If your brand is providing customer support on Twitter, your handle probably gets a lot of mentions from people who are looking to reach out to you. Some of these tweets might be casual in nature with the user not expecting an instant reply, while others might be more important issues that will need immediate action from your side. Therefore, it becomes essential to quickly find the most important tweets that require action amongst all the tweets directed at you.
With the Social Signals feature in Freshdesk, we use Artificial Intelligence to help you automatically find the relevant tweets quickly and cancel out the noise in your Twitter account.
Setting up Social Signals inside Freshdesk
Login to your support portal as an administrator.
Add your Twitter account to Freshdesk.
Once it is added, go to Admin > Channels > Twitter.
From the list of added handles, choose the one you want to associate Social Signals with and click Edit.
Enable the ‘Start converting tweets into tickets’ checkbox.
Once the above steps are completed, you will get to choose between two options on how you want Social Signals to work.
2 ways of configuring Social Signals
Using only Social Signals
If you choose the option that says ‘Convert relevant @mentions into tickets’, Social Signals will intelligently convert only your relevant mentions on Twitter into tickets in Freshdesk.
You can also choose to assign these tickets to a particular group.
Combining Social Signals with keyword based filters
Companies who use specific business keywords can choose the option that says ‘Convert tweets to tickets using filters’ and enable Social Signals here.
This will combine functionality from the keyword based filters and Social Signals to fetch a set of relevant tweets containing the specified keywords.
It is important to note that when you are using a combination of these two filters, the keyword based filter will be given precedence over Social Signals.
Note: Please note that Social Signals will work only on tweets in the English language.
How does Social Signals learn and adapt with feedback?
The Social Signals feature is powered by a machine learning algorithm and can learn to automatically distinguish between a casual mention and a serious attempt to engage in a conversation on Twitter.
The machine learning algorithm of Social Signals will also be constantly learning from the feedback provided to its predictions. It looks into the brand’s activity on hundreds of tickets and adapts to suit their style better. Here’s an example of how this works.
Let’s say Social Signals converts a tweet into a ticket. You realize that it was a spam mention and you mark the ticket as Spam. Social Signals will learn to not convert such tweets into tickets in the future.
Assume that Social Signals does not convert a tweet into a ticket. If it sees an agent manually clicking on Convert to ticket on this tweet from the Social tab, it will learn that these kinds of tweets need to be automatically converted into tickets.