What Is a Good Connect Rate for Cold Calling? Why 5-10% Is Normal
The median B2B sales development rep places 44 phone dials a day and gets 4.1 quality conversations out of them, according to The Bridge Group's 2025 SDR Models, Motions & Metrics report, which surveyed 351 B2B companies. That's roughly 9 percent. Meanwhile Cognism puts the average cold calling success rate, meaning dials that end in a booked meeting, at 2.3 percent. Those two numbers describe the same activity and they're eight times apart, because they measure different things.
That gap is the whole problem with connect rate benchmarking. Before you can decide whether your team's number is good, you have to know exactly what sits above and below the line. This guide gives real ranges, names the sources, and shows why two honest teams can report wildly different connect rates while doing identical work.
Key Takeaways
- Median dial-to-conversation sits near 9%: 44 dials for 4.1 quality conversations (Bridge Group, 2025).
- Connect rate is a ratio, so a bad denominator inflates or destroys it.
- Phone data decays 20-25% a year (ZoomInfo), which depresses connect rate before a rep speaks.
- 80% of people ignore unknown numbers (Pew Research).
What Counts as a Good Connect Rate in 2026?
For most B2B outbound teams, a dial-to-conversation rate between 5 and 10 percent is normal, and the 9 percent implied by Bridge Group's 2025 data (44 phone dials, 4.1 quality conversations per day) is a fair median anchor. Anything above 12 percent usually signals verified mobile data, tight targeting, or a generous definition of "connect."
The reason nobody can give you one clean number is that connect rate is a ratio, and ratios are easy to move without improving anything. Dial a smaller, cleaner list and your rate climbs. Dial a bigger, older list and it collapses. Neither change says much about how well your reps sell.
The published benchmarks, side by side
| Metric | Median / average | Source |
|---|---|---|
| Phone dials per SDR per day | 44 | Bridge Group, 2025 |
| Quality conversations per day | 4.1 | Bridge Group, 2025 |
| Dials per day, phone-centric teams | 56 | Bridge Group, 2025 |
| Conversations per day, phone-centric teams | 4.6 | Bridge Group, 2025 |
| Conversations per day, email-centric teams | 3.4 | Bridge Group, 2025 |
| Cold call success rate (meeting booked per dial) | 2.3% | Cognism, 2025 |
| Attempts needed to reach a prospect | 3 | Cognism |
| Average cold call duration | 93 seconds | Cognism |
Notice that phone-centric teams dial 27 percent more than the median and only convert about 8 percent of those dials into conversations, slightly below the blended figure. Volume alone doesn't buy you a better ratio.
What "good" actually means
Good means your connect rate is high enough to hit conversation targets at a cost per conversation you can defend, and stable enough that you can tell when something breaks. A team holding 7 percent for six months is in better shape than one that swings between 4 and 14 percent depending on which list got loaded.
What Exactly Is the Denominator in "Connect Rate"?
This is where most benchmarking falls apart. If you count every dial attempt, including redials to the same contact, your denominator inflates fast: Cognism reports it takes an average of three attempts to reach a lead, and 93 percent of conversations have happened by the third call. Count attempts and you divide by three. Count unique contacts and you don't.
Both are legitimate. Only one is comparable to a published benchmark. Before you compare yourself to anything, write down which of these you're using.
The three common denominators
Total dial attempts. Every time the system placed a call, including retries, voicemail drops, and calls that never reached the network. This is the most conservative and the most common in dialer reporting.
Unique contacts touched. One entry per person, no matter how many attempts. This roughly triples the apparent rate versus attempts.
Completed call attempts. Attempts that actually reached a carrier and rang, excluding disconnected numbers and network failures. This is the fairest measure of rep effectiveness, and the rarest, because most teams can't isolate it.
Why the numerator is just as slippery
Does a five second "wrong person, sorry" count as a connect? Does a gatekeeper? Does a voicemail system that picked up on ring two? Teams that count any answered audio as a connect will report double what teams counting qualified conversations report. Bridge Group's 4.1 figure is explicitly quality conversations, which is a stricter bar than "somebody said hello."
Pick one definition, write it into your reporting spec, and stop changing it. A benchmark you can't reproduce next quarter is a vanity number.
Dial-to-Connect, Connect-to-Conversation, Conversation-to-Meeting: What's the Difference?
These are three separate ratios and they fail for three separate reasons. Dial-to-connect is a data and deliverability problem. Connect-to-conversation is an opening problem. Conversation-to-meeting is a message-market fit problem. Gong's research found that stating your reason for calling boosts success 2.1x, which affects only the third ratio.
Blending them into one "connect rate" hides which stage is actually broken.
Stage one: does the phone ring, and does a person pick up?
This stage is decided before the rep speaks. Is the number live? Does it belong to the person on the record? Is your caller ID trusted enough to survive the screen? Pew Research found that 80 percent of Americans don't generally answer cellphone calls from unknown numbers, and only 19 percent say they do. No script fixes that.
Stage two: does the person stay on the line?
Here the opening matters enormously. Gong found that asking "Did I catch you at a bad time?" reduces the odds of booking a meeting by 40 percent. Successful cold calls also run nearly twice as long as unsuccessful ones, with reps talking about 55 percent of the time.
Stage three: does the conversation become a meeting
Cognism puts the average at 2.3 percent of dials, with their own team reporting 6.7 percent using verified data and a structured methodology. Against a 9 percent connect rate, a 2.3 percent dial-to-meeting rate implies roughly a quarter of conversations turn into something booked.
Track all three. When a manager says "our connect rate dropped," the follow-up question is always: which one?
How Much Do List Quality and Bad Phone Data Move the Number?
A lot, and mostly invisibly. ZoomInfo reports phone numbers decay 20-25 percent per year and job titles 25-35 percent, with aggregate B2B database accuracy loss running anywhere from 22.5 to 70 percent annually depending on the field. A list bought 12 months ago is a materially different asset than the one you paid for.
The cruel part is the asymmetry. A bad email bounces and tells you so. A dead phone number just rings, or rings somewhere else, and consumes a dial slot exactly like a good one. Your denominator grows while your numerator doesn't.
Modelling the effect
The table below is an illustrative model, not measured data. It applies the sourced decay rates above to a hypothetical list to show how a stable underlying pickup rate produces very different reported connect rates.
| List condition (illustrative) | Assumed unreachable share | Reported dial-to-connect at a constant 11% true pickup |
|---|---|---|
| Freshly verified | 5% | 10.5% |
| 6 months old | 12% | 9.7% |
| 12 months old | 22% | 8.6% |
| 24 months old, never refreshed | 40% | 6.6% |
Nothing about rep skill changed across those rows. The reported number fell by a third anyway. If your board is comparing this quarter to last, list age alone can explain the move.
Verification as a continuous process
Batch cleaning helps, but it expires. The stronger signal is the dial itself, because the attempt already touched the number. This is why platforms like Personnect built verification into the dialing workflow rather than bolting it on beforehand, treating every attempt as a data event even when they don't pick up. Their positioning, "Every Call Counts," follows directly from that idea.
The operational question is where the result lands. Verification that stays in a vendor dashboard is a report. Verification written back into the CRM changes what the next rep dials.
Why Does Spam-Flagged Caller ID Depress Connect Rate?
Because the call never gets a fair hearing. Caller ID Reputation's 2025 report logged 14.5 million caller ID flag events in 2025 alone and 45 million since 2021, with roughly half of 2025's flags coming from consumers manually marking calls as spam. Once a number carries a label, pickup collapses regardless of who's dialing or what they'd have said.
Combine that with Pew's 80 percent screening rate and you get a structural ceiling that has nothing to do with rep performance.
How numbers get flagged
Flags accumulate from complaint volume, short call durations, high dial velocity from a single number, and calls to numbers that never answer. Notice that last one: dialing a decayed list actively damages the reputation of the number you dial from. Bad data and spam labels feed each other.
What teams do about it
The common playbook is number rotation, carrier registration, monitoring reputation across the network, and keeping dial velocity per number inside sane limits. Personnect's approach here is a dedicated-number model, so one team's dialing behaviour can't affect another's caller ID and callbacks route back to the team that placed the call.
Separate the two problems
Verification cleans the numbers you dial to. Reputation management protects the number you dial from. They're different fixes and solving one does nothing for the other. Teams that buy a data vendor and expect their spam labels to disappear are disappointed every time.
How Do Benchmarks Change by Seniority, Industry, and Time of Day?
Seniority moves it more than most people expect, and in the opposite direction to intuition. RAIN Group's research across 488 buyers and 489 sellers found 57 percent of C-level and VP buyers prefer to be contacted by phone, versus 51 percent of directors and 47 percent of managers. Senior people are harder to reach, but more receptive when you do.
Seniority
Executives have gatekeepers, packed calendars, and aggressive screening. Expect a lower dial-to-connect rate targeting VPs than targeting individual contributors. Expect a higher conversation-to-meeting rate. If you're comparing two SDR pods calling different levels, connect rate alone will make the enterprise pod look lazy.
Industry and list source
Fast-moving sectors churn staff faster, so their data decays faster. A technology list ages harder than a manufacturing list. Similarly, inbound-sourced and event-sourced numbers usually outperform purchased lists, because someone typed them recently.
Time of day
Cognism reports the strongest windows fall between 10 and 11 in the morning and 2 to 3 in the afternoon, with 7 to 9am, the lunch hour, and 5pm performing worst. Their data also puts Tuesday ahead of Friday for booked meetings. If your team dials whenever, you're volunteering a few points of connect rate.
Territory coverage compounds this. A rep in one time zone calling a national list spends part of the day dialing into dead hours. Matching dialing windows to the prospect's clock rather than the rep's is one of the cheapest gains available, and it's why Personnect surfaces where contacts actually sit geographically before a calling block starts.
How Should You Set a Connect Rate Target for Your Team?
Start from your own baseline, not from a published number. Measure four weeks with a fixed denominator definition, then set a target 15-20 percent above that baseline. Given the median sits near 9 percent (Bridge Group, 2025), a team at 6 percent should aim for 7, not for someone else's 14.
Segment before you compare
Split by target seniority, list source, and territory. A blended company-wide connect rate averages away the only signal you'd act on. Report the segments separately and the diagnosis usually writes itself.
Watch bad-number rate as a leading indicator
Bad-number rate, meaning disconnected lines and wrong-person outcomes as a share of attempts, moves before connect rate does. When it climbs, your list is aging or your source degraded. It's a much cleaner management metric than the ratio it feeds.
Pair it with cost per conversation
Connect rate on its own can be improved by dialing less and dialing better, which is usually right, or by cherry-picking a small list, which is usually wrong. Cost per conversation catches the second case. With Salesforce reporting that reps spend under 30 percent of their time actually selling, the real currency is conversations per rep hour, not the percentage next to it.
Frequently Asked Questions
What is the average cold call connect rate in 2026?
Roughly 9 percent on a dial-to-conversation basis, derived from Bridge Group's 2025 data of 44 phone dials producing 4.1 quality conversations per day. Most teams land between 5 and 10 percent. Higher figures usually reflect verified mobile numbers or a looser definition of "connect."
How many dials does it take to book one meeting?
Cognism puts the average cold call success rate at 2.3 percent, which works out to roughly 43 dials per booked meeting. Their own team reports 6.7 percent, closer to 15 dials. The spread comes mostly from data quality and targeting, not scripting.
Does calling more improve my connect rate?
No. More dials raise conversation volume but leave the ratio flat or lower it, since extra volume usually comes from lower-quality list segments. Bridge Group's phone-centric teams dial 56 a day and still convert about 8 percent, slightly under the 9 percent blended median.
Can I verify phone numbers without calling them?
Yes, through database lookups, but those results age quickly given 20-25 percent annual phone decay. The dial attempt itself is the strongest signal. Personnect works on this principle, returning verified data from calls that go unanswered and syncing the result back to the CRM.
Why did our connect rate drop without any change in staffing?
Check list age and caller ID reputation first. Both degrade silently. With 14.5 million caller ID flag events logged in 2025, a newly labelled number can cut pickup overnight while every activity metric on the dashboard looks completely normal.
The Bottom Line
A good connect rate for cold calling is somewhere between 5 and 10 percent, with about 9 percent as the honest median. But that range is only useful once you've defined the denominator, separated dial-to-connect from connect-to-conversation from conversation-to-meeting, and accounted for the two forces that move the number without anyone touching a script: decaying phone data and flagged caller ID.
Those two account for more variance than coaching does. A list that's 22 percent unreachable and a number carrying a spam label will hold a strong team below a mediocre benchmark indefinitely, and no amount of activity coaching will surface the reason.
So measure your own baseline, segment it, and watch bad-number rate and cost per conversation alongside the ratio. Then improve against yourself. Comparing a blended company number to a published median is how teams end up solving the wrong problem for a full quarter.
