In the realm of enterprise sales, where each rep actively handles dozens of calls per week (go team!), keeping track of every call and identifying those in need of attention can be a formidable task for revenue leaders.

Enter Call Signals – Call Signals allow you to filter and receive summaries only for calls that satisfy certain subjective outcomes determined by AI, such as customer sentiment, rep performance, win probability, and many more.

This is particularly useful if you want to reduce noise and be alerted (with a summary, based on the Call Signal) for only specific calls.

Call signals work along with any other conditions and prompts you may have set. They are an additional layer of intelligent filtering.

Use cases

Send the summary of a call to a Manager when the customer/prospect sentiment is negative. Allows for quick coaching, feedback, and immediate actions that can be taken to save a deal.
Notify a PMM on Slack (along with call context) when a competitor is mentioned in a deal. Allows for quick enablement and effective follow-up actions.
Acknowledge on a public channel or send feedback to her manager when a seller has done really well on a call.
Notify the manager when the win probability is lower than a % based on how a call went.

How it works

Set Call Signal conditions

Your Meeting Intelligence workflow has a section called Call Signals, with a list of pre-programmed prompt insights. Set these conditions to filter out calls that meet specific signal criteria.

Rattle understands the context of your call, including identifying the internal and external stakeholders, as well as the purpose (intro, demo, support issue).

To set up a condition:

Select the prompt insight > relational operator > response value/s. Depending on your selection, the response values may be in the form of a rating scale, percentage, or text.
Click the pen symbol to edit the signal, and the '+' symbol to create a new one.

In the example below, we've programmed the conditions to send summaries of calls where the Seller rating is less than 4, so that the revenue leader is alerted only when there is a potential risk to the account's health.

Create a new Call Signal

MI's Call Signals are structured with two values:

1. Prompt value

This refers to the question the AI will use to extract the required information from the call transcript. E.g. “How engaged was the customer during the call? Did they ask questions, provide feedback, etc.?".

Click the '+' symbol to create a new prompt, or the pen symbol to edit and customize an existing prompt. You can be as descriptive as you'd like here - feel free to specify sales personas, and frameworks for a precise outcome.

2. Response value

This refers to the format of the Call signal, you can choose between:

a rating scale - Add minimum and maximum values. eg: 1 - 10
percentage - Add minimum and maximum values. eg: 0 - 100%
custom text fields - Add text fields separated by a comma. eg: Likely, Unlikely, Neutral, Positive, Negative

Once you're done setting up your prompt, add a condition to filter calls basis the Call signal you've chosen.

In the example provided, we've chosen the 'prospect sentiment' as the prompt value with three potential response values: negative, neutral, or positive. In the conditions dropdowns, we've specified - Prospect sentiment > Less than > Negative. This configures the workflow to exclusively send call summaries requiring immediate attention.

The image below shows what the triggered alert will look like. This represents an actual call with live data, and as such, personal information has been redacted for privacy.

And that's it, that's everything you need to get started!