Lead lists · 2026
Where to get lead lists for cold calling
Last updated September 3, 2026
Almost every article answering this question is an affiliate page. This one sells a dialer, not data, so we have no reason to steer you anywhere: below are the paid sources with the prices they actually publish in September 2026, the free and public sources that beat them for certain markets, what a single usable phone number costs on each plan, and the compliance layer that comes attached to a bought list. Where a vendor does not publish a price, we say so instead of inventing one.
The short answer
Match the source to the target, not to the review site. Local businesses: the Google Places API, about four cents a record with a listed phone, entirely within terms. Regulated trades: the state licence registry, free and more current than anything you can buy. Named people at real companies: a contact database, where UpLead and Lusha are the cheapest credible entries and Cognism owns European mobiles. Whatever you pick, buy the smallest sample first and dial 100 numbers before you buy volume — and budget for the fact that the data usually costs more than the software and the telephony combined.
Paid providers, with the prices they publish
All figures are list prices from the vendors' own pricing pages, checked September 2026, except where noted as reported by third parties because the vendor publishes nothing. Prices are per user per month unless stated.
| Source | What you get | Price | Phone data | Notes |
|---|---|---|---|---|
| Apollo | Large contact database plus sequencing and a dialer add-on | Reported $49–$119/user/mo annual ($59–$149 monthly) | Broad coverage, variable on mobiles; mobiles use separate credits | Self-serve, free tier to test. Dialing needs the Advanced Dialer add-on (~$119/mo annual). |
| ZoomInfo | The largest US B2B dataset, intent data, org charts | No published price; third parties report low five figures a year and up | Best-in-class US direct dials | Annual contract, seat minimums, sales process. Overkill below ~10 reps. |
| Cognism | EMEA-strong data with human-verified mobiles | No published price; contracts reported in the tens of thousands a year | The strongest European mobile coverage we could find | Markets itself on GDPR posture and notified-of-processing workflows. |
| Lusha | Contact lookup, browser extension, prospecting lists | $69.90/user/mo ($52.45 annual), 600 credits | 10 credits per phone number, 1 per email | Free tier of 40 credits a month. Cheapest credible way to sample real data. |
| Seamless.AI | Contact search with real-time verification claims | Basic reported ~$147/user/mo annual, 250 credits/mo | Mixed reports; verify with a sample before committing | Pro and Enterprise are quote-only. Check the auto-renewal terms. |
| UpLead | Contact search with per-contact credits, verification on unlock | $99/mo Essentials (170 credits) or $74/mo annual; $199/$149 Plus (400) | One credit unlocks email and mobile together | The clearest per-contact pricing of the group; extra credits $0.50–$0.60. |
| LeadIQ | Capture-to-CRM workflow, contact tracking on job changes | $200/mo Pro for up to 5 users (200 credits); $150/mo annual | 10 credits per number — 20 numbers a month on Pro | Free plan of 50 credits. Priced for workflow, not for volume list building. |
Two honesty notes. Apollo's comparison table would not render for us on the day we checked, so its band is marked as reported rather than confirmed; the one Apollo figure we are confident in is that power and parallel dialing require the Advanced Dialer add-on at around $119 a month on annual billing, on top of a seat (see Apollo dialer minutes and limits). And ZoomInfo and Cognism genuinely publish no prices — every specific dollar figure you will read about them online is somebody's reported contract, not a rate card.
The number that actually compares them: cost per phone number
Sticker prices are not comparable because credits are not equal. A phone number costs 10 credits at Lusha and 10 at LeadIQ against 1 credit for an email; Apollo bills mobiles from a separate mobile-credit pool; UpLead spends one credit per contact and includes the mobile in it. Divide the plan price by the phone numbers the plan can actually unlock and the ranking changes completely:
| Plan | Monthly | Phone numbers / mo | Cost per number |
|---|---|---|---|
| Google Places API (listed line) | pay as you go | unlimited | ~$0.04 |
| UpLead Essentials (monthly) | $99 | 170 | ~$0.58 |
| Seamless.AI Basic (reported) | ~$147 | 250 credits | ~$0.59 |
| Lusha Professional (annual) | $52.45 | 60 | ~$0.87 |
| LeadIQ Pro (annual, up to 5 users) | $150 | 20 | ~$7.50 |
Seamless is the one soft row: it publishes a credit allowance rather than a phone-credit multiplier, so that line assumes one credit per contact and is a floor, not a rate. Everything else is arithmetic on published figures.
That is a 180-fold spread, and it is not a quality ranking — a Places API record is a switchboard, a Lusha record is a named person's mobile, and they are not the same product. But it does explain why a team calling local businesses should never be on a credit-based database, and why a team chasing named decision-makers should never judge a plan by its monthly price. Work out how many numbers you need a month, then divide.
Data usually costs more than everything else combined
Here is the line item people forget when they budget an outbound motion. Take 1,000 dialable leads called over roughly ten working days at 100 dials a day — that is 910 billed minutes on the maths in how many minutes cold calling uses (voicemails bill a whole minute each, which is where most of it goes).
| Line item | Cheap source | Mid source | Expensive source |
|---|---|---|---|
| 1,000 phone numbers | $40 | $580 | $7,500 |
| Twilio, 910 min + one number | $14 | $14 | $14 |
| Dialer software (free plan / Pro) | $0–29 | $0–29 | $0–29 |
| Data as a share of the bill | 48% | 93% | 99% |
Telephony is $14 at Twilio's ~$0.014/min plus a ~$1.15/month number, and the software is free for one person up to 500 calls a month or $29 for 3 seats. The data is the budget. Which means the highest-leverage cost decision in outbound is not which dialer you use — it is whether you are paying database prices for numbers a public registry publishes for nothing.
Free and public sources worth the effort
| Source | What you get | Price | Phone data | Notes |
|---|---|---|---|---|
| Google Places API | Local business name, address, listed phone, category | ~$35 per 1,000 requests at the field tier including phone; ~1,000 free calls/mo | Main line, not a person. Accurate because the business maintains it. | Supported and within terms. Scraping the Maps interface is not — do not. |
| LinkedIn Sales Navigator | Targeting, org research, job-change alerts | $119.99/mo Core, $159.99/mo Advanced | None. It does not expose phone numbers. | Automated extraction breaches the user agreement. Use it to build the target list, source numbers elsewhere. |
| Chamber of commerce directories | Local member businesses with a contact number | Free to browse; membership sometimes unlocks the full list | Main line, generally current | Small volumes, unusually good fit. Check the directory terms before bulk copying. |
| Trade association member lists | Members in one vertical, often with specialisms | Free to free-ish | Business line, self-maintained | The highest fit-per-row of any source if you sell to one trade. |
| State licence registries | Every licensed contractor, agent, broker, practice in a state | Free; many publish bulk downloads | Business phone on the licence record, kept current as a condition of licensure | Public records. The best free source in existence for regulated trades. |
| County / assessor property records | Owner name, parcel, sale history, sometimes a mailing address | Free or a small per-search fee | Rarely included; needs a separate append | Consumer data, so consumer rules. Scrub against the DNC registry. |
On scraping, our position is boring and firm: do not extract data from a site in violation of its terms of service. Google and LinkedIn both prohibit automated collection from their interfaces, and LinkedIn has litigated it. Beyond the legal exposure, a scraper is a maintenance job that breaks whenever the page changes. Where a supported API exists — as it does for Google Maps data — use it; the Places route is cheaper than a database anyway. Where a directory is genuinely public, read its terms before bulk-collecting it, because "publicly visible" and "free to redistribute" are different things.
The compliance layer on a bought list
The obligations follow the caller, not the seller. A vendor's compliance page describes how they gathered the data; it does not transfer to you a right to call it. Four things to know before the first dial:
- United States. Consumer telemarketing must be scrubbed against the National Do Not Call Registry. FTC access fees are $82 per area code for fiscal year 2026 with the first five free and a nationwide cap of $22,626; the FTC has published an increase to $85 per area code and $23,425 nationwide effective 1 October 2026. Internal do-not-call requests must be honoured regardless of registry status, and calling hours are 8am to 9pm in the contact's local time. B2B calls are frequently exempt from the registry — frequently, not always. Detail in our DNC and TCPA guide.
- United Kingdom and EU. Live B2B calls are permitted under PECR without prior consent, but you must screen against the Telephone Preference Service and its corporate register, CTPS, and stop on request. Under UK GDPR you need a lawful basis; for cold outreach that is legitimate interests, and the ICO expects a written legitimate interests assessment done before you call. You also need to satisfy yourself the vendor collected the data lawfully and can evidence it. See power dialers in the UK.
- Canada. Telemarketer registration with the CRTC is free; a National DNCL subscription is not — an all-area-code annual subscription ran to $62,166 following the September 2025 rate increase. Exempt callers do not need a subscription, but the exemption is narrower than people assume. See power dialer Canada.
- Everywhere. Keep the provenance. Record which vendor or registry each row came from and on what date, in a column on the list. When someone asks where you got their number — and eventually someone will — the answer "a spreadsheet" is the wrong one.
None of this is legal advice; it is the shape of the questions to take to someone who gives it.
The quality test: dial 100 before you buy 10,000
Every vendor quotes an accuracy percentage. None of them define it the same way, and none of them are measuring your segment. The only test that means anything is your own dial tone. The procedure:
- Buy the smallest sample the vendor will sell, or use the free tier. Lusha, Apollo and LeadIQ all have one.
- Clean it properly first, or you will measure your own formatting rather than their data — see cleaning a lead list.
- Dial 100 numbers in one or two sessions, same rep, same script, same time of day.
- Count four things: rang at all, reached the named person, reached someone else, plainly dead.
- Compare against a control — a sample from a free source or your existing list, dialed the same way.
Read the result carefully. One hundred dials at a 12% connect rate carries a standard error of roughly three percentage points, so 8% and 16% samples can come from identical lists; do not re-rank two vendors on a four-point difference. What 100 dials does catch, unambiguously, is the failure case: a fifth of the numbers dead, or half of them reaching a receptionist who has never heard of the contact. That is the outcome you are paying $50 of sample credits to discover before you commit to a year.
A power dialer makes this test cheap, because 100 dials is about an hour rather than an afternoon and the dispositions are already counted for you at the end. Which is, honestly, the only reason we are qualified to write this page at all.
Test the list before you buy the list
Import a sample, dial 100 numbers, read the connect rate off the dashboard. Free for one person up to 500 leads and 500 calls a month, no credit card, running on your own Twilio at ~$0.014/min with no markup.
Dial a sample freeQuestions people actually ask
Where is the best place to get phone numbers for cold calling?
It depends on who you call. For local businesses with a published main line, the Google Places API returns a listed phone number for roughly four cents a record and is fully within terms of service, which beats every database on price. For named individuals at larger companies, you need a contact database: UpLead and Lusha are the cheapest credible entries, Cognism is the strongest for European mobiles, and ZoomInfo is the enterprise default at enterprise money. For regulated trades — contractors, agents, brokers — the state licence registry is free, public, and more accurate than anything sold to you.
How much does a B2B lead list cost in 2026?
Published self-serve pricing in September 2026: UpLead Essentials $99 a month for 170 credits ($74 a month annually), Lusha Professional $69.90 a month for 600 credits ($52.45 annually), LeadIQ Pro $200 a month for 200 credits across up to five users ($150 annually), and Seamless.AI Basic reported at about $147 a month for 250 credits a month. Apollo is reported at $49 to $119 per user per month on annual billing. ZoomInfo and Cognism publish no prices at all; third-party contract data puts typical annual commitments in the tens of thousands of dollars. The number that matters is cost per usable phone number, not the sticker.
Is it legal to buy a lead list and cold call it?
Buying the list is generally legal; calling it is where the obligations attach, and they attach to you, not to the vendor. In the US, consumer calls must be scrubbed against the National Do Not Call Registry ($82 per area code in fiscal 2026, rising to $85 from 1 October 2026, first five area codes free) and must honour internal opt-outs regardless. In the UK you must screen against TPS and CTPS and record a legitimate interests assessment under UK GDPR. In Canada, telemarketers register with the CRTC free and pay for a National DNCL subscription. A vendor telling you their list is "compliant" is describing how they collected it, not licensing you to call it.
Are free lead sources any good for cold calling?
For some markets they are better than paid data. Licence registries for regulated trades — contractors, electricians, real-estate agents, insurance brokers, medical practices — are public records, published by the regulator, usually downloadable in bulk, and carry a business phone that the licensee is obliged to keep current. Chamber of commerce and trade association member directories are small but exceptionally well targeted. County property and assessor records are the standard source for real-estate calling. What free sources will not give you is a named decision-maker with a direct dial at a 500-person company; that is what you pay databases for.
How do I test a lead list before buying a big one?
Buy the smallest sample the vendor will sell, clean it, dial 100 numbers, and count three things: how many rang at all, how many reached the named person rather than someone else, and how many were plainly wrong numbers. Treat the result as a smoke test, not a measurement — at a 12% connect rate, 100 dials carries a standard error of about three percentage points, so a sample showing 8% and a sample showing 16% may be the same list. What 100 dials does reliably catch is the disaster case: a list where a fifth of the numbers are dead or belong to someone else. That is the outcome you are testing for.
Can I scrape Google Maps or LinkedIn for phone numbers?
Not within their terms of service, and we do not recommend it. Both Google and LinkedIn prohibit automated extraction from their interfaces, and LinkedIn has litigated the point. The supported route to the same Google data is the Places API, where the field tier that includes a phone number bills about $35 per 1,000 requests with roughly 1,000 free calls a month — approximately four cents a record, cheaper than any contact database and without the terms problem. LinkedIn has no supported equivalent because Sales Navigator does not expose phone numbers at all; it is a targeting and research tool, priced at $119.99 a month for Core.
Why do phone numbers cost so many more credits than emails?
Because they are harder to get and harder to keep current. An email address can often be inferred from a naming pattern; a direct dial or mobile cannot be guessed and has to be sourced, contributed or human-verified. Vendors price that in through credit multipliers: Lusha and LeadIQ both charge 10 credits for a phone number against 1 for an email, and Apollo charges a separate mobile credit. The practical effect is that a plan advertised at 600 credits is a 60-phone-number plan if you only want numbers, which is why cost per phone number rather than cost per credit is the figure to compare across vendors.