How Consilium Energy increased profit margins by 700% with an AI-powered chat agent

The challenges

Consilium Energy had a problem that many growing businesses would envy - they had too many leads. But in the competitive consumer energy market, having leads isn't enough. You need to quickly identify which ones are worth pursuing, and that's where Consilium was struggling.

The challenges were stacking up:

• Manual qualification bottleneck: Sales agents were spending countless hours reviewing potential customer inquiries, assessing energy needs, and determining which leads deserved follow-up. It was exhausting and slow.

• Hiring wasn't the answer: To handle growing inquiry volumes, they would need to hire additional sales staff. But the maths didn't work—the cost of new agents was eating into profits faster than they could close deals.

• Time-to-response problem: In consumer busienss, speed matters. Every hour a qualified lead waited for assessment was an hour they could be talking to competitors. Consilium's manual process meant they were losing hot leads to faster-moving competitors.

• Inconsistent qualification: Different agents had different approaches to qualifying leads. This created unpredictable sales funnel quality and made it difficult to optimise their sales process.

• No scalability: The manual approach created a hard ceiling on growth. More leads meant more agents, which meant more costs, which meant thinner margins. The business model couldn't scale without fundamentally changing the economics.

Consilium needed a way to handle higher volumes while maintaining, or even improving qualification quality, all without the crippling cost of endless hiring.

Our approach

We partnered with Consilium Energy to build an AI-powered solution that would transform their entire lead qualification process:

Phase 1: Understanding their sales intelligence

We didn't just study what questions their agents asked—we learned why they asked them. What signals indicated a high-quality lead? What information was essential versus nice-to-have? We distilled years of sales expertise into a framework that AI could apply consistently.

Phase 2: Building the conversational engine

We developed a custom chatbot powered by Claude (Anthropic's AI) that could engage potential leads through WhatsApp where their customers already were. We integrated into Meta WhatsApp using their webhook architecture and then allowed the chat agent to proceed with lead qualification before it would then pass to the sales agent leads that fit a particular criteria.

Phase 3: Intelligent lead scoring

We created a sophisticated lead qualification system that could assess customer responses and automatically score leads based on Consilium's criteria. The AI agent could identify high-value prospects, flag concerns, and route qualified leads to the appropriate sales team members with complete context.

Phase 4: Seamless integration

We integrated the chatbot with Consilium's existing CRM and technology stack. Qualified leads flowed automatically into their sales pipeline with all the information their agents needed to close deals quickly. We implemented security protocols to protect sensitive customer data throughout the automated workflows.