Solar Lead Qualification AI: 68% Reduction in Rep Time on Dead Leads
A UK MCS-certified installer was drowning in unqualified leads from Google Ads. Deploying AI qualification (ownership, bill, EPC rating, roof type) filtered 71% of leads before a rep ever saw them, freeing sales for high-intent conversations that closed 2.4x more often.
What was broken
- 70%+ of Google Ads leads were renters, low-bill households or unsuitable roofs
- Reps spent 5–6 hours a day dialling leads that had zero chance of closing
- MCS + SEG paperwork requires accurate property + ownership data — often missing at lead stage
- Close rate on 'qualified' leads was only 12% because the qualification bar was too low
What we built
AI qualification bot on every Google Ads Lead Form and web submission
Structured qualification: ownership, monthly bill in £, EPC rating, roof pitch and orientation, timeline
Instant utility + EPC lookup enriches lead automatically
Only leads scoring 6+ out of 10 land on a rep's calendar
Sub-6 leads enter a 60-day nurture with SEG rate updates and MCS-installer education content
How it runs end-to-end
Lead arrives
Google Ads / Meta / website / Bark form fires webhook.
AI qualifies
5-question conversation via WhatsApp or voice — under 3 minutes.
Enrichment
EPC register + postcode + property data enrich the profile.
Score + route
Score 6+ books an appointment; 4–5 enters nurture; 0–3 gets a polite decline + resource.
Rep works only hot leads
Sales calendar filled with pre-qualified consults; time on dead leads collapses.
The stack we used
AI Voice: LiveKit + GPT-4o Realtime + ElevenLabs
WhatsApp: MessageBird WhatsApp Business
Enrichment: EPC Register API + Postcodes.io + ATTOM
CRM: Salesforce Sales Cloud
Scoring engine: custom rules + LLM sentiment on transcripts
Notifications: Slack for reps + weekly ops dashboard
What it looks like in production
5-question WhatsApp flow that scores every lead in under 3 minutes.
Score, reasoning and enriched property data on every lead in Salesforce.
Only 6+ leads land on the rep's day — full context in the calendar invite.
Sub-6 leads warmed over 60 days with SEG + MCS content.
The numbers that moved
"We stopped hiring more reps and instead let AI filter leads. Same headcount closes 2.4x more installs — best decision of the year."
Answers to the usual questions
Doesn't filtering leads mean losing pipeline?
No. Filtered-out leads are renters or unsuitable roofs — they cannot close regardless. Filtering just stops reps burning hours on them.
How is scoring calibrated?
We start with a rules-based scoring model, then tune weekly against actual close outcomes from Salesforce. Continuous learning.
Does it handle MCS-specific compliance?
Yes. MCS + SEG paperwork prompts are baked in; the bot pre-collects the exact data an MCS installer needs for the final quote.
Other solar deployments
Residential Solar AI: 4.1x Booked Consults in 45 Days
A Texas residential installer was spending $46,000/month on Google Ads and only reaching 22% of leads. Within 45 days of deploying our AI voice + WhatsApp + HubSpot stack, cost per booked consult dropped 61% and monthly installs climbed from 11 to 45.
ReadCommercial Solar AI: 2 x 400kW Deals Won From Weekend RFQs
A Pune-based C&I solar developer was losing weekend RFQs from procurement teams. Deploying an AI receptionist with intelligent routing captured every enquiry, qualified MW target and financing structure, and pinged the right named account executive on Slack — winning two 400kW deals in month one.
ReadSolar Audit AI: Free Online Audit Booked 340 Consults in 60 Days
A Florida installer replaced their generic 'get a quote' form with an AI-powered instant solar audit. Homeowners input their address and bill; the tool returned an estimated system size, savings and payback in 45 seconds — and 3.2x more visitors booked a consult.
ReadReady for a similar deployment?
Book a 20-minute audit. We will show a live demo trained for your service area, panels and CRM — with real numbers, not slides.