Blog

AI for Utility and Energy Retailers' Support Teams

August 2026 · 6 min read · Industry Guide

Notebook sketch of a gear with a terracotta centre, representing support operations
← Back to all posts

A challenger energy retailer running a lean support team, often 15 to 40 people covering tens of thousands of customer accounts, faces a support load that larger retailers absorb with sheer headcount. Billing queries, connection and disconnection requests, and hardship applications all carry specific obligations under the Australian Energy Regulator's rules, and getting a response wrong or slow is not just a customer service problem, it is a compliance one.

Where the support queue actually clogs

The bulk of inbound contact to a small retailer's support team is repetitive in structure even when it feels urgent to the customer: a bill that looks higher than expected, a request to move a connection date, a question about a concession or rebate eligibility. Each of these has a mostly standard answer that depends on pulling the right account data and applying it correctly, which is exactly the kind of task Claude handles well when connected to the retailer's billing and CRM systems.

What it does not do, and should not do, is make hardship or disconnection decisions. AER rules around hardship customers are specific and carry real regulatory risk if handled badly, including obligations around payment plans, disconnection warnings and the sequence of contact attempts before disconnection is permitted. Claude's role here is to draft the response and flag the account status correctly, surfacing every hardship indicator to a trained support officer, who makes the actual call under the retailer's hardship policy.

The workflows worth building first

  • Billing enquiry triage and drafted responses pulling live account and usage data

  • Connection and disconnection request processing against the correct notice periods

  • Concession and rebate eligibility checks drafted against current state and federal schemes

  • Hardship indicator flagging that routes straight to a trained officer rather than auto-resolving

A Victorian retailer with around 60,000 residential accounts and a 22-person support team was running average response times of just under 48 hours on email enquiries during winter billing peaks, close to breaching their own service commitments. After building a Claude-assisted triage and drafting layer across billing and connection queries, average response time dropped to under 8 hours, with the support team spending their time on the harder calls, including hardship conversations, instead of the routine billing questions that made up close to 60 percent of volume.

The cost and compliance case together

Outage communication is another area where speed and accuracy both matter and where a small retailer's team is most exposed during storm season. When a network outage hits a service area, the retailer needs to notify affected customers quickly and accurately, distinguish network faults from billing-side issues in the response, and keep a consistent message across email, SMS and any social channel the retailer runs, all while the phone lines are already overwhelmed. Claude can draft that outage communication set in minutes from the network operator's outage feed, freeing the support team to handle the direct calls from customers who need more than a standard notification.

Meter and billing dispute handling also benefits from having a consistent first-pass response. A customer disputing a bill after a meter reading correction, or querying a sudden usage spike that turns out to be a faulty appliance, needs a response that pulls the actual interval data and explains it clearly, not a generic template. Claude can draft that explanation directly from the usage data, which both resolves more disputes on the first contact and reduces the number of enquiries that escalate to a supervisor purely because the first response felt like a brush-off.

Getting the rollout order right matters for a small retailer's risk tolerance. Starting with billing enquiry drafting, the lowest-risk and highest-volume workflow, lets the compliance team see exactly how the drafting and review process behaves before extending it to connection requests and eventually the hardship-flagging layer, which needs the most careful review given the regulatory weight attached to it.

Setup for a retailer this size typically runs $15,000 to $28,000, largely driven by the integration work needed to connect securely to the billing platform and maintain a clean audit trail of what was drafted versus what a human sent. That audit trail matters as much as the time saving, because a retailer under AER scrutiny needs to demonstrate exactly how a customer enquiry was handled, and a system that can show a complete, timestamped record of draft, review and send is a stronger compliance position than a support team working from memory and personal notes, even before counting the roughly $180,000 a year in reduced overtime and after-hours contractor cover that retailer avoided in its first full year running the workflow.

Ready to move from AI pilot to production?

We help mid-market Australian businesses deploy AI automations that actually reach production and deliver measurable ROI.