthly example so you can picture the order of magnitude. They all assume a **direct Meta line sending to Colombia**, **40 credits per agent reply**, **10 credits** for a reply your team types in Inbox, and **1,000 USD = 1,000,000 credits** ($1 = 1,000 credits). Meta's first **1,000 service messages per number each month are free**, and that is already taken off.

They exclude your Dapta subscription, phone-number rental and any setup charges. Rates differ by country — see [How credits work](/my-account/how-credits-work-in-dapta.md) and [WhatsApp billing](/ai-text-agents/channels/whatsapp-billing-and-meta-charges.md) for yours.
{% endhint %}

***

## 1. Sell on WhatsApp

**For:** ecommerce and retail teams whose customers ask "do you have this in blue?" before they buy.

Your customer messages the store number. The agent answers from your real catalogue — sizes, stock, prices, shipping times — recommends alternatives when something is out, and passes the ones who are ready to buy to a person.

**What you set up**

1. Purpose **Ecommerces Sales Agent**, with your tone in the instructions.
2. A [Brain](/ai-text-agents/setting-up-your-text-agent/brains-for-text-agents.md) holding your catalogue or product sheets.
3. An [Action](/ai-text-agents/setting-up-your-text-agent/text-agent-actions.md) that checks stock or creates the order in your system.
4. A [human transfer](/ai-text-agents/inbox.md) for anything about payment or a complaint.

**What it might cost.** A store taking **800 conversations a month**, about 8 agent replies each:

|                                                            |            |
| ---------------------------------------------------------- | ---------- |
| 6,400 agent replies × 40 credits                           | $256.00    |
| Meta delivery: 5,400 chargeable service messages × $0.0008 | $4.32      |
| **Monthly total**                                          | **≈ $260** |

**Watch:** whether the agent is inventing availability. If it is, the Brain is missing the data — add it there, not in the prompt.

***

## 2. Qualify the leads your ads bring in

**For:** marketing teams running click-to-WhatsApp ads, where most of the budget is wasted on people who were never going to buy.

Every click lands as a WhatsApp conversation. The agent asks the three or four questions your sales team would ask, writes the answers into your CRM, and only books time with a human for the leads that pass.

**What you set up**

1. Purpose **Qualify Inbound Leads**, with your qualifying questions in the instructions.
2. [Post-conversation analysis](/ai-text-agents/setting-up-your-text-agent/post-conversation-analysis.md) with a variable per thing you need to know — budget, timeline, company size, product.
3. [CRM Pipeline Sync](/ai-text-agents/setting-up-your-text-agent/crm-pipeline-sync.md) so a qualified lead lands in the right HubSpot or GoHighLevel stage on its own.
4. The [Meta Ad Conversions](/integrations/meta-ads-conversions.md) credential, so Meta learns which conversations became customers and optimises your ads on real outcomes.

**What it might cost.** **500 ad conversations a month**, about 6 agent replies each:

|                                                            |            |
| ---------------------------------------------------------- | ---------- |
| 3,000 agent replies × 40 credits                           | $120.00    |
| Meta delivery: 2,000 chargeable service messages × $0.0008 | $1.60      |
| **Monthly total**                                          | **≈ $122** |

Conversations opened by a Click-to-WhatsApp ad carry a 72-hour free delivery window, so in practice your delivery cost here is usually lower than shown.

**Watch:** your cost per *qualified* lead, not per conversation. That is the number this setup moves.

***

## 3. Handle support without a night shift

**For:** any team whose customers ask the same forty questions, at every hour.

The agent answers from your help material, and the moment a question is genuinely new — or the customer is upset — it hands the conversation to a person in [Inbox](/ai-text-agents/inbox.md), with the whole history already there.

**What you set up**

1. Purpose **Customer Services And Support**.
2. [Brains](/ai-text-agents/setting-up-your-text-agent/brains-for-text-agents.md) built from your help centre, policies and FAQ.
3. A [Transfer to a human action](/ai-text-agents/setting-up-your-text-agent/text-agent-actions.md) pointing at the team that should pick those chats up.
4. [Post-conversation analysis](/ai-text-agents/setting-up-your-text-agent/post-conversation-analysis.md) to record the reason for each contact.

**What it might cost.** **1,200 conversations a month**, about 5 agent replies each, with 15% handed to a person:

|                                                            |            |
| ---------------------------------------------------------- | ---------- |
| 6,000 agent replies × 40 credits                           | $240.00    |
| 900 replies typed by your team × 10 credits                | $9.00      |
| Meta delivery: 5,900 chargeable service messages × $0.0008 | $4.72      |
| **Monthly total**                                          | **≈ $254** |

**Watch:** the reasons for escalation. The ones that repeat are missing pages in your Brain, and each one you add is a shift your team gets back.

***

## 4. Fill the calendar

**For:** clinics, services and sales teams that lose bookings to phone tag.

The agent proposes times, books the appointment, and sends the reminder the day before — all inside the same WhatsApp thread the customer already opened.

**What you set up**

1. Purpose **Schedule Meetings**, or **Confirmations** and **Reminders** for an agent that only confirms.
2. An [Action](/ai-text-agents/setting-up-your-text-agent/text-agent-actions.md) against your scheduling system.
3. A reminder [Sequence](/ai-text-agents/campaigns.md) using a **Utility** [template](/ai-text-agents/channels/template-types.md) — reminders go out after the 24-hour window has closed, so they must be a template.

**What it might cost.** **400 bookings a month**, about 6 agent replies each, plus one reminder per booking:

|                                                            |           |
| ---------------------------------------------------------- | --------- |
| 2,400 agent replies × 40 credits                           | $96.00    |
| 400 utility reminder templates × $0.0008                   | $0.32     |
| Meta delivery: 1,400 chargeable service messages × $0.0008 | $1.12     |
| **Monthly total**                                          | **≈ $97** |

**Watch:** no-show rate before and after. It is the clearest number this one produces.

***

## 5. Bring quiet contacts back

**For:** teams sitting on a list — old leads, lapsed customers, people who abandoned a cart.

A [Sequence](/ai-text-agents/campaigns.md) sends an approved opening message to your list on a schedule you set. Anyone who replies stops being a campaign and becomes a conversation: the agent picks it up and takes it as far as it can.

**What you set up**

1. A **Marketing** [template](/ai-text-agents/channels/template-types.md), approved before you schedule anything.
2. [Imported contacts](/ai-text-agents/campaigns/importing-contacts.md), cleaned, with consent to be messaged.
3. [Sending rules and limits](/ai-text-agents/campaigns/message-logic-and-limits.md) — pacing, retries, suppression.
4. Optionally, a [call fallback](/ai-text-agents/campaigns/call-fallback.md) so a voice agent phones the ones who never read the message.

**What it might cost.** A **5,000-contact re-engagement campaign** with a 12% reply rate, about 5 agent replies per reply:

|                                                            |            |
| ---------------------------------------------------------- | ---------- |
| 5,000 marketing templates × $0.0125                        | $62.50     |
| 3,000 agent replies × 40 credits                           | $120.00    |
| Meta delivery: 2,000 chargeable service messages × $0.0008 | $1.60      |
| **Monthly total**                                          | **≈ $184** |

This is the case where the connection route matters most. On a legacy line those same 5,000 marketing templates are charged at 26 credits each — **$130 instead of $62.50**.

**Watch:** your quality rating in the connection details. Sending marketing to people who did not ask for it is the fastest way to get a number restricted. Start small and grow the list.

***

## 6. Answer on your website too

**For:** teams who want one agent, not one per channel.

The same agent that answers on WhatsApp answers on your site through the [widget](/ai-text-agents/channels/widget-embed-your-text-agent-on-a-website.md). Both land in the same [Inbox](/ai-text-agents/inbox.md), and the **Source** filter tells you where each conversation came from.

**What you set up**

1. The widget, embedded on the pages where people hesitate — pricing, checkout, contact.
2. The same Brains and actions the WhatsApp agent uses.

**What it might cost.** **600 website conversations a month**, about 5 agent replies each:

|                                  |            |
| -------------------------------- | ---------- |
| 3,000 agent replies × 40 credits | $120.00    |
| Meta delivery                    | $0.00      |
| **Monthly total**                | **≈ $120** |

Widget conversations do not travel over WhatsApp, so Meta charges nothing for them.

**Watch:** which source converts better. It usually decides where your next effort goes.

***

***

## Work out your own number

Three inputs decide almost all of it:

1. **How many conversations a month**, and roughly **how many times the agent replies** in each one. Multiply them: that is your agent replies.
2. **How many messages you start yourself** — campaign sends, reminders, codes. Those are [templates](/ai-text-agents/channels/template-types.md), priced by category and country.
3. **Who delivers them.** On a direct line Meta bills delivery to your Meta account; on a legacy line it is all Dapta credits.

```
agent replies × credits per reply          → your Dapta agent cost
+ templates you send × your country rate   → Meta template cost
+ (service messages − 1,000) × your rate   → Meta delivery cost
```

The number people usually get wrong is replies per conversation. Run twenty real conversations, count the replies, and use that instead of a guess.

## Before you build any of them

Whatever you picked, the order is the same: [create the agent](/ai-text-agents/create-your-first-text-agent-in-dapta.md), test it in the [Playground](/ai-text-agents/setting-up-your-text-agent/playground-testing-your-text-ai-agents.md) until it answers the way you would, then [connect WhatsApp](/ai-text-agents/channels/connect-your-agent-with-whatsapp.md). Testing costs you nothing, and it is much cheaper than letting a customer find the gap.

{% content-ref url="/pages/TXgOhDoCH3jfppJaD1K2" %}
[Dapta and WhatsApp Business](/ai-text-agents/dapta-and-whatsapp.md)
{% endcontent-ref %}


---

# 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://docs.dapta.ai/ai-text-agents/use-cases.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.