Sentiment analysis

Analysing and scoring your inbound messages

ThinkAutomation comes with an inbuilt sentiment analysis engine. It helps you extract meaning from millions of messages, and quickly identify issues or opportunities.

1. Parse inbound message

ThinkAutomation runs messages through a sentiment analyser

2. Calculate sentiment score

It examines keywords, tone and syntactic effect to produce a sentiment score

3. Run rule-based actions

Based on this score, ThinkAutomation performs the relevant follow-up action

Route customer emails based on sentiment
Prioritise and escalate support tickets
Alert sales of an upsell opportunity
Push great feedback to a web feed
Bridge triggered calls for critical issues
Crawl big data to generate a report
Forward strongly negative complaints to a manager
Segment customers based on satisfaction
Send auto-responses to emotive tweets
Run competitor analysis and market research
Evaluate the customer response to a campaign
Track customer sentiment over time
Send a case study request to happy customers
Compile a “heat map” of common customer phrases
Alert managers of non-compliant emails
Pull a list of indifferent/passive customers
Run HR reports on employee engagement
Determine key promoters and detractors
Send special offer to loyal customers
Track trends in complaints and queries
Email positive feedback to marketing
Analyse online reactions to a product
Track how a change in service affects sentiment
Compile a spreadsheet of customer comments
Aid decision-making for financial trade and investments
Shortlist CVs based on keywords
Identify top-performing employees for customer care
Conduct ongoing brand analysis and monitoring
Make connect calls between sales and hot web leads
And countless more

Hundreds of combinations

Sentiment analysis could end with the production of a sentiment score. Or, you can act on those sentiment scores and incorporate them into any number of larger automated workflows.

You design the workflows that work for you using “if” rules. (If sentiment score equals this value, then run that action.) For example:

If a support ticket contains strongly positive feedback

Parse comment into a company intranet page

Send email requesting permission to share on your site

If a tweet mentioning %company% has low sentiment score

Parse the tweet into an email

Send to customer services department

If email scores highly for up-selling potential

Forward to the sales team

Add a CRM note against the customer record

Ready to get more value from your data?

Start monitoring and measuring customer sentiment with a free trial of ThinkAutomation.

GET STARTED