A list of 500 contacts can hide three very different lists. Some numbers ring in the prospect’s pocket, some ring on a desk that may be empty, and some ring at reception, where a person or a menu decides whether you get through. Most teams measure one global connect rate and conclude that a list is good or bad. The real driver sits one level below: the mix of number types. This page defines each type, shows how to recognize it in the exports of the main data providers, and gathers the connect rates published for each one, with the source, the date and the limits of every figure.
Four kinds of numbers, four different calls
Before any rate, a definition. Sales teams use the words loosely, and data providers label the same number differently. Here is what each term means when you are about to dial.
A fifth category cuts across the others: VoIP and virtual numbers. Many companies now give every employee a number from a cloud phone system. That number behaves like a direct dial, and some apps forward it to a mobile. But the VoIP family also includes generic online numbers, call tracking numbers used in ads and lines that nobody monitors. A line-type lookup labels them VoIP; only a call tells you whether a person answers.
The distinction between your own number and the prospect’s number matters too. This page is about the number you dial. Which number you call from, mobile or landline, local or national, is a separate question with its own data; our French guide on choosing a mobile or landline caller ID covers that side.
How to spot each type in a data export
Data providers do not use the same column names, and the label is a hint, not a guarantee. Three of the most common exports, read the way a sales ops lead would before importing them.
Apollo.io
- Mobile
- Mobile Phone
- Direct dial
- Work Direct Phone
- Switchboard
- Corporate Phone, Company Phone
- Other columns
- Home Phone, Other Phone; phone values often start with an apostrophe
Lusha
- Phone columns
- Phone 1, Phone 2
- Type
- Phone 1 type, Phone 2 type
- Watch out
- The values of the type column are not documented; check a sample
ZoomInfo
- Mobile
- Mobile phone
- Direct dial
- Direct Phone Number
- Switchboard
- Company HQ Phone
- Extra
- Contact Accuracy Score and Grade per record
Column names come from Lusha’s export documentation and from real Apollo and ZoomInfo files published on GitHub, checked in November 2026. Apollo lets each user choose the export columns, so your file may differ. Other providers (Cognism, Kaspr, FullEnrich, Surfe) do not publish a stable list.
When the column is missing or unreliable, four checks classify a number without calling it:
- 01
Read the numbering plan:
In most of Europe a mobile has its own prefix: 06 and 07 in France, 07 in the UK, 015, 016 and 017 in Germany. The number itself tells you. In the US and Canada mobiles and landlines share the same area codes, so the plan says nothing and you need a carrier lookup. - 02
Count duplicates by company:
a number that appears on three or more contacts at the same company is the switchboard, even in a column called Direct Phone. Two contacts sharing a mobile is usually a job change or a data error, not a switchboard. - 03
Compare with the company phone:
if the contact’s number equals the company’s main number, it is the switchboard. Toll-free and premium-rate prefixes (800, 0800, 0870) almost always are too. - 04
Look at the ending:
a number ending in 000 or 100 is often the head of a block that a company routes to reception. It is a weak signal: some real direct dials end that way too.
Our direct dial checker runs these rules on a whole list in the browser, shows the reason behind each label and sorts the file into a calling order. Fix the format first with the phone number validator so that duplicates written differently are caught.
Connect rates by number type: what is published
There is no neutral, large, recent study that measures connect rates by number type across countries. What exists is a handful of vendor figures, each with its own method. Read together, they agree on the shape and disagree on the numbers.
| Source | Mobile | Direct dial | Switchboard | What it is |
|---|---|---|---|---|
| SMARTe Updated October 26, 2026 | 11 to 14% (verified) | Not published | 3 to 5% | Vendor of mobile data; survey of 1,200+ sales pros, attribution of these rates unclear |
| Octave March 16, 2026 | 5 to 10% average, 10 to 15% good | 1 to 6% by segment | 1 to 2% average | Benchmark table with no stated data source |
| ScaleX, via SalesIntel Early 2020 | 10 dials per conversation | 73 dials per conversation | 59 dials per conversation | Small experiment on VP and C-level targets, vendor-published |
| Gong Labs July 2024, updated May 2026 | Not split | Not split | Not split | 5.4% average, 13.3% top quartile, all numbers mixed |
Three readings come out of this table.
Mobiles win everywhere. Whether the source is a vendor survey, a benchmark grid or a controlled test, mobiles connect two to seven times as often as the other types. The reason is physical: the phone is on the person.
Direct dials do not always beat switchboards. Octave places direct dials three to five times above switchboards. The ScaleX test found the opposite: 73 dials per conversation on direct dials, 59 on switchboards. The test is small and dates from early 2020, just as offices were emptying, which is exactly the situation in which a desk line rings for nobody. The honest conclusion is that a direct dial is worth what the prospect’s working habits are worth. For a plant manager on site every day it is excellent; for a remote sales director it can be worse than reception.
Averages hide the mix. Gong’s 5.4% across more than 300 million calls is the most robust figure of the lot, and it says nothing about number types. A list that is 60% switchboards and a list that is 60% mobiles can both “average” 5% if the first one is better targeted. That is why the type mix has to be measured before the rate is judged.
A direct dial is worth what the prospect’s working habits are worth: excellent for someone at their desk every day, worse than reception for someone who never is.
Two vendor claims circulate widely and deserve a word. The statement that direct dials are “147% more likely” to reach a VP is repeated on many blogs; we could not trace it to a published study, and the sources that cite it attribute it to different origins. Cognism’s claim of 1.55 dials per connect with its verified mobiles comes from the vendor and has no published method. Cognism’s own Redgate case study is more modest and more useful: the connect rate rose from 8% to 11 to 12% after switching to its data. For the general US benchmark, read our cold call connect rate benchmarks; this page does not repeat its ranges.
The rates to plan with, and how many dials each type costs
To size a campaign, you need one working number per type. Ours are deliberately prudent and sit inside the published ranges. They are the default rates of our direct dial checker and of the phone enrichment pricing tool, so every calculation on the site uses the same base.
| Number type | Planning connect rate | Dials per conversation | Published range (sources above) |
|---|---|---|---|
| Mobile verified or recent | 14% | about 7 | 5 to 15%; up to 22% on verified mobile lists |
| Direct dial landline or VoIP DID | 8% | about 13 | 1 to 12% depending on segment and remote work |
| Switchboard main number, reception, IVR | 3% | about 33 | 1 to 6% |
| Unknown VoIP generic or virtual number | Measure first | Unknown | No published figure |
Dials per conversation is simply 100 divided by the connect rate. It is the figure that turns a list into hours. With these rates, reaching 100 people takes about 700 dials on mobiles, 1,250 on direct dials and 3,300 on switchboards.
Why the gap between mobiles and desk lines keeps widening
The published figures span six years, and the direction is the same in all of them: the share of conversations that happen on a mobile grows, the share that happens on a desk line shrinks. Four forces push in that direction, and none of them is about to reverse.
- 01
Hybrid work emptied the desks:
a direct dial rings where the person sits on office days. With two or three days at home, the same line reaches its owner less than half of the working week, unless the company forwards it to a softphone or a mobile. - 02
Cloud phone systems blurred the line:
when a direct dial is a VoIP number that also rings in an app on the person’s phone, it behaves like a mobile. That is why some direct dials connect very well and others not at all; the label no longer describes the device. - 03
Reception got automated:
many switchboards are now IVR menus or virtual receptionists. They pick up every call, which inflates pickup rates, and route almost none to a named decision maker without an extension. - 04
Providers invested in mobiles:
data vendors compete on mobile coverage because buyers judge them on it. The supply of verified mobiles keeps growing, especially in Europe, which makes a mobile-first list easier to build than it was in 2020.
The counterweight is screening. Mobiles are where operating systems and carriers filter unknown callers most aggressively, and a flagged caller ID cancels part of the mobile advantage. The practical consequence is simple: the more mobiles a list contains, the more your caller reputation matters. A team that moves to a mobile-heavy list should check its numbers’ reputation at the same time, and rotate them if the volume per number gets high; the phone number rotation calculator gives the number of lines a given call volume needs.
Country and time of day change the picture
Number types do not behave the same everywhere. Three effects show up in every team that splits its results by country.
Plan by local time rather than your own. The when to call tool shows each prospect’s local time and the windows that tend to work for desks and for mobiles, and our best time to cold call guide gathers the published studies by day and hour. The cold calling cell phones guide covers screening and carrier lookups in more depth, including what a line-type lookup can and cannot tell you.
What the mix changes in practice
Once you know the share of each type in a list, three decisions follow almost mechanically.
| Your list | Calling order | Dialing mode | Data to buy |
|---|---|---|---|
| Mostly mobiles 60% or more | Mobiles first, the rest in a second pass | Power or parallel dialer | Little; refresh dead numbers |
| Mixed a third of each | Mobiles, then direct dials, switchboards last | Power dialer, switchboards by hand | Mobiles for the best-fit accounts |
| Mostly switchboards 50% or more | Gatekeeper session to collect names and lines | Manual, with a script | A mobile-first provider before calling |
The calling order. Dial mobiles first, while the list is fresh and the rep’s energy is high, then direct dials, then switchboards in a separate session with a gatekeeper script. Mixing them in one session is the most common waste: the rep keeps changing gear between a prospect who answers personally and a receptionist who asks who is calling.
The provider. A list that comes back mostly with switchboards says something about the provider’s coverage for your market, not about cold calling. Compare what each one charges per mobile found with the phone enrichment pricing tool, check coverage by country, and read our B2B phone number enrichment guide for the test protocol on 100 contacts.
The dialing mode. On a mobile-heavy list, the limit is the time between rings, so automation pays: a power dialer chains the calls, and a parallel dialer rings several numbers at once and connects the rep to whoever answers first. On a switchboard-heavy list, automation mostly multiplies receptionists; manual dialing with a script is often better. Our comparison of parallel and power dialers covers the trade-offs.
The traps that make a list look better than it is
A column called “Direct Phone” filled on 90% of the rows looks like a great list. Five traps explain why it may not be.
Before the first session: classify the list, count duplicates by company, keep extensions separate, check your caller ID, and plan mobiles first.
Measuring your own rates by number type
Published rates are a starting point. After two weeks of calling, your own figures are worth more than any benchmark, provided you log them properly.
- 01
Tag every number with its type before calling:
mobile, direct dial, switchboard or unknown, using the rules above. Without the tag, the dialer cannot split the results. - 02
Log three outcomes, not one:
no answer, answered by someone else (reception, colleague, voicemail), and conversation with the right person. The second outcome is where switchboards hide. - 03
Keep the time slot and the country:
a rate by type that mixes 9 a.m. and 4 p.m., or France and the US, will mislead you. Two weeks is usually enough to see a stable picture per slot. - 04
Compute dials per conversation per type:
then cost per conversation, by dividing the cost of the data and of the rep’s time by the number of conversations. That is the figure to compare providers on. - 05
Feed the results back to the data:
numbers that never connect after three attempts at different times are probably wrong. Flag them, ask the provider about refunds where its policy allows, and refresh the list.
How Skipcall fits this workflow
Skipcall is a calling platform: it does not sell contact lists and has no contact database. Its built-in enrichment completes the contacts you already have, imported by CSV or synced from your CRM, so you can call them. Once the numbers are in, the parallel dialer calls up to four of them at once and connects the rep to whoever picks up, and each outcome is logged on the contact record.
Questions to ask a data provider about number types
Before you buy, five questions tell you more than a coverage percentage.
- 01
How do you label a mobile?
By numbering plan, by carrier lookup or by declaration? In North America only a carrier lookup is reliable. - 02
Can a direct-dial field contain a switchboard?
Ask what the provider puts in the field when it has no personal line, and whether it flags shared numbers. - 03
How old is the number?
Ask for a last-verified date per record. A mobile verified this quarter is worth more than a direct dial verified two years ago. - 04
What do I pay when the number is wrong?
Refund policies for wrong numbers vary from none to automatic credit returns. Get the rule in writing and keep your call logs as proof. - 05
What is your mobile share in my country?
Coverage differs sharply from the US to Europe. Run a test on 100 contacts from your real target before signing.
To go further on vocabulary, our cold calling glossary defines direct dials, switchboards, HLR lookups, waterfall enrichment and do-not-call lists. To find the mobile of one specific decision maker, read our guide on how to find a cell phone number by name, and for the files themselves, our guide to B2B lead lists for cold calling explains why a calling list needs direct dials and mobiles rather than switchboards.
Sources, checked in November 2026: SMARTe, B2B prospecting statistics (updated October 26, 2026); Octave, GTM engineer’s guide to connect rates (March 16, 2026); ScaleX experiment published by SalesIntel (2020 case study PDF) and cited by Prospeo; Gong Labs, insights from 300 million cold calls (July 2024, updated May 2026); Cognism, Redgate case study; Lusha export documentation; Apollo and ZoomInfo export columns observed in public files. This page gives general information, not legal advice.
