You can use proactive customer contact if you see certain behavior from your customer that you would like to anticipate, among other things to prevent the customer from having to contact your organization. When you contact a customer based on expectations that emerge from data analyses, you speak of predictive customer contact. In this edition of Ziptone's TechUpdate we take you through the possibilities of predictive and proactive customer contact.

tech update is the Ziptone section on emerging technology relevant to customer engagement. We briefly explain what it is, how it works, why it is relevant, what the pitfalls are and we ask the reader's opinion: Hot or Not?

What is it

Proactive customer contact can be based on a prediction, but can also be based on business rules: if certain conditions are met, we contact the customer.

Predictive customer service is customer service that anticipates a certain outcome that you want to do something with. That outcome can be certain customer behavior, but also predicted events with customers or signals from systems. Anticipating is done on the basis of recognizing patterns and predicting based on data.

Examples of predictive customer contact

• intervene on shopping cart abandonment: prevent a customer from abandoning a filled shopping cart without paying or from partially emptying the shopping cart;
• intervene in search behavior: for example, when someone lingers too long on certain input fields of an online form;
• make an appointment for a technician to visit because a device is malfunctioning and may become defective over time.

Examples of proactive customer contact

• as a health insurer, approach your customers before the switching/renewal period begins;
• contact a credit card holder regarding possible payment problems or suspicion of fraud;
• contact an addressee because a package is delivered later than agreed.

How it works

Predictive customer service can be fueled by different types of events. By evaluating historical data and behavioral patterns, AI can anticipate customer needs and preferences to deliver a fast and personalized experience. This requires several steps.

• collect data (online tracking data, information from customer conversations, CRM data);
• analyze data (using machine learning/algorithms and previous customer journeys that are already known) to make predictions about customer needs based on patterns;
• define actions: criteria for which action, at what time/in what situation via which channel should be taken; think of customer value, priority/urgency, costs, etc.;
• using data to initiate actions;
• perform actions (e.g. call, email, chat or SMS) and follow up.

Advanced CCaaS solutions have the necessary functionality for predictive engagement on board. As you want to further personalize predictive customer contact, you can also use data from other sources, such as CRM systems and connected devices.

Why is it relevant

With proactive customer service you can anticipate potential questions and respond to specific customer needs at an early stage or prevent future problems. Predictive customer service can both improve service provision through good timing and personalization, and directly contribute to commercial results, for example through higher customer appreciation, higher customer value (through additional spending).

Also read: Proactive customer contact, who wouldn't want that? – Preventing problems instead of solving them – or from the perspective of the contact center manager: avoiding customer contact instead of resolving it. This sounds attractive and has great advantages, but it should not be confused with classic programs aimed at call reduction.

 

Predictive customer service and proactive customer contact can also reduce the pressure on customer service. On the one hand by using off-peak hours for outgoing contact and on the other hand by preventing incoming customer contact.

Technology

More and more CRM and customer contact platforms are equipped with functionality in the area of ​​proactive and predictive customer contact. Think of Zendesk Relay (in partnership with Meta), Genesys predictive engagement, Twilio CustomerAI Predictions and NICE Enlighten AI. Also CRM applications like Salesforce (Einstein Analytics) and Pega (Next Best Action) have functionalities on board, primarily aimed at defining actions. You can use different channels for proactive customer contact: from an outbound phone call to a push message or text message or from an email notification to a pop-up for a chatbot or live chat. WhatsApp is a popular inbound channel, but outbound communication is more difficult to be arranged. SMS can be used without any problems. Direct messages (DMs) via social media channels such as X are used less and less.

Just pay attention

1. Predictive and proactive customer contact is not something you can simply unleash on your entire customer base. Because you actively approach your customer, there is a risk of irritation. With shopping cart abandonment, for example, you will have to focus on customers for whom the chance is greatest that you will intervene effectively in a process. Customers who you can predict will almost certainly (or almost certainly not) pay offer a lower chance of success.
2. Predictive is often promoted as a possibility to achieve less contact, a faster solution or even no contact at all. The question is whether this is realistic. More interesting is the question how you can use predictive and proactive customer contact to strengthen the relationship with your customer.
3. When using proactive customer contact, you should take into account that this can lead to additional costs. For example, when you initially approach a customer via automated e-mail or chatbot, after which the customer would like to be helped further via a live channel. The costs must be in balance with the intended commercial result, even when live contact arises from an automated predictive message.
4. There are rules for contacting customers in specific situations. These rules relate to opt-in for customer contact (consent) but also to privacy. These rules are relevant for well-intentioned actions such as service calls, happy calls and welcome calls, but also for actions that affect privacy, such as shopping cart abandonment or warnings about limits for calling bundles and bank balances. Therefore, involve your CISO or privacy officer in the design and implementation of proactive and predictive customer contact.

(Ziptone/editors)

Thanks to Kim de Boer, Front Line Solutions

Featured, Knowledge base, Knowledge partners, Technology, tech update Front Line Solutions