Industry overview
What is shaping electric power
Power utilities serve millions of households and businesses, and are judged on reliability, billing accuracy and how well they communicate during outages. Smart meters are changing the relationship.
- Smart meters are changing billing and payment behaviour
- Outage communication strongly affects satisfaction
- Digital payments are replacing counter payments
Client background
A power distribution utility, working in Malaysia. Client identity is withheld to protect confidentiality.
- Organisation
- A power distribution utility
- Sector
- Electric power
- Team engaged
- IT & Development
- Services
- AI & automation solutions, WhatsApp marketing, API development & integration
Client challenge
Call centres were overwhelmed by routine questions about bills, payments and outages, especially during storms. Customers often waited a long time on hold just to check their bill amount or report an outage.
Research objectives
- Deliver on the brief: a WhatsApp assistant for electricity bill queries.
- Define requirements and success criteria the solution would be tested against.
- Deliver a reliable, secure solution the team can run and extend.
Target audience
Scope: Bill, payment and outage queries.
Audiences we typically reach in electric power
Residential consumersCommercial and industrial consumersSmart-meter usersField and call-centre staffRegulators
Geographic coverage
Coverage: Malaysia. Market: Malaysia.
- Malaysia
Research methodology
An iterative delivery approach: discovery, design, build and test in short cycles, with the client reviewing each step.
- 1Brief
- 2Contact reason analysis
- 3WhatsApp assistant
- 4Insight
We analysed contact reasons and built a WhatsApp assistant in Malay and English connected to billing and outage systems, handing complex cases to agents. The assistant was trained on approved answers, tested with customers before launch and designed to hand over to agents with the full conversation history.
AI & automation solutionsWhatsApp marketingAPI development & integration
Sampling strategy
- Users and stakeholders from every affected team took part in discovery workshops and acceptance testing.
Data collection
- Requirements were gathered through workshops, process walk-throughs and a review of existing systems and data.
Data quality & validation
- Peer code review on every change
- Automated and manual testing
- User-acceptance testing with the client
- Release checklist and rollback plan
Data analysis
- Usage and performance monitoring after launch
- Defect and support-ticket trends
- Process time before and after
Key findings
- Routine queries moved to self-service, waiting times fell and customers could check outage status instantly.
- Agents spent more time on complex cases such as billing disputes and new connections.
Connecting the assistant to live systems made it genuinely useful.
Recommendations
- Build on the core insight: connecting the assistant to live systems made it genuinely useful.
- Track adoption and support tickets, and plan the next release from real usage.
Business impact
- During storms, the assistant handled outage queries at scale, reducing pressure on call centres when it was highest.
Impact is described qualitatively.
Deliverables & next steps
- Deployed solution with documentation and handover
- Support and maintenance plan
- Repeat the measure after changes are in place to track progress.
- A natural follow-on question: which payment channels do consumers prefer?
Client identity is withheld to protect confidentiality.