Create dynamic customer support while protecting sensitive data

Requirements

If you use a chat like Intercom for your SaaS product or e-commerce site with a customer support team, you may encounter the following problems:

• Too many unqualified incoming requests if chat is present throughout the site for all users.

• Your chat may trigger unwanted calls from users who don't need immediate assistance.

• Default support choices for automatically responding to visitors can be difficult to navigate.

• Your chat may not be adapted to your interlocutor's location and language.

Put yourself in the customer's shoes. Just like in a store, you don't want to be mobbed by a sales assistant as soon as you arrive. You also don't want to have to search for hours if you have any questions. However, support is important to reassure you during the purchase (as in an Apple Store, where an advisor will guide you right up to the point of cashing up).

Objectives

Set up precise scenarios that trigger customer service intervention via Intercom. You'll be able to offer contextual assistance, respond to the specific needs of each user and avoid irrelevant requests. Imagine being able to guide your customers through their purchasing journey, like a dedicated advisor in a store, while giving them the autonomy they need to explore your products or services.

Proposed solution

A turnkey solution for personalizing your Intercom chat messages. By defining scenarios tailored to your needs, you'll be able to automatically trigger context-sensitive customer service assistance, based on user actions and profiles. To do this, we'll use variables and events collected via the data layer and transmitted to Intercom via Google Tag Manager.

Steps

• Define scenarios and customize chat messages to be triggered.

• Use or update data collection in the data layer.

• Use Google Tag Manager to pass variables and events to Intercom, which will return the correct configuration to trigger a chat.

• Intercom settings to manage scenarios

Implementation

Examples of customer support scenarios

‍

💡 Avoid exposing your product margins in the dataLayer with Cloud Firestore
‍Grâceaux variables asynchrones récemment introduites dans GTM server-side, il est maintenant possible d'utiliser Cloud Firestore pour enrichir les données côté server sans exposer certaines données côté client dans le dataLayer.

For example, simply collect item_id to a product page view. Then you can do a lookup with your data in Cloud Firestore's collections to return certain attributes such as margin which in this case is sensitive data.

Architecture implemented

Project duration

The project can take from a few days to a few weeks to implement, depending on customer development cycles and the number and types of scenarios.

‍