Automation of contact centres in Salesforce

Virtually everyone has encountered the basic version of contact centre automation as a customer. It consists in speeding up the processing of customer requests by obtaining basic information about the enquiry and then directing it to an operator who has the appropriate, relevant knowledge and access to resources to be able to resolve the request. In call centres, IVR (Interactive Voice Response) systems are used for this type of automation - namely automatic voice systems to which customers respond by pressing the appropriate buttons. In digital text communication, such enquiries can be directed through contact forms on a website, where the customer fills in the required fields (such as an order number, etc.) and selects the type of request (complaint, invoicing, delivery of goods, etc.). But Salesforce takes this process even futher. It also allows you to work with the analysis of the text in messages from customers and direct the request to a suitable operator based on keywords.

"With standard Salesforce tools, you can set up the paths customers go through during calls so that they are automatically directed to the right operator who already has the information and experience necessary to solve the problem. Salesforce can also analyse the content of text communication. Experience shows that the success rate of correctly assigning an enquiry based on text analysis is up to 80%," explains Josef Gorgan.

Although keyword-based automatic sorting of emails may seem relatively easy, experience has shown that too much freedom in defining keywords (or a large number of people who can define them) can cause confusion and delays in handling customer requests. This is why only a designated Salesforce administrator should add keywords after proper data analysis and approval for test environments.

However, IVR systems and contact forms have their limits. They only provide a limited range of sorting and client request routing, which cannot be extended indefinitely. If the IVR system on the call centre line contains too many dial steps, the customer will terminate the call, or he will try to bypass the automation and request the fastest possible connection with the operator. Customers will also not have the patience to choose from dozens of contact form fields.

This is why the deployment of chat and voice robots - chatbots and voicebots, is a clear trend in the automation of contact centres. But while Salesforce has integrated technology for text communication with customers, including its own artificial intelligence called Einstein, a voicebot must be deployed on the PBX side. The openness of this Salesforce platform is also an advantage, as it enables connection to voicebot solutions in PBXs from established suppliers, supported by various types of artificial intelligence solutions. However, the deployment of sophisticated chatbots or voicebots usually leads to the deployment of an MVP (minimum viable product), which makes basic functions available, subsequently extended based on customer feedback, end users and data analysis.

Systems driving voicebots and chatbots utilize Salesforce's integration with other corporate systems (such as ERP, warehousing, logistics, etc.) to collect data to respond to customer enquiries.

"Salesforce provides a variety of automation features and tools, many of which do not require programming skills for their deployment, they are relatively easy to set up in a visual environment. The Apex programming language opens up more space that makes the possibilities of automation almost unlimited," adds Josef Gorgan, Salesforce Practice Manager at Merkle.

An article by Michal Mravinac, Regional Sales Manager, Central and Eastern Europe (CEE) at Salesforce.