Skip to main content

Echo: Use AI to Identify Risk and Expansion Opportunities at Scale

Using AI to surface risk and expansion opportunities from customer conversations

Updated over a month ago

Written by: KJ Close

What is Echo?

Echo is Hook’s AI-powered sentiment intelligence layer. Echo ingests data across your emails, meetings and support tickets and leverages an LLM to surface key insights to your whole business. Pairing this with your Hook ML engagement level gives the most complete end-to-end view of customer health available.

Set up Echo for your organisation

There are four steps to set up Echo for your organisation:

  1. Set context about your organisation

  2. Decide what Echo should listen for

  3. Create automated actions based on triggers

  4. View this information across your book of business

Set context about your organisation

  1. Go to Org Configuration in Settings.

  2. Complete the About your business and Your competition fields in the AI Agent data section to provide Echo with the relevant context about your organisation and competition.

    Organisation Configuration settings with AI Agent data section

If your organisation has multiple products in Hook, you can also provide context about these products using the Agent data field in Products settings.

Complete product specific context for Hook's AI agent in Product settings

Enter product specific context for Hook's AI agent in Product settings

Decide what Echo should listen for

  1. Go to Automations (in Settings) and select Create automation.

    Create an automation

    Create a new automation

  2. Add a name and select Begin configuration.

  3. In the list of automation triggers, select Hook’s AI Agent detects something.

    List of trigger options for the automation

    Select Hook's AI Agent detects something as the automation trigger

  4. Add your instructions for Echo, Hook’s AI agent. Make sure to include:

    1. Detection category: The type of information you are looking to surface (e.g. risk, upsell opportunity).

    2. Source: The data source you would like to be analysed (e.g. meetings, emails, support tickets).

    3. Prompt: The prompt you would like the AI agent to leverage.

      Hook's AI agent form

      Enter the details for Hook's AI agent

  5. Echo will follow these instructions for all customers as default. Use Add filter to specify which customers Echo should use for insights.

Use Add filter to specify which customers Echo should use for insights

Reach out to your AM if you would like help with prompt engineering.

Create automated actions based on triggers

  1. After adding instructions for Echo, choose what happens next by clicking Next: Add action.

  2. Choose which action should be taken when the AI agent detects something:

    1. Create an alert.

    2. Add to a cadence or sequence.

    3. Create a risk.

    4. Populate a custom field.

      List of actions to take in this automation

      Choose an action to be taken

  3. Complete the relevant fields.

  4. Click Create to finish configuring the automation.

  5. To enable the automation for your organisation, make sure it is set to Active by toggling the switch in Automations.

List of active and inactive automations

Set you automation to Active

Use Echo to create sentiment based risks

  1. In Automations, follow these steps and select the Hook’s AI Agent detects something trigger and the Create a risk action.

  2. Complete the required fields:

    1. Risk: Choose the risk flag for the risk, e.g. High Risk. These are configured in settings.

    2. Reason: Choose the reason for the risk e.g. Adoption. These are configured in settings.

    3. Executive summary: A summary of why the risk is being raised.

    4. Pre-existing risk: Choose which action to take if Hook’s AI agent detects a risk for a customer that already has an open risk.

    5. Product: For organisations with multiple products, select which product this risk should be associated with.

    6. Set how often this risk re-runs if nothing’s changed:

    Choose Never if the risk should only be created once when the conditions are met. Choose After a certain amount of time if the risk should be created on a timely basis every time the conditions are met e.g. once a week. This time interval can be customised.

Create a risk in automations using Hook's AI agent trigger

Fill out the details to have a risk created when the automation is triggered

Let Hook’s AI agent choose the values for each of these fields by selecting the Hook AI agent variables. These will automatically populate the following fields:

  • Risk: Hook AI Agent flag

  • Reason: Hook AI Agent reason

  • Raised date: Hook AI Agent raised date

  • Executive summary: Hook AI agent summary

  • Product: Hook AI agent product

Hook AI Agent field names

View this information across your book of business

Information created in Hook by Echo can be seen in various parts of Hook.

For AI Agent risks, this is in the Customers table and account pages.

Hook AI Agent risk created on the account page

Hook's AI Agent created a risk on this account page

To see why the AI Agent surfaced certain information, click the ? icon next to any sentiment. You'll see the reasoning and data sources Echo used.

Agent rationale for creating the risk

Hook's AI Agent's reasoning for creating the risk on this account

❓ Frequently Asked Questions

How can I get access to Hook’s Echo?

Reach out to your AM today to find out more about Echo.

What data sources can you support?

Echo supports the following data sources:

  • Meetings

  • Emails

  • Support

Data can be used from a variety of platforms including Gong, Salesloft, Salesforce, Intercom and more. If you’re not sure, just ask your AM.

How often is data refreshed?

Hook analyses data in real time. Every new call, email or support ticket that is logged is analysed within 5 minutes. If a trigger is met (e.g. a risk is identified), it will immediately appear in the Hook UI.

What models does Echo use?

At Hook, we use Anthropic’s Claude.

Did this answer your question?