Best time to make cold calls: what the studies say and how to test it

There is no single best time to make cold calls. Most studies point to late morning and late afternoon in the middle of the week, but they disagree on the details. A 2007 study by InsideSales.com and MIT found 4 to 6 pm best for reaching leads, while PhoneBurner's 2021 data put 10 am on top. Use the studies as a starting point. Then run a simple test on your own list, because your prospects are the only ones whose habits count.

What the studies say

Here are four sources we could check, side by side. Each one measured something slightly different, which matters later.

The 2007 MIT study

This is the most cited one. It found that 4 to 6 pm was 114% better for reaching a lead than 11 am to noon, and that 8 to 9 am was 164% better for qualifying a lead than 1 to 2 pm, just after lunch. Thursday was 49.7% better than Tuesday, the worst day, for making contact. The same study is the source of the famous five-minute rule for speed to lead.

Keep its limits in mind. It is nearly twenty years old, it covers only six US companies, and it is about people who had just filled in a web form. Those are not cold prospects. It is also a vendor's own summary: the link goes to a PDF copy hosted by a third party, because the original page is no longer online.

The dialer datasets

PhoneBurner analysed the calls its users made in 2021. The hour from 10 to 11 am did best, and the window from 10 am to 3 pm was strong overall. Tuesday won, but PhoneBurner itself says the margin was slim and that Monday and Wednesday were nearly tied. Yesware looked at which calls lasted over five minutes in its own dialer data, and those clustered between 3 and 5 pm on Tuesdays and Thursdays.

The Cognism report uses data from WHAM, a calling agency. Tuesday had the most booked meetings. Calls lasted longest between 10 and 11 am. Cognism notes that the best time for WHAM may not be the best time for you.

Why the studies disagree

Four studies, four different answers. That is not because one is right and the others are wrong. They simply measure different things in different places.

• Different outcomes. Reaching someone, qualifying them, a long call and a booked meeting are four different things. The best hour for one is not automatically the best hour for another.
• Different samples. Web leads in 2007, dialer users in 2021, an agency in 2025. Each has its own mix of markets and roles.
• Different time zones. PhoneBurner reports in US Central Time. A study that does not convert to the prospect's local time blurs the pattern.
• Vendor data. Most studies come from companies that sell calling tools and analyse their own users. That is useful, but it is not a neutral sample.
• Where people call most. If most callers call in the morning, the morning has the most data. That does not prove the morning is better.
• Small real differences. PhoneBurner itself calls some of its differences small. Time of day is a smaller lever than list quality, your opening and how often you try.

So what is left? A rough pattern: late morning and late afternoon, midweek, and avoid the hour right after lunch. That is a good starting point, nothing more.

Think about your own prospects first

Before you test, picture a normal day for the person you are calling. This costs nothing and often explains more than any study.

• A contractor or installer is on site during the day. Early morning or late afternoon may work better.
• A shop or restaurant owner is busy when customers are in. Try quiet hours.
• An office manager or finance lead has a desk day. Mid-morning is often fine.
• A CEO of a small company may pick up before the team arrives or after it leaves.

Also watch the calendar. Lunch, the weekly team meeting, school holidays and industry seasons all shape when people answer. These are hypotheses, not facts. The test below tells you which ones hold.

How to test your own best time to call

You do not need special software. You need two campaigns, one outcome per call and a few weeks of patience.

1. Pick one question. For example: morning (9 to 11) or afternoon (15 to 17)? Test one thing at a time.
2. Split one list at random. Take one list of similar leads and split it in two, for example by putting every other row in a second file. Do not split by region or sector, or you test the list instead of the time.
3. Make two campaigns. Call one only in the morning block and the other only in the afternoon block.
4. Swap callers halfway. If one caller does mornings and another does afternoons, you are also comparing people. Swap after two weeks, or let the same people do both.
5. Log one outcome per call. Reached or not, conversation, meeting. Same definitions for everyone.
6. Compare connect rates per campaign. Connect rate = conversations ÷ call attempts. Also look at meetings per attempt, since an hour can be good for reaching people but bad for booking.
7. Keep the winner and test the next question. Then try Tuesday versus Thursday, or an early block versus your normal block.

How many calls before you trust the result?

More than you think. Here is a rough guide, worked out with standard statistics. If your connect rate is around 20% and each campaign has about 300 attempts, a difference smaller than about 6.5 percentage points can easily be chance. With 100 attempts each, even a 10-point gap can be luck. So keep going until each campaign has a few hundred attempts, and be careful with any result based on a single week.

Worked example

The numbers below are an illustration, not a benchmark.

In this example, the afternoon gap of 8 points is larger than the noise margin, and it also brings more meetings. The team moves its main calling block to the afternoon and keeps the morning for call-backs and email. A month later, it tests the next question.

If you want to see how a higher connect rate changes your whole funnel, use the formula in how many cold calls it takes to get one customer. For the definitions of each metric, see cold calling KPIs.

Doing this in Ringlane

Ringlane has no reports by hour of the day or day of the week. You can still run this test. Import the two halves of your list as two campaigns, call each one in its own time block, and log one outcome per call with a single key. The reports in Ringlane show call attempts, connect rate and meetings per campaign and per caller, so you can put the morning and afternoon campaign side by side. Protecting those time blocks is the other half of the work: see call blocks for focus.

Common mistakes

• Copying a study's best hour without testing. A 2007 US result is a guess for your market, not an answer.
• Testing two things at once. A new list and a new hour in the same week tells you nothing about the hour.
• Comparing callers instead of times. Swap people halfway or let the same people do both blocks.
• Stopping after a few days. Small samples swing wildly.
• Only calling at the "best" time. For repeat attempts on the same lead, vary the hour. If you always call at ten, you may hit the same weekly meeting every time.

Frequently asked questions

Sources

1. InsideSales.com and James Oldroyd: Lead Response Management Study (2007, vendor study, third-party PDF copy)
2. PhoneBurner: Best time of day for sales calls (2021 data)
3. Cognism: The State of Cold Calling in 2025 (with WHAM data)
4. Yesware: Reignite your cold outreach strategy