In your phone system, send calls to a VoiceGenie number when no one picks up after four rings, and all calls after hours. Daytime calls still ring your team first. Voicemail stays as the fallback.
AI reaches every lead in a minute. Your closers close.
The AI answers every call and calls every form lead, asks the fit questions and books. Your closers keep the buyers, the price talks and the named accounts, so you get speed and judgment together.
- Answers on the first ring
- Warm transfer with full context
- Writes to your CRM and calendar

After-hours sales call
A fleet buyer rings at 7:20 pm. The AI asks the fit questions and books or transfers.
New web-form lead
A demo request comes in at 4:50 pm. The AI calls within a minute.
Ready to buy now
A contract ends in three weeks. The AI hands the call to a closer with notes.
Price talks and named accounts
A discount or a VP at a target account goes straight to a person.
Why lead phones will run on AI and people together
Lead phone work has a timing problem and a judgment problem. Of 9,538 companies tested, 47% never answered a web inquiry, and those that phoned took a median of 3 h 8 min. A people-only team cannot answer every 7:20 pm call or every form inside a minute. An AI-only line is unsafe when a buyer with a three-week deadline needs a decision maker.
- Answers every sales call on the first ring, even at 7:20 pm
- Calls a consented form lead inside a minute, every time
- Asks the same four fit questions on every call, no skipped steps
- Reads the buyer and finds the real problem behind the quote request
- Handles price talks, contract terms and multi-site deals
- Carries the relationship with named accounts and executives
- The 7:20 pm fleet buyer gets a voice, then a closer who already knows the deadline
- Closers stop screening vendors and wrong numbers and spend the day on buyers
- Every closer rejection sharpens the AI's questions, so fit gets better each week
Anything where minutes matter and the questions repeat goes to the AI. Anything with money, terms or a relationship at risk goes to a closer.
Do not make the AI close. Let it gather fit and timing, then pass the call warm so the lead repeats nothing.
Accepted and rejected leads are the best training data you have. Review them every week and tighten the questions.
How do you qualify leads by phone with AI and people together?
Let the AI answer or call every lead within a minute, ask four or five fit questions, and then act on the result: warm-transfer leads ready to buy, book a meeting for clear fits, and tag the rest for nurture with a reason. Closers keep pricing talks, named accounts and negotiation. Start with one lane, such as sales-line calls nobody answers today.
- 47% of 9,538 companies tested never responded to a web lead (InsideSales, 2013 test)
- Calling at 5 minutes vs 30 raised contact odds about 100x (15,000+ leads; close rates not measured)
- A 4-minute qualification call costs about $0.56 of AI time at $0.14 per minute, billed per second
- Human seats start at $29 a month on annual billing; no minimum commitment
Most leads are lost in the first hour, before anyone pitches.
You already paid for the demand. What decides its value is how fast a live voice answers and whether the right questions get asked. The best-known studies are older and measure contact, not closed deals, but recent audits show the gap is still open.
of 9,538 companies tested never answered a web inquiry; those that phoned took 3 h 8 min at the median (2013 test)1
of 6,346 company forms got no email reply, and 96% of companies never phoned (Clay audit, September 2026)2
higher odds of reaching a lead called at 5 minutes than at 30 (6 companies, 15,000+ leads, 2007 study)3
of businesses asked callers for the sale or appointment, as scored by AI (Invoca, 70M+ calls, 2026)4
Where AI and closers work together, from first inquiry to win-back
A lead is not one call. It is a form, a qualification, a meeting, a quiet stretch, a handoff and sometimes a second chance. Here is who does what at each stage, so no lead waits and no closer starts cold.
Read down a column for one stage, or across a row to see what the AI and your team each do from first contact to win-back.
Which lead calls the AI owns, which it starts, and which go straight to a closer.
Sort by what is at stake and how much judgment the call needs, not by how the lead arrived. Most volume is simple fit questions. The few calls with money, relationships or regulated advice go to a person.
One web lead, start to finish, in about four minutes.
A fleet manager submits a quote form at 4:50 pm. This is the path that beats the next-morning callback, and the points where a person takes over.
Form submitted, checks run
The form carried the call-consent wording. VoiceGenie checks the number, the do-not-call list and the lead's local time, then reads the CRM for an owner and territory.
The AI calls
It says who it is, that it is an AI and that the call is recorded, then asks if now works. If not, it texts a link to pick a time.
Four fit questions
What prompted the request, how many vehicles or seats, what they use now, when they need it. Budget comes last, as a range, and can be skipped.
Decide on the call
Clear fit and urgent: transfer to the closer on duty. Clear fit, not urgent: book a meeting. Not a fit: thank them, send resources, tag nurture with the reason.
Closer picks up with context
The closer sees a one-line summary, the transcript, the lead's mood and the answers. The lead repeats nothing. If nobody answers in 20 seconds, the AI books.
Review and feedback
Notes, disposition and meeting are in the CRM. A supervisor scores a sample of calls. Closers mark each lead accepted or rejected, and rejections improve the questions.
Start with sales-line calls nobody answers today.
Pick one lane where today's result is voicemail. You cannot make speed to lead worse there, and you learn the whole path: answer, ask, book, hand off. Add one form callback only after closers trust the notes.
Unanswered sales calls: after hours and overflow after four rings
These callers reach voicemail now, and under 3% of people sent there leave a message (Invoca platform data). They asked to talk, so there is no outbound consent question. Closers see results within days.
Definitions and baseline
- Agree with closers on what "qualified" means: three must-haves, and the reasons a lead is not ready.
- Pull 6 to 8 weeks of call logs and form timestamps. Count missed calls, voicemails and minutes to first live touch.
- Write four or five questions, the meeting rules, and who takes a warm transfer at each hour.
- Load the do-not-call list, the opt-out phrases and the AI and recording notice.
After hours only
- Forward nights and weekends only. Voicemail remains the fallback.
- Run 20 test calls: price shopper, wrong number, vendor, angry caller, "let me talk to a person".
- Listen to every call the next morning and fix wrong answers the same day.
- Check each booking lands in the CRM with owner, source, answers and the call record.
Add overflow and live transfer
- Add daytime overflow after four rings. Name one closer on duty for each hour.
- Review every call in week 3 and a 25% sample in week 4, plus every lead a closer rejects.
- Time each warm transfer, from "one moment" to a closer saying hello. Set a fallback booking at 20 seconds.
- Send closers a daily list: booked, transferred and nurture, with the reason.
Add one form callback, then decide
- Add one web form that shows call consent. Call within 60 seconds, 8 am to 8 pm in the lead's time zone.
- Leads that arrive out of hours wait for an 8 am local call, never a night call.
- Compare with the week 1 baseline. Hold a 30-minute review of the ten worst calls with closers.
- Decide in writing: widen to more forms, fix and extend two weeks, or stop.
Expand when all of these are true
- At least 95% of sales-line calls get a live answer, against your week 1 baseline
- Closers accept at least 70% of AI-booked meetings, or within 10 points of person-booked ones
- Held-meeting rate on AI-booked meetings is within 10 points of person-booked meetings
- No missed opt-outs and no dropped warm transfers in the audit
The numbers to manage by, with honest baselines.
Agree the definitions with closers before you launch, then measure each one from the CRM and call record. Count a response only when a live attempt happens. An auto-reply email is not one.
Planning ranges are targets to test in a pilot, not promises. Measure your own baseline for two weeks before you change anything.
What a mid-sized lead line might return.
Illustrative inputs for a team with 1,200 lead calls and form leads a month. Replace each with your own numbers; the page computes results from them.
A worked example. Move the sliders to match yours.
AI at $0.14 a minute with volume credit applied, plus Insights seats for the people who take handed-over calls. Phone line charges are your carrier's and are not included. Illustration, not a quote. Open the full calculator
Where these numbers come from
- Cost per person: median SDR pay of $80,000 8, divided by 0.70 because benefits are 30% of private-industry pay 9, is about $9,500 a month before tools and management.
- Calls per person: SDRs log about 44 phone activities and 4.1 quality conversations a day 8. Inbound qualification calls need notes, so 30 is a planning input.
- Missed share: 47% of tested companies never answered a web lead 1 and two in three forms got no reply 2. 30% is deliberately lower than both.
- Conversion: 15% is near the low end of the 13% to 21% reported MQL to SQL range 5.
- Outcome value is yours to set. Example: $6,000 gross profit per won deal times a 20% meeting-to-win rate. Call length of 4 minutes is a planning input.
What the dollar figure leaves out
- Closers open each conversation with the answers already written down.
- A lead who calls at 7:20 pm hears a voice, not a voicemail greeting.
- Marketing sees which sources produce real conversations, not just form fills.
The rules that decide whether you can call a lead at all.
US-first, as of November 2026, and not legal advice. Calls to leads are telemarketing in most readings, so consent, hours and opt-outs matter more here than on inbound-only lines.
Written consent before an AI calls a lead10
The FCC treats AI voices as artificial voices (2024). Telemarketing with one needs prior express written consent: a signed agreement, e-signature allowed, naming the number and saying consent is not a condition of purchase.
One seller per consent: struck down, still wise11
The FCC's one-to-one rule was vacated in January 2025 and the text deleted, so shared-lead consent is judged under the older definition. Name your company on the form anyway, and keep the wording, time and number.
Calling hours and state limits12
The federal window is 8 am to 9 pm at the lead's location. Oklahoma and Washington stop at 8 pm, and Oklahoma caps sales calls at three per 24 hours. Other states vary, so default to 8 am to 8 pm and three tries a day.
Do-not-call lists and the inquiry window13
An inquiry gives a 3-month window to call under the FTC's business-relationship rule, unless the person says stop. Scrub the National Registry within 31 days of calling and keep your own list. Fines reach $53,088 per violation, and each call can count.
Opt-out in any words, honored fast14
Honor revocations and internal do-not-call requests within 10 business days at most, and keep a do-not-call request for 5 years. From November 25, 2026, an opt-out from one marketing call covers all your marketing calls and texts, and you may name the opt-out method on each call.
Say it is an AI and recorded, first15
Thirteen states need all-party consent to record. Put the AI notice and recording notice in the first sentence, before any question, and offer a person at any point. One script meets every state.
Four questions to ask any voice AI vendor, including us
- Can you log consent wording, time zone and do-not-call checks for each outbound call, per lead?
- Do you use my call audio to train your models, and can that be turned off in the contract?
- If no closer answers a warm transfer, how soon does the lead get a booking instead?
- Does the AI say it is an AI at the start and hand to a person whenever the caller asks?
This is general information, not legal advice. Rules differ by state and country and change often. Confirm with your own counsel before you call.
Where lead data lives, and what the AI reads and writes.
VoiceGenie typically connects to these through its open API, webhooks or Zapier. The CRM stays the system of record: the AI reads the owner and history, then writes the disposition back.
CRM and lead records
- SASalesforce
- HUHubSpot
- ZCZoho CRM
- PIPipedrive
- MDMicrosoft Dynamics
Forms and lead sources
- HFHubSpot Forms
- MAMarketo
- TYTypeform
- GLGoogle lead forms
- MLMeta lead ads
Routing and scheduling
- LELeanData
- CPChili Piper
- CACalendly
- GCGoogle Calendar
Phone and sales engagement
- TWTwilio
- RIRingCentral
- Any SIP trunk or PBX
- OUOutreach
- SASalesloft
One platform for the dialer, the AI agent and your closers
You do not need one vendor for the dialer, one for the AI agent, one for the closer desk and one for tools that help reps. VoiceGenie is all four on one record. You set it up once, and every input improves both the AI and your closers.
Four tools, four set-ups
own training · own reports
own training · own reports
own training · own reports
own training · own reports
One platform, set up once
- Your phone numbers and carrier
- CRM and calendar connections
- Knowledge and call rules
- Recordings, QA and analytics
Five ways lead qualification projects go wrong.
Questions people ask before they start
What is lead qualification, and how does AI do it?
Lead qualification is deciding, early, whether a lead fits what you sell and is ready to talk to a closer. The AI answers or calls the lead, asks a short set of fit questions, such as need, timing and decision maker, and records the answers. It then books, transfers or tags the lead for nurture, using rules you set.
How fast should we respond to a new lead?
Aim for a live attempt inside one minute, and never leave a lead waiting for hours. The best-known study found that calling at 5 minutes instead of 30 raised contact odds about 100 times. It measured contact and qualification, not closed deals, and the data are old, so treat it as direction.
Can AI call people who fill out a web form, and is it legal?
Yes, if you have the right consent. In the US, an AI voice is an artificial voice, so telemarketing calls need prior express written consent: a signed agreement, e-signatures allowed, naming the number. Stay inside 8 am to 8 pm local time, honor opt-outs, and check with your counsel before you launch.
Will leads hate talking to an AI?
Some will. In a 2023 Gartner survey of 5,728 customers, 64% said they would prefer companies not use AI in customer service, and their top worry was reaching a person. So the AI says what it is, answers fast, and passes the lead to a closer on request, with the notes, so nobody repeats themselves.
What questions should the AI ask to qualify a lead?
Four or five: what prompted the request, how big the need is, what they use today, who decides, and when they want it solved. Ask budget last, as a range, and allow a skip. Match the questions to your closers' definition of qualified, and cut any question they never use.
What does AI lead qualification cost?
AI minutes are $0.14 per minute, billed per second, and the phone line is extra. A four-minute qualification call is about $0.56 of AI time. Human closers on the Human Agent Desk start at $29 per seat per month on annual billing. There is free trial credit and no minimum commitment.
How does the hand-off to a closer work?
The call moves to the right closer by skill, queue and hours, with a one-line summary, the transcript, the lead's mood and what the AI already did. The lead repeats nothing. If no closer answers within the time you set, the AI books the next open slot so the lead is not left on hold.
What if the AI wrongly turns away a good lead?
Prevent it by design. The AI tags a lead as nurture with a reason, never deletes it, and any caller who asks for a person gets one. Each week, listen to a sample of nurture calls, count how many a closer would have taken, and tighten the questions until that number is near zero.
Related guides
From the VoiceGenie blog
Hear it handle your calls before you change anything.
Talk to our AI agent in your browser, or start a trial and point one line at it. Your team keeps every call that needs them.
No credit card · No minimum commitment · Keep your carrierSources
- InsideSales.com (XANT), 2014 Lead Response Report (tests run in 2013) (2014).
- Clay, We asked 6,346 companies for a demo. Most never wrote back. (2026).
- InsideSales.com and MIT researcher, The Lead Response Management Study (Oldroyd with InsideSales.com) (2007).
- Invoca, Lead Conversion Rate: 2026 Benchmarks From 70M+ Calls (2026).
- LeanData, Why B2B lead-to-opportunity conversion rates decline (2026).
- RevenueHero, Inbound lead response time benchmarks by company size (2026).
- Glencoco, 2026 State of Sales Development Report (2026).
- The Bridge Group, SDR Models, Motions & Metrics: 2025 Research Report (2025).
- U.S. Bureau of Labor Statistics, Employer Costs for Employee Compensation, June 2026 (2026).
- Cornell Legal Information Institute (eCFR text), 47 CFR 64.1200, Delivery restrictions (2026).
- Womble Bond Dickinson, FCC repeals one-to-one consent rule following Eleventh Circuit decision (2025).
- Manatt, Phelps & Phillips, What you need to know about Washington's and Oklahoma's telephone solicitation laws (2022).
- Federal Trade Commission, Q&A for telemarketers and sellers about DNC provisions in the TSR (2026).
- National Law Review, FCC Overhauls TCPA Consent Revocation Rules (2026).
- King & Spalding (KTS Law), Wiretap laws in the United States (2024).
- Cooley LLP, FCC: AI-generated robocalls are illegal under the TCPA (2024).
- Gartner, Gartner survey finds 64% of customers would prefer that companies didn't use AI for customer service (2024).
- Invoca, How much missed sales calls cost home services businesses (2024).