The shop, before

The customer is a Pacific US residential solar installer running a 6-rep inside-sales team backing a smaller field team. Lead sources were standard for the industry: paid social, paid search, partner referrals, and a meaningful chunk of organic from the marketing site.

Their problem was the structural problem of solar lead-gen: most leads aren't qualified prospects, and the team finds out the hard way. The shape of the leak looked like this:

  • The unqualified majority. By their own data, about 62% of inbound leads failed at least one of the four basic qualifiers (roof age < 15 years, monthly utility bill > $80, homeowner status, suitable roof material). Most of these were discoverable in a 60-second conversation.
  • The SDR cost. The 6-rep team was spending the bulk of its time on initial qualification calls — many of which ended at the 90-second mark with the rep politely closing out a non-prospect. By the time the team got to a qualified consult, the rep was already context-fatigued.
  • The follow-up debt. Because every initial call took 15–25 minutes regardless of outcome, the team accumulated a follow-up backlog on the qualified prospects — leads that should have been closing this week were getting their second call in two weeks.

"My reps were spending 60% of their week qualifying leads we already knew weren't going to qualify. The data was there. Nobody was running the filter early enough."

— Sales director (anonymised at customer request)

What we deployed

Standard 14-day deployment timeline. Two channels were the primary deployment surface, with a third added in week three:

  • Web chat on the homepage, the savings-calculator page, and all paid landing pages, with the AI Employee handling first-contact qualification end-to-end.
  • Facebook / Instagram lead-form webhook — every paid social lead immediately got an AI Employee outreach within 60 seconds, while interest was still warm.
  • Inbound voice routing (added week three) — first-touch on inbound calls during business hours, escalated to a human SDR only when qualification cleared.

The solar qualifier was tuned around six gating data points: roof age, roof material (composition shingle vs. tile vs. metal vs. flat), homeownership status, monthly utility bill range, primary motivation (savings, environmental, off-grid, EV), and timeline.

Critically, the AI Employee was tuned to politely close out non-qualifying leads with a short explanation of why the timing wasn't right, plus a low-pressure invitation to reach back out when circumstances changed (e.g. roof replacement scheduled, or planned move into a homeowned property). This preserves the brand experience for non-prospects, which matters for word-of-mouth in dense suburban service areas.

The 90-day comparison

We compared the 90 days post-deployment against the matching 90-day window from the prior quarter. Lead-source mix and ad spend were within 8% across the windows.

Before · 90 days

  • 1,830 raw inbound leads
  • ~38% qualified (per post-call data)
  • SDR time per booked consult: 4h 12m
  • Avg first response: 4h 22m
  • Consult-to-close rate: 18%

After · 90 days

  • 1,902 raw inbound leads
  • 62% pre-qualified before SDR contact
  • SDR time per booked consult: 2h 04m (−51%)
  • Avg first response: 47 sec
  • Consult-to-close rate: 25% (+38%)

What the AI Employee actually did

The headline efficiency number is the SDR-hour reduction. Underneath it, three behaviours mattered.

1. It pre-qualified before the SDR ever picked up the lead. Of the 1,902 raw inbound leads in the post-deployment window, the AI Employee handled the entire qualification on roughly 62%. SDRs only saw the leads that had already passed the four gating qualifiers, with structured answers attached.

2. It politely closed out non-qualifiers with the brand intact. About 28% of inbound leads were politely closed out by the AI Employee (rented home, roof too old, utility bill too low, planning to move). These customers received a clear, accurate explanation and an open invitation to come back. Net Promoter feedback from the closed-out segment was actually higher than from the converted segment — likely because they got a fast, honest answer instead of a 20-minute sales-track call that ended in disappointment.

3. It moved the consult booking upstream. When a lead did qualify, the AI Employee booked the consult directly into the SDR's calendar — with the customer's qualification data already filled in. The first SDR touch was the consult itself, not a screening call. That single change is most of the +38% consult-to-close rate lift.

"My reps started showing up to consults already knowing the customer's roof age, utility bill, and timeline. Their first 30 seconds went from 'tell me about yourself' to 'based on what you've told the team, here's what we can do.' That changed everything."

— Sales director

What this didn't fix

  • It didn't increase top-of-funnel volume. Lead volume was effectively flat across the comparison windows. The lift came from doing more with the same lead flow, not from generating more leads.
  • It didn't replace the SDR team. The team's headcount stayed the same; their time mix shifted from qualification to closing.
  • It didn't fix lead-source quality. If anything, it made the lead-source quality picture sharper — the team now has clean per-source qualification data, which has changed how they think about ad spend.

What it would cost a similar installer

For a residential solar installer with a similar inside-sales team and lead-mix, the typical setup is:

  • Setup and 14-day deployment — one-time fee.
  • Monthly subscription — based on inbound lead volume and channels deployed.
  • Per-channel pricing for installers running multi-source paid acquisition.

For specifics tuned to your team size, lead-source mix, and trade, book a 20-minute call.