Bought a Lead List? How to Turn Purchased Leads Into Booked Appointments With AI Calling

You can buy leads in an afternoon. Turning them into booked appointments is the part most teams get wrong, and they get it wrong in the first hour. Harvard Business Review's 2011 audit of 2,241 US companies found that firms contacting a lead within an hour were nearly seven times as likely to qualify it as those that waited even an hour longer — and the average response time among companies that responded at all was 42 hours. A purchased list that sits in a spreadsheet until Monday has already lost most of what you paid for.

This is a working guide to the whole chain, from buying a verified list without trusting a stranger with your money to measuring what an appointment actually costs.

How to buy leads without getting burned

Most advice on buying lead lists is about vetting the seller: ask for a sample, ask where the data came from, start small. All of it is sound, and none of it protects the money. The usual failure is a stale or mis-targeted file discovered after payment, with nothing to be done. The fix is to make payment conditional on the leads checking out — which is what escrow is for.

On LeadPedlar's lead marketplace that is how every trade works. Sellers are reviewed and their lists verified; a listing shows sample rows (phone numbers and emails masked), the seller's badges and the delivery terms, and the buyer pays into escrow in USDT or other major crypto assets — the seller is not paid at that point. The full list unlocks for the buyer, who can dial a share of the leads to check connection quality and data accuracy, then release the funds or, if the data is invalid, open a dispute for mediation. Signing up is free, with no card and no KYC.

What to inspect before you pay, on any marketplace: that records carry a phone number and a location (you will need the lead's time zone for calling hours), how and when the leads were generated, whether the geography matches where you sell, and whether the price is per lead or for the full dataset. Then test before you release.

Import into the CRM before anyone dials

A purchased list is a file; a lead is a record with a history. The import is where one becomes the other, and skipping it is how teams call the same person twice from two spreadsheets. Bring the file in as one list, put every record in one pipeline stage (“New — purchased, batch 14”), and merge it against the leads you already have so an existing customer is never greeted as a stranger.

Two fields matter more than the rest: a dialable phone number — the test sample will have shown you how clean they are — and a location, because the campaign's calling hours are judged where the lead is, not where your office is. Tag the source and the price paid on the list; the cost model below needs both.

In LeadPedlar the CRM holds lists, pipeline stages, notes, callbacks, dispositions including sale amount, and a timeline per lead, so the batch, the seller and every call that follows sit on one record.

• ~7×More likely to qualify a lead when contacted within an hour than an hour later — audit of 2,241 US companiesHarvard Business Review (2011)
• 42 hAverage first response time among companies that responded at all; 23% never respondedHarvard Business Review (2011)
• 100×More likely to make contact when calling at 5 minutes rather than 30 — Lead Response Management study, telephone contactsLead Response Management study (InsideSales.com & Dr. J. Oldroyd, 2007)
• ~5.4%Average cold-call connect rate across 300M calls — most dials end in a machine, a silent line or no answerGong Labs, 300M cold calls (2024)

The first call within minutes — and why that means AI

The response-time evidence is unusually consistent. The Lead Response Management study — InsideSales.com's call data analysed by Dr. James Oldroyd, telephone contacts only — found that calling within five minutes rather than thirty made contact 100 times more likely and qualification 21 times more likely. HBR's audit found only 37% of companies responded within an hour. Both studies are old, but nothing since has reversed the direction; our speed-to-lead article walks through the decay curve.

A purchased list breaks the rule by construction. An import creates a thousand “new” leads in the same second; a team of five can call five in the first five minutes, and the other 995 wait — some for days. Only software can give every record a first call in the first minutes, and that software is an AI calling agent, not a dialer that still needs a free human for each answer. Run the list as a campaign in AI first mode and the agent works through it from the moment it is loaded, inside the calling hours you set.

What the agent says comes from the campaign's context — a plain-text brief the team leader writes and reviews with the AI: the business name, the opening line, the offer, the goals (transfer to a live person first, otherwise book a callback; or callbacks only), what never to promise, and the voicemail. Write it like a briefing for a new hire who will make a thousand calls tomorrow. Our walkthrough of how an unattended campaign runs shows each screen; agentic lead qualification covers writing questions the AI can use.

What the AI does with each kind of answer

At Gong's 5.4% connect rate for cold calls, nineteen of twenty dials do not reach a person. A purchased list usually does better, but most first attempts still meet a machine or nothing. The agent starts by listening first rather than talking over whoever — or whatever — picks up.

• A person answers. The AI opens with the line from the context, confirms who it is speaking to and works towards the goal with short turns that stop when interrupted. An interested lead is handed live to a closer or a SIP extension with a summary; someone interested but busy gets a callback at the time they choose; “not interested” and do-not-call requests are respected and recorded.
• An answering machine. The AI recognises the greeting, waits for the beep and leaves a message — written for that lead from the campaign context, or the campaign's own recorded voicemail — and stops if a person picks up mid-message. Phone numbers are read digit-group by digit-group.
• A silent line or a carrier intercept. The AI classifies the answer, ends the call cleanly, and the lead goes back into the queue for another attempt at a different time of day.

Voicemails are not a consolation prize. Gong's data shows a cold call ending in a voicemail lifted the follow-up email's reply rate from 2.73% to 5.87%, so an agent that leaves a clear, personalised message on every machine is doing half of the next touch. Our article on answering machine detection and AI voicemail drop explains why the beep matters.

Getting the warm ones to closers: live transfer, callbacks and SIP extensions

The point of qualifying purchased leads with AI is that a person only talks to someone who wants to talk, and the hand-off wiring decides whether the closer's day is a stream of warm conversations or a series of missed transfers. In LeadPedlar, hand-off rules decide who is rung — a team, particular people, team leaders, admins or SIP extensions — and in what order, and every hand-off's result is recorded: connected to a person or an extension, nobody available, all agents busy, or no answer and back to the AI.

The live roster shows who is online, on a call, ready, paused or offline, so transfers go to people who can take them. When nobody is free the AI does not leave the lead on hold; it books a callback and the lead lands in the CRM with the time agreed. Closers who prefer a desk phone take transfers on a softphone — Zoiper, X-Lite or Bria — through a SIP extension.

Every transfer arrives with a summary of who the lead is and what was said, so the closer does not re-ask what the AI just asked. That matters for trust as much as efficiency: in a Gartner survey of 5,728 customers, 64% would prefer companies did not use AI for customer service, and the top concern was that it would become harder to reach a person.

Redialing by outcome: working the list to the end

One pass never finishes a purchased list. After the first run, the campaign stats show what you have: calls today, total dialed, connected, voicemails left, transfers, callbacks booked, sales and sales amount, leads remaining. In LeadPedlar each number opens the leads behind it with two buttons — Call with AI and Call myself — so redialing by outcome is a click, not an export.

A sensible cadence, which you set yourself: unreached leads get another attempt at a different hour of their day; leads with a voicemail left get a second attempt the next day and a second voicemail only if still unreached; callbacks are made at the time agreed — by a person if the lead asked for one; “not interested” is left alone. Because calling hours are configured per campaign by the lead's time zone, redials stay inside the window on their own.

Decide in advance how many attempts a lead gets before it is marked exhausted, and record the share of each batch you reached. Reached per thousand bought is the honest quality score for a seller.

1. 1

   Step 1: Buy through escrow

   Inspect sample rows, seller badges and delivery terms; pay into escrow; test a share of the delivered leads; release the funds or open a dispute.

2. 2

   Step 2: Import into one list

   One pipeline stage for the batch, duplicates merged, every record with a phone number and a location.

3. 3

   Step 3: Write and review the context

   Business name, offer, opening line, goals (transfer first or callbacks only), what never to promise, the voicemail — reviewed with the AI.

4. 4

   Step 4: Start the campaign AI first

   The agent calls every new lead within minutes, inside calling hours set by the lead's time zone, and hands over to a person on request.

5. 5

   Step 5: Wire the hand-offs

   Hand-off rules name who is rung and in what order; nobody free means a callback, not a hold.

6. 6

   Step 6: Redial by outcome

   Open each stat, pick the group — unreached, voicemail left, callbacks due — and send it back to the AI or call it yourself.

7. 7

   Step 7: Measure cost per kept appointment

   List price plus AI calling time plus closers' hours, divided by appointments that happened.

What an appointment actually costs: an illustrative model

The list price is the number everyone negotiates and the smallest number in the model. Here is an illustrative batch with every assumption stated, so you can swap in your own. One thousand leads bought at $0.50 each: $500. The AI makes up to three attempts per unreached lead over a week — roughly 2,400 dials. Thirty per cent of leads are reached in a live conversation (300); one in five of those is qualified and handed off or booked (60); two thirds of the appointments are kept (40).

The AI's time: 2,100 attempts ending in a machine, a silent line or no answer at about half a minute each (around 1,050 minutes, voicemails included) plus 300 conversations at three minutes (900 minutes) — roughly 1,950 minutes, or 33 hours, of AI calling. Published list prices for voice-AI platforms run from $0.07 to $0.31 a minute before telephony (Retell's price page; Bland lists $0.12–$0.14), which puts the AI stack at roughly $140–$600 for the batch; LeadPedlar's own rates are on the pricing page. The illustrative total is about $640–$1,100 for 40 kept appointments — $16–$28 per appointment before telephony — with the closers' time spent on 60 warm conversations rather than 2,400 dials.

Work the same list by hand and the shape changes before the price does. At 45 seconds per attempt, 2,400 attempts are 30 hours of dialing plus 15 hours of conversation plus wrap — more than a working week for one rep at the US median customer-service wage of $21.53 an hour (May 2025) before overheads. The money is comparable; the week is not — most leads get their first call days after import. Our AI call centre cost breakdown builds the full model.

How LeadPedlar takes a purchased list from escrow to appointment

Across several tools the thread breaks at each hand-off: the marketplace does not know the list was called, the dialer does not know what the AI said, the CRM does not know a callback was promised. LeadPedlar puts the four steps on one platform — an escrow-protected lead marketplace, a CRM, a cloud dialer that calls from the browser through any SIP provider, and AI calling agents — with teams, roles and permissions across all of them.

• Fully autonomous. AI start to finish works the list unattended: first calls within minutes, voicemails on the beep, callbacks booked, “not interested” respected, calling hours judged in each lead's time zone — for callback-only and appointment campaigns.
• Blended. AI first, people when the caller asks is the default for purchased lists with closers on hand: the AI qualifies, and interested leads go live to a person or a SIP extension with a summary. AI steps in covers the other direction: people dial, the AI takes the overflow and the voicemails.
• Human-only. Manual dialing from the record, or autodial for teams with the dialer feeding the next lead and answering-machine handling options — with recording and live monitoring.

The mode is a campaign setting, so one batch can start AI first and move its callbacks to people, or start human-only and send the unreached remainder to the AI. The AI agent's brain can be the platform's AI or your own connected model, and AI calling hours are paid in crypto. The features page lists the campaign modes; our comparison of AI, predictive and power dialers explains why the AI first touch leaves nobody hearing dead air.