Understanding Automated signals
Signals are surfaced on an account to give insight into renewal and upsell likelihood. Signals can represent risk or opportunity and are triggered either manually or automatically.
Automatic signals can be generated from two different sources:
Echo AI
These are signals that our Echo agent has heard from across your customer interactions (meetings, support, email and Slack data). You configure the types of signals you want Echo to look for, give it some context and it does the hard work. Once it's identified a signal, it can review product usage information in Hook as well as information from external sources to get extra context and bring this signal to life. Follow the steps here to customise Echo signals for your organisation.
Follow the steps here to customise Echo signals for your organisation.
Automations
These are signals that you configure to get created automatically, based on data in Hook. You do this through Hook's automations feature, by saying when this specific set of criteria is hit, create a signal. These are what you would set up to have signals generated based on product usage, such as a risk signal created when adoption drops.
In order to create a signal using automations, first consider what your trigger might be:
A trend in customer data across the whole product or specific features e.g. increase in active users in the last 30 days, 50% increase in usage of settings page.
Hook automation trigger: A trend in customer data occurs
Usage patterns of specific user roles, e.g. 25% increase in dashboard feature usage by admins in the last 30 days.
Hook automation trigger: User filters are met
Decreases in customer engagement e.g. last meeting was more than 3 months ago.
Hook automation trigger: Customer filters are met
It's best practice to first consider what you would deem of significant importance to create a risk or an upsell signal, before you start creating them.
Only signals created by our Echo AI agent get associated Playbooks and actions right now.
Create a Signal using Automations
Navigate to your name in the top right-hand corner of Hook, then go to Organisation Settings.
On the left-hand side, select the Automations icon.
Select Create Automation and give it a name.
Choose whether this automated signal will be based on:
Customer Activity: based on customer-level activity, like account engagement, lifecycle events or total company adoption.
User Activity: based on user-level activity, behaviour or role-specific activity.
Select Begin Configuration.
Click the + in the If section and choose the trigger that should create the signal for you. Once you have completed the trigger fields, hit Create.
Click the + in the Then section and select Add a Signal.
Choose the Type and Category of that signal and the Owner, in the same way that you have instructed Echo AI to when it finds a signal.
You can add an Executive Summary and determine whether the signal is to be detected (accepted before being promoted to an active signal) or whether it immediately becomes active.
Lastly, choose whether you want this signal to re-run after a period of time and hit Create.
You have now created an automated signal.
Examples of Automated Signals
There are hundreds of possibilities of automated signals that could add a lot of value to a Customer Success or Account Management team. Here are some best-in-class examples:
Risk Signals
Automated signals created when:
Inactive users increases past a certain threshold
Active users drops below a certain threshold
Usage of key parts of your product drops by a percentage
The EB or Champion doesn't log in
Open support tickets increase above a certain threshold (can also be filtered by priority)
No customer interactions (emails, support tickets or meetings) in a given period
A customer hasn't used a product they've purchased within a given period
Upsell Signals
Automated signals created when:
A customer hits 80%+ utilisation, to signal upsell
A customer uses a part of their product not included in their package, as part of a fair usage policy
A customer goes into overages or soft limits
A customer meets a specific set of criteria that makes them a good fit for a new product or add-on upsell
❓ Frequently Asked Questions
What types of signals can be captured in automated signals which aren't flagged by Echo?
Any risk or upsell opportunity that can be identified from usage metrics requires an automated signal to be created. Echo signals are created based on customer interactions (unstructured) data.
How can I identify which signals are created by automations and which ones are created by Echo?
Echo signals can be identified by adding a filter on the Signals table for Source: Echo (AI Agent). Automated signals can be identified by adding a filter for Source: Automation. Signals on an account page will either state: Signal detected by Echo or Created by an automation.
Are automated signals always created as detected signals, or can they be active?
You can decide whether the signal is created as a detected signal (needs to be accepted) or an active signal. This can be configured in the Signal status: detected or active part of the Add a signal automation output.
Can I create an automated signal for accounts that already have a signal?
Accounts can have multiple signals (detected or active). The automation behaviour can be configured based on whether there are existing active signals on the account e.g. close the pre-existing active signal, add the automated signal as a detected signal etc. This can be configured in the Signal status: detected or active part of the Add a signal automation output.
I've created automated signals as a test but want to delete the ones I've created. How can I do this?
Reach out to [email protected] who can help with bulk deleting signals created from automations.
Does a signal created from an automation close automatically if it's no longer relevant based on metric changes?
It's not yet possible for these signals to close automatically. Only Echo is configured to automatically dismiss detected signals based on new context.