A 4 percent connect rate: is it the data or the time of day?
By Romain Eliard, Co-Founder of Listar. Updated .
You cannot know without sorting your call outcomes. An overall connect rate tells you nothing, because it mixes five causes with different fixes: invalid number, wrong person, call filtered before it rings, right person unavailable, right person reached and declining. Until those five lines are separated, changing calling hours is a gamble.
Key points
- Timing only explains the fourth category. If most of your failures sit elsewhere, moving the calls changes nothing.
- A call that never rings is not a missed call: it is either a dead line or carrier filtering. Both are fixable, with different remedies.
- A connect rate only means something alongside the number of attempts per contact. 4 percent on first attempt and 4 percent after six describe different problems.
- Caller ID reputation acts before the data and before the hour: a flagged number stops ringing for part of your list.
- The diagnosis takes two weeks and costs one extra column in the CRM.
The five categories to separate
Sort every attempt into exactly one of these: 1, invalid or disconnected number; 2, answered by someone other than your target; 3, rings with no answer at any hour, a sign the call is filtered; 4, right person reachable but unavailable; 5, right person reached, who accepts or declines. A rep takes two seconds to fill that column. A fortnight of calling is enough for a usable distribution, provided attempts are spread across several time slots. Right people reached over numbers dialled is your connect rate.
Category 1: invalid numbers
If more than one attempt in ten hits a disconnected number, the problem is source freshness, not timing. A carrier lookup before the campaign removes those lines in one query, with nothing visible to the recipient. It is the cheapest and most immediate fix. See how to check a mobile is still active.
Category 2: the wrong person
If your reps often land on a switchboard, a namesake or the previous holder of the role, the number was valid but wrongly attached. This is the most expensive defect, because it consumes a full call and sometimes a meeting. The fix is to require that the provider checks the match to the person, not only the line. See phone ownership check.
Category 3: the call filtered before it rings
This is the most commonly missed cause. Carriers and handsets apply spam filters that stop some calling numbers from ringing at all. The symptom is distinctive: an unusually high rings-no-answer rate, even across every time slot. The test is to redial a sample from a different calling number and compare.
Category 4: the right person at the wrong moment
This is the only category timing explains. You recognise it by an uneven distribution across slots and days. If this is your main case, optimising hours pays, and it is measured by comparing connect rates in two-hour bands rather than following a general rule found online: the right slot depends on the role you are calling.
How to read the distribution
| What the distribution shows | Likely cause | What confirms it | The fix |
|---|---|---|---|
| Many disconnected numbers | Stale source | Carrier lookup on the sample | Verify before calling |
| Many wrong people | Unverified match | Names obtained do not match | Require a match check |
| Rings with no answer at every hour | Caller ID filtering | Test from another calling number | Change and warm the calling number |
| Rings with no answer, varying by hour | Availability | Clear gaps between time slots | Redeploy the slots |
| People reached but refusing at once | Targeting or opener | The refusal reason recurs | Revisit the target or the script |
How we test
This diagnosis rests on classifying call outcomes row by row, over at least two weeks and several time slots. Rates published by providers do not replace this measurement: see what vendors leave out of their rates.
Where Listar fits
Listar acts on categories 1 and 2, not on the others: the search queries around thirty providers in a waterfall, then applies four verification layers including WhatsApp to drop dead lines and confirm the number matches the right person. No commitment, one credit equals one euro. If your distribution mostly shows caller filtering or an hours problem, changing data provider will fix nothing: it is your telephony and your organisation to look at. Before switching tools, measure on a sample. Listar is published by the author of this page.
FAQ
What is a normal B2B connect rate? There is no usable benchmark, because the figure depends on the role you call, the number of attempts and the quality of the numbers. Compare yourself with your own history, not with a published average.
How many attempts per contact? Enough that a one-off absence is not counted as a failure, varying the time slots. A rate computed on a single attempt mostly measures luck.
How do I know whether my caller ID is filtered? Redial a sample from another number and compare the rings-no-answer rate. A large gap points to a reputation problem on the calling number.
Should you display a local number? It often improves connects, but a local number used at volume gets flagged like any other. The effect lasts until its reputation degrades.
Does the 11 August 2026 French law affect B2B calls? No. The prior consent introduced on that date covers prospecting towards consumers. Calls to professionals about their work remain possible with information and a right to object.