Product
MCP serverREST APICaller appChatGPT, Claude & Muse
Use cases
Lead follow-upAI SDR phone stepAgencies
Developers
DocumentationQuickstartSandboxllms.txt
More
CustomersPricingGuidesSign inGet API key
Rank.ai Customer story

How Rank.ai calls every grader lead within minutes.

Rank.ai's free website grader turns local businesses into warm leads. Its agent now writes a call brief from each grade, and a human calls while the owner is still looking at their score.

Pilot in progress

The figures below are illustrative targets from the pilot design. We'll replace them with measured results, including the email-only holdout, once the pilot completes.

<10 minFrom grade to first call attempt, down from next-day email follow-up
3.6×More booked meetings per lead than the email-only holdout
38%Of called leads reached a decision-maker
$61Cost per booked meeting, including free calls and bonuses

The company

Rank.ai helps local businesses (dentists, plumbers, med spas, law offices) win more customers from search and AI answers. Its free website grader scores a site, flags the gaps, and builds a refreshed preview at {business}.rank.dev.

The problem

Grader leads are the warmest Rank.ai gets: an owner who just watched their site score 41/100. But follow-up was email-only, and most local-business owners don't live in their inbox. By the time a reply came, the moment had passed.

Calling every lead by hand wasn't realistic, and AI voice agents were off the table. AI-generated voices are treated as artificial-voice calls under the TCPA. Here's why.

The solution

  1. Grade completes → agent writes the briefRank.ai's agent turns the grade into a 150-word brief: the score, the top three gaps, the preview link, and the ask (a 15-minute walkthrough).
  2. One tool callThe agent calls create_call_task with the business's main line (consent: "b2b_main_line", or inquiry when the owner entered their own number), the state, and a 30-minute deadline.
  3. A human callsA Ring a Human caller claims the task, opens the grade and the preview, and dials during the business's local hours.
  4. Outcome back into the funnelThe webhook writes the disposition and notes to the lead. Booked meetings land on the calendar; voicemails trigger a tailored follow-up email.

A real-shape brief

You're calling on behalf of Rank.ai about the free website grade Bright Smile Dental ran at 10:02am. Score: 41/100. No review schema (Google hides their 4.8★), 7.2s mobile load, no booking button above the fold. We built a refreshed preview: brightsmile.rank.dev Ask: 15 minutes this week with Dr. Patel or the office manager to walk through it. If "we have a web guy": "Great, it's free; can I send the report so they can use it, and book 15 minutes so it makes sense?"

How the pilot is measured

  • Randomized holdout: 50% of grader leads with a phone number get a call task plus the usual emails; 50% get emails only.
  • Primary metric: booked meetings per lead. Secondary: time to first touch, connect rate, paid conversions, cost per booked meeting.
  • Guardrails: calls only 9am–8pm local, one attempt per lead, do-not-call checked before every dial.

Results (illustrative)

MetricEmail onlyEmail + Ring a Human
Time to first touch~20 hours<10 minutes
Reached a decision-makern/a38%
Booked meetings per 100 leads3.111.2
Cost per booked meetingn/a$61

Illustrative pilot targets. Measured results will replace this table.

Run the same playbook.

Start in the sandbox, then 2 free real calls.

Get API key →