TL;DR
- A phone field is not a call-ready contact. Mobile, person-assigned direct dial, shared office line and headquarters switchboard have different value. Ask vendors to separate them in the trial export, then measure whether the number reaches the intended person.
- ZoomInfo belongs on the shortlist for a team already buying an enterprise contact-data platform. Require a scoped quote and a territory-by-territory phone sample; public database size does not establish your connect rate.
- Cognism Pro is the clearest purchase here for teams that want on-demand mobile verification in a five-seat prospecting package. It is quote-priced, and API or bulk delivery requires separate commercial scope.
- Lusha works for rep-led prospecting and a scoped API deployment. A phone reveal costs five credits, with request charges on top. Budget API access on Premium or Scale, as its current public FAQ specifies; generating an API key does not establish the purchased endpoint entitlement.
- Datagma is a focused pay-on-found mobile layer. Its phone page lists 100 phones for $39/month on annual billing and larger bundles, while warning it does not provide mobiles for contacts based in France. Confirm the checkout because some price cards on its site disagree.
- LeadMagic suits an API-first mobile lookup and waterfall. Its current Essential plan is $99/month, with 5,000 credits; a successful mobile costs five credits, while a lower-tier search API request can add one credit. That is different from the obsolete $49.99 figure in the old article.
First, decide which phone number you are buying
A sales rep sees one field labeled “phone,” but it can represent four outcomes. A mobile is assigned to a person, although ownership and job association still need checking. A direct dial routes to an individual line or extension at a company. A shared office number reaches a location or team. A headquarters number reaches a switchboard. All can be useful in some campaigns; none should be merged into a single “direct contact coverage” rate.
The difference shows up after purchase. Suppose a provider returns numbers for 80 of 100 target people. If 30 are company switchboards and 20 belong to people who changed jobs, the useful phone coverage is at most 30, before unanswered calls. “80% phone coverage” would conceal the buying problem. The numerator for a calling program should be correct-person, permitted-to-use numbers of the required type. Track actual connection separately; dialing time, time zone and rep behavior also affect it.
Verification claims need the same care. A number can be correctly formatted and active on a carrier yet reach somebody else. A recent automated match may indicate a current association, but it is not the same as a human-confirmed contact. Cognism, for example, explicitly sells phone-verified mobiles and on-demand verification in its higher prospecting tier. That is a product distinction worth testing, not a reason to assume every number in its database was recently called.
Before a demo, write the target: “Find the current VP of Sales or Sales Director at 200 named companies in Germany, the UK and the US, return a mobile or person-assigned direct dial, and exclude shared company lines.” Change countries and roles to match your market. Tell every vendor whether a second-best title counts, whether voicemail on the correct person's line counts and how to handle a mobile associated with a former employee. If you call into several jurisdictions, have your team review the relevant calling, suppression and data-use requirements for each market.
The five products below suit different procurement models. Two are sales platforms with phone data as one component; one is a credit-based prospecting interface; two are focused API enrichment layers. The “best” option depends on whether you need a rep to reveal a number on demand, a monthly bulk refresh, or a fallback lookup when the primary source misses.
Five providers at a glance
| Provider | Best buying situation | Public phone-cost signal | Main caveat |
|---|---|---|---|
| ZoomInfo | Enterprise team buying contact data and workflow scope together | Quote required for the selected package | Test actual mobile/direct-dial mix in each territory |
| Cognism | Five-seat prospecting team needing on-demand mobile verification | Quote; one credit per revealed contact | Pro verification, API and bulk rights must be scoped separately |
| Lusha | Reps revealing contact fields and a bounded API pilot | Five credits per phone reveal | API request overhead and empty-response minimum can raise cost |
| Datagma | Pay-on-found mobile lookup or a second source in a waterfall | Phone page says $39/100, $79/250, $209/750 per month on annual billing | Vendor cards disagree; no France-based mobile data per vendor |
| LeadMagic | API-first mobile lookup with a shared credit pool | $99/month Essential; five credits per successful mobile | Search API request may cost an additional credit on lower tiers |
This table shows public terms on 27 September 2026, not negotiated quotes. The useful unit is not a website's “price per lookup” but a correct-person number your rep can use. The same vendor may be the cheapest source in the UK and the weakest in a different country. Neither company-wide record counts nor an undifferentiated phone percentage resolve that question.
1. ZoomInfo
ZoomInfo is a broad sales-intelligence purchase. Phone numbers are one field among contact search, company information and workflow features. That makes it a reasonable candidate for an established sales operation replacing several data subscriptions at once. It is less obvious as the first purchase for a small team that only needs a few hundred mobile numbers each month.
The distinction to test is mobile versus person-assigned direct dial versus switchboard in your own account list. Ask for a sample export with number type and country, then have reps validate an agreed sample before signing. Count known wrong-person and disconnected numbers, not only blank rows. A vendor can provide a phone on almost every record and still make a team spend most of its dialing time on routing menus.
ZoomInfo's public pricing path does not give a single phone-data price that can be responsibly compared with Datagma's or LeadMagic's self-serve plans. The quote should state seats, the exact contact-data product, mobile and direct-dial access, territory, export or API permissions, credits, renewals and any minimum spend. If your existing contract already includes most of those, the incremental cost of phone data could be attractive. If you need to buy a platform only to unlock phone fields, compare its full contract with a focused lookup layer.
Its non-fit is a buyer seeking a transparent, low-volume pay-per-found mobile bill. Its fit is a team prepared to test a large, repeatable slice of its actual market and use more of the platform than one phone column. Avoid the old article's unsupported claim that it is necessarily the volume winner for US mobiles or that it has a fixed $15,000 entry floor; neither is established for your quote.
2. Cognism
Cognism's current pricing page names Standard and Pro for sales prospecting, each with five seats included. Pro adds the premium mobile filter and on-demand mobile verification. Its separate phone-verified data page describes the proposition: reduce wasted dials by checking whether the number reaches the right decision-maker. The page is vendor positioning, so the buyer's own correct-person trial remains the proof.
The current pricing page says one credit reveals one contact and previously revealed contacts can be viewed again without another charge. It does not publish a simple Standard or Pro monthly amount. Request a quote for the actual number of users, contact reveals, mobile verification requests and territories. Standard now lists MCP and API access, using the shared credit pool; CRM Enrichment and custom Data-as-a-Service delivery remain separately scoped. Do not assume a five-seat package includes unlimited API or bulk delivery.
Cognism deserves priority when wrong-person calls are costly and verification of high-value contacts is more valuable than maximum raw coverage. A team calling senior people in several European countries may find that workflow compelling. Still, the quote and trial should tell you which records are phone-verified, which are merely present, how recent the verification is and what happens when a requested verification fails.
It is a weaker fit for a one-person team wanting a handful of phone lookups without a sales conversation, or for a developer expecting a self-serve, fixed-price API. Its old Diamond/Platinum language can also confuse procurement; buy against current Standard and Pro terms, and put the verification entitlement in writing rather than relying on an older tier name.
3. Lusha
Lusha is the most direct choice here for a rep who discovers a prospect, opens a profile and wants a number immediately. Its current credit documentation states one credit for an email and five credits for a phone number in the platform and API. Revealing both can therefore use up to six data credits. Search and viewing company information in the platform do not themselves consume reveal credits.
The API has additional rules that the old draft missed. Lusha's documentation says a single API request costs one credit per returned result and even an empty request has a one-credit minimum. Bulk requests charge one request credit per 1–25 results, on top of data-point credits. Thus “five credits per phone” is the field cost, not necessarily the entire API workflow cost. A batch of 100 phone reveals could use 500 phone credits plus request credits; calculate the actual call pattern before deciding an annual allowance is enough.
The same current document explicitly says every user has an API key by default, including on the Free plan, although an administrator can disable access. That corrects the old article's claim that Lusha API access requires an enterprise Scale tier. An API key's existence does not guarantee that the free plan's allowance, endpoint rights or rate limits will support a production job. Test those from the account you intend to buy.
Lusha's indexed pricing page shows a Pro annual-billing equivalent of $52.45 per month with 7,200 credits per year at the time of this review. The live plan card did not render in our browser reader, so confirm the checkout. At the five-credit field cost, 7,200 credits would be a theoretical maximum of 1,440 phone reveals if no other activity used credits; API request costs and mixed email work reduce that number. Do not present that theoretical count as guaranteed current plan capacity.
Lusha fits a small team that values quick rep-driven reveals and can control consumption. It fits less well when numerous empty API calls, repeated research tasks or non-phone credit use make the simple five-credit arithmetic misleading. During the pilot, export account-level usage by feature and reconcile it to the actual numbers accepted by reps.
4. Datagma
Datagma's Phone Finder is a narrower proposition: find a direct mobile from a professional profile or a full name plus company. The page says it does not return switchboards or desk direct lines and charges only when it finds a mobile. That makes it useful as a fallback after a primary contact database fails, or as the primary source for a team specifically seeking mobiles rather than a complete prospecting system.
The detailed phone-page FAQ says Regular $39/month for 100 phones, Popular $79 for 250 and Expert $209 for 750, with 12-month rollover for unused credits. Its visual card describes the $79 tier as “from $0.28/phone,” although $79 divided by 250 is $0.316; the $209/750 tier is about $0.279. The general pricing page also displays crossed-out prices and annual-billing language. Confirm the exact billing term and current checkout before quoting a unit price to a client. A found number can still be wrong for your prospect; pay-on-found is not pay-on-correct-person connection.
The vendor says its strongest match rates are in the US, UK and Western Europe, with coverage also in India and parts of APAC. It explicitly says mobile data are not available for contacts based in France. That exception matters more than a general “European coverage” label. If France is a key market, do not build a program assuming Datagma will fill that segment. Test each country and role separately. The site's advertised 35–45% match rate is a vendor claim, not a guarantee for a particular customer list.
Datagma supports an API and key-based integrations, so it can sit in a waterfall enrichment workflow without forcing a rep to manually search every missing person. Configure the call only for records whose mobile field is still missing. Log the original identifier, number type, country, credited success and later correct-person outcome; those last two are not always the same.
Its strongest fit is a measured pay-on-found mobile job with a manageable monthly volume and compatible geographies. It is less suitable if the team needs a full company-and-contact research suite or guaranteed French mobile coverage. Get the order terms and sample output before treating its front-page unit cost as final.
5. LeadMagic
LeadMagic combines several enrichment functions under one credit pool. The current pricing page lists Essential at $99/month with 5,000 monthly credits, Growth at $249 with 20,000, Professional at $499 with 50,000 and Ultimate at $849 with 100,000. All current plans include the enrichment API. Successful Mobile Finder results cost five credits; empty or failed mobile matches do not deduct those result credits. This is materially different from the old article's $49.99 entry claim.
For a phone-only job on Essential, 5,000 credits can theoretically fund 1,000 successful mobile results if no other operations consume credits. But an API workflow may need people or company searches first. LeadMagic's page says its search API costs one credit per request on Essential and Growth; the $499 Professional and $849 Ultimate tiers include unmetered search API subject to fair use and account-wide rates. In-app search has different rules. A developer must price the actual sequence, search, select, mobile lookup, validation and export, not just five credits times desired mobiles.
The pay-on-valid-result structure is attractive when the team already has reliable person identifiers and only needs to fill missing phones. If half the input list has no match, no mobile-result credits should be charged for those misses under the stated policy. Still, “valid result” is the vendor's billing definition; sample the output for wrong-person and stale-company errors. A successful result that reaches the wrong person is a procurement loss even if the API behaved as designed.
LeadMagic's broader pool can also pay for emails, company search and job-change operations. That helps a GTM engineer consolidate usage, but it means the phone budget competes with every other team using the account. Put per-workflow consumption reports and a monthly alert around the pool. Its public pricing page also states internal-use and fair-use restrictions for unmetered search; if you are reselling data or embedding a customer-facing lookup, seek explicit commercial terms.
LeadMagic best fits an API-first workflow with a defined phone-missing trigger. It is less compelling if a rep needs a polished browser-first prospecting interface as the main buying criterion, or if search request charges make a poorly filtered batch expensive.
What a phone number really costs
Suppose your team needs 500 accepted correct-person mobiles each month. “Accepted” means a current target employee, a number of the required type and country, and permission to use it in the planned workflow. That target is harder than “500 phone fields returned.”
The public numbers give only an initial floor. Datagma's $209 Expert tier advertises 750 found phones, so 500 found phones fit inside its stated allowance. LeadMagic Essential's 5,000 credits can fund up to 1,000 successful five-credit mobile lookups before other activities; its $99 subscription is not a promise that 500 accepted mobiles will be found. Lusha's five-credit phone reveal means 500 revealed phones use 2,500 field credits, plus any API request charges. Cognism and ZoomInfo require scoped quotes. The total bill is therefore incomparable until you measure each vendor on the same sample and the same acceptance rule.
Use this model: subscription + phone credits or overages + search/request charges + validation + review labor, divided by accepted correct-person phones. For example, if a $99 subscription produces 200 found numbers but only 120 pass review, the subscription-only cost is $0.83 per accepted number ($99 ÷ 120), not $0.50 per found number. That is an illustrative calculation, not a LeadMagic performance result. Add any request charges, other subscriptions and labor to get the true amount.
For a quote-led platform, calculate both the full contract and the incremental cost if you already own the platform. A sales organization that uses ZoomInfo or Cognism for other purposes should not assign every subscription dollar to phone data, but it also should not pretend mobile access is free. State the allocation method. Keep country and seniority segments visible so strong coverage for easy records does not mask a costly hole in your actual ICP.
A 200-person trial that exposes the difference
Select 200 named people from accounts your team actually pursues. Include 50 senior and 50 operational contacts in your largest market, then 100 across the smaller territories that often fail. Remove duplicates. Use the same names, current employer and profile identifiers for each vendor, and record any extra input step a provider requires. Make the test recent enough that your own ground-truth employment check is meaningful.
Create separate output columns for mobile, person-assigned direct dial, office and headquarters. Do not let a generic “phone” column pass. Independently check a reasonable sample by careful, permitted calling or another reliable method, recording wrong person, correct voicemail, correct live connection, disconnected and untested. A voicemail on the intended person's line can be useful but is not a live connection. Report both. Treat carrier or format checks as supplementary evidence, not proof of ownership.
Compare the accepted-number rate by market and role, the incremental numbers found by a second provider, the credits actually deducted and the number of wasted dials. Also record permission or suppression decisions relevant to your countries. If a vendor offers a phone-verified subset, score it apart from the unverified pool. This is the evidence a buyer needs to decide whether a higher quote buys more conversations or merely more displayed numbers.
The trial is small enough to complete without a large implementation, but it has a limit: 200 people cannot establish a statistically precise global ranking. Use it to reject a poor fit, expose plan mechanics and estimate the next purchase size. Once you have a primary source, test a second source only on missing accepted phones, rather than buying two full databases because their coverage claims look additive.
Which provider should you buy?
Buy ZoomInfo when phone data is part of a broader enterprise-data and workflow purchase, and its same-list trial justifies the quote. Buy Cognism Pro when on-demand phone verification and a five-seat prospecting workflow are material enough to warrant a sales-led contract. In both cases, negotiate against correct-person numbers in your territories, not a global record count.
Buy Lusha when reps need fast reveals and can monitor both field and API-request credits. Buy Datagma as a focused pay-on-found mobile source in supported markets, especially for missing records after another provider; verify the price card and exclude France-based contacts from any assumed coverage. Buy LeadMagic when a developer needs a result-based mobile API and can account for search calls and shared-pool use.
If your team needs a larger contact database rather than phone-only enrichment, our B2B contact databases for sales teams guide compares the broader purchase. The phone-data decision should still be won on numbers that reach the intended person.
Use the Clay Waterfall Enrichment Guide to separate primary phone coverage from incremental numbers found by a second provider.
FAQ
Is a direct dial the same as a mobile number?
No. A person-assigned direct dial routes through a company's phone system; a mobile is assigned to a handset or personal mobile service. Both can be valuable, but a headquarters or office line should not be counted as either. Ask for a number-type field and validate it in a sample.
Which provider has the best mobile coverage in Europe?
No public global claim establishes that for your list. Cognism sells a premium verification workflow; Datagma names strong European markets but explicitly excludes France-based mobiles. Test your specific countries, roles and company sizes on the same names before ranking them.
Does Lusha charge only five credits for an API phone lookup?
Five credits is the phone reveal cost. Lusha's current docs also describe API request charges, including a one-credit minimum for an empty request and bulk-result charges. The full API path can therefore cost more than five credits per accepted number.
Does Lusha require an enterprise plan for API access?
Lusha's current public FAQ places API access on Premium and Scale, so those are the purchase tiers to budget. Developer documentation describing default API keys, including free accounts, does not establish access to every endpoint. Price the intended phone endpoint, request charges and field credits in that qualifying plan.
What happens when Datagma or LeadMagic finds no number?
Both vendors state that their phone-result credits are charged on a successful found result, not an empty mobile lookup. LeadMagic may still charge for a separate search API request on lower tiers. A found number also needs correct-person review; “billable match” and “usable phone” are different outcomes.
Should we buy two phone-data providers?
Only if the second finds enough additional accepted numbers among the first provider's misses to justify its cost and complexity. Measure overlap on the same people, then trigger a waterfall enrichment workflow only for records still missing the required phone type.
