Why Your Contact Rate Is Low and How to Tell Which Cause It Is

Contact rate is unique records reached by a human divided by records attempted — 430 of 2,000 records is 21.5%. Five causes explain almost every low contact rate: too few attempts, the wrong calling hours, a single-channel sequence, invalid data, and a caller ID with a reputation problem. A 200-record redial control tells you which one.

At a glance

• The formula: contact rate = unique records that reached a live human ÷ unique records attempted, over a fixed window.
• The five causes, in the order they are usually the answer: attempt depth → calling hours → channel mix → data validity → caller-ID reputation.
• The discriminating test: the redial control — 200 records already marked “no answer”, redialled on a fresh outbound number at a different hour, with three more attempts.
• Objective tests exist for four of the five: an invalid-number rate, an email spam-complaint rate, a Do Not Call wash, and a free caller-reputation report.
• Not a contact-rate fault at all: a list so old the numbers have churned. That has its own test, and it is a different one.

Do the arithmetic before you argue about it. Take a real campaign shape: 2,000 records, 5,200 dial attempts, 430 unique records where someone picked up and spoke. Contact rate by record is 430 ÷ 2,000 = 21.5%. Contact rate by attempt is 430 ÷ 5,200 = 8.3%. Both are true, both get called “contact rate”, and they sit two and a half times apart. Diagnose with the by-record number, because that is the one the causes below move; plan capacity with the by-attempt number.

Why is my contact rate low? The five causes, ranked by how often they are the answer

A list of causes with no discriminating test is useless, because every cause sounds plausible when you are staring at a bad number. Each candidate below is paired with the measurement that confirms it or rules it out. The thresholds in the last column are the working rules we use on our own campaigns, not published industry benchmarks — a starting line to argue with, not a standard.

The rule to take away: a cause you cannot rule out with a number is not a diagnosis, it is a hunch with a job title.

The redial control: how to tell a bad list from bad calling before you buy another one

The redial control. Take 200 records already marked no-answer, unreachable or dead. Do not clean them, do not enrich them, do not change the script. Change exactly three things: a different outbound number, a different block of hours from the one that produced the original attempts, and three more attempts spread across three separate days. Then compare that cohort’s contact rate against what the same records produced the first time, which was by definition close to zero.

• The control comes back materially above zero — the records were reachable all along. Your problem is attempts, hours or number reputation, and a new list would have bought you the same failure on fresh data.
• The control comes back near zero too — data validity and list age are now the live suspects, and rows 4 and 6 of the table above are the tests to run next.

Three variables at once is a compromise you make because you want an answer this week; if the control lights up, unpick it by re-running with only the number changed, then only the hours. Below about 200 records, skip the control and run validation instead — that is roughly the smallest cell where a gap between 2% and 15% is not just noise.

Cause 1: not enough attempts, spread across too few days

This is the most common answer and the least interesting one, which is why it gets skipped. The tell is that people quote an average: “we average 4.2 attempts per lead.” Averages hide the shape — in most underperforming files a small tail has been dialled fifteen times and the bulk once. Plot the histogram instead. If the median record has one or two attempts, you do not have a contact-rate problem, you have a coverage problem wearing one as a disguise. The sibling failure is compressing attempts into a single day: six dials on a Tuesday is one sample of that person’s week, not six.

There is a real ceiling here, and it is worth stating because most advice pretends there is not. The 2007 InsideSales.com/MIT study by Dr James Oldroyd — three years of data, six companies, more than 15,000 web-form leads and over 100,000 call attempts — found that after roughly 20 hours from lead creation, additional dials actively hurt the ability to contact and qualify that lead. It measured North American web-form leads in 2007, so read it as a shape rather than a constant: attempts have a window, the window is short, and persistence inside it beats persistence after it.

Cause 2: you are calling outside the hours people answer

Timing is the second-most common cause and it has the cleanest test: plot connect rate — never raw connect counts, which flatter the hours your team already prefers — by hour of day and day of week across at least 500 attempts.

The same 2007 MIT study found the 4–6pm block was the best window for making contact, by 114% over the worst block, and Wednesday and Thursday the best days, by 49.7% over the worst. It also concluded that time of day mattered more than day of week, and response latency more than either. Latency is a separate lever with its own evidence; it lives on our speed-to-lead 5-minute rule page, not here.

In Australia the window is narrower than most teams realise, because it is set in law. Under the Telecommunications (Telemarketing and Research Calls) Industry Standard 2017, telemarketing calls are permitted weekdays 9am–8pm and Saturdays 9am–5pm, and are not permitted on Sundays or public holidays. The highest-yield block in the research — late afternoon and early evening — is therefore legally available and is exactly the block most in-house teams have already gone home from. Add the two-to-three hour spread across Australian time zones and a team calling “9 to 5” from Melbourne is hitting Perth before people are at their desks.

Cause 3: one channel is carrying the whole sequence

A voice-only sequence measures one thing: whether that person answers an unknown number. A sequence that also texts and emails measures whether they respond to you. The diagnostic mistake is treating a silent channel as a rejected channel when very often the message never arrived.

Each channel publishes delivery evidence. For email, Google’s sender requirements for bulk senders — in force since 1 February 2024 for anyone sending more than 5,000 messages a day to Gmail — require SPF, DKIM and DMARC authentication, one-click unsubscribe, and spam rates in Postmaster Tools kept below 0.30%, with below 0.10% recommended. If you are not reading Postmaster Tools, “no reply” is not evidence of anything. For SMS in Australia, check delivery receipts and your sender ID: under the ACMA’s SMS Sender ID Register, from 1 July 2026 texts sent with an organisation’s name at the top of the message must have that sender ID registered. Both tracks are covered in detail in email and SMS deliverability for outbound at scale and SMS sender ID registration in Australia.

A channel with no delivery evidence is not a channel that failed — it is a channel you did not measure.

Cause 4: the data is wrong, which is not the same as the list being old

These two get merged constantly, and they have different tests and different fixes.

Wrong data means the number does not resolve to a working service: mistyped, transposed, a landline recorded as a mobile, a switchboard where a direct line was expected. The test is validation before dialling, not disposition analysis after it. Push the file through number and email verification, then count invalids and hard bounces as a share of records. If invalids run above roughly 5%, stop diagnosing your calling patterns — you are dialling a file a cheap pre-flight check would have fixed.

The second half of the data test is the wash. Australian calling lists have to be checked against the Do Not Call Register, and the register’s industry guidance states that if a marketer has washed its list against the register during the 30 days prior to the call and the register did not show the number as listed, the marketer will not be in breach. Two consequences for diagnosis: a wash older than 30 days is a compliance exposure, and a wash you have not run yet means your dialable universe is smaller than your CRM thinks — which quietly changes the denominator you have been dividing by.

An old list is a different animal: the numbers are valid, the people are simply further from the moment they raised their hand. Split contact rate into recency cohorts — 0–3 months, 3–12 months, 12 months and older — instead of reading the blended figure. Flat across cohorts means age is not your problem. A cliff means a lead recency gap, which is a reactivation problem with a reactivation method, not a dialling fault. For what old records can still do when worked properly rather than redialled harder, our database reactivation record of 4.4% average and 8.9% peak conversion on dormant leads is our own campaign data from the Colliers era, not an industry benchmark.

Cause 5: your caller ID has a reputation problem

This is the least-diagnosed cause and the most technical, and it has grown teeth because carriers now block at network level rather than leaving it to the handset. The regulator’s own reporting makes the scale plain: in its action on scams, spam and telemarketing report for October to December 2025, the ACMA states that telcos reported blocking more than 2.8 billion scam calls since December 2020 and more than 1.0 billion scam SMS since July 2022, with 109.9 million calls and 41.1 million texts blocked in that quarter alone. Those filters are pattern-based, and high volume from a single number with short call durations, a low answer rate and a high hang-up rate is the exact pattern a legitimate outbound campaign produces — and the exact pattern a scam campaign produces.

On the handset side, Hiya’s State of the Call 2026 report, based on a survey of more than 12,000 consumers across six countries, reports that 86% of unknown calls go unanswered. That reframes the diagnosis: a low contact rate on unknown numbers is the default state of the market, not an anomaly, and the question is not why people ignore you but what your number has to do to stop being unknown.

Two tests, in order. First, pull a caller-reputation report for every outbound number you use — the analytics providers whose labels the carriers consume publish free lookups, and a number flagged “likely spam” shows up there long before it shows up in your close rate. Second, run a matched split: identical records, identical hours, identical script, one cell on the incumbent number and one on a number never used for outbound. If the fresh number wins meaningfully, the incumbent is burnt and extra attempts will not fix it. The remedies are structural rather than clever: rotate a pool of numbers instead of hammering one, keep per-number daily volume inside what a human could plausibly dial, register your business identity with the caller-ID providers, and use local presence honestly rather than spoofing, which is the behaviour the blocking regime exists to stop.

What running this diagnosis actually costs you

None of the five tests requires buying anything. What they require is instrumentation and patience, and it is worth being blunt about how much of each.

The arithmetic from there is yours. In our own client work we typically see doubling contact rate roughly double downstream conversion, because contact is the gate every later stage sits behind; that is operator experience from the campaigns behind LeadsNow’s 50,769+ AI-booked sales appointments since 2017, not a study, and we do not publish it as one. The same applies to the figure we quote for speed to lead, roughly a threefold effect on its own. The honest wrinkle most agencies leave out: those numbers do not multiply. Three times two is not the lift you get, because the levers overlap — responding faster is part of how contact rate improves, and a higher contact rate is part of how set rate improves. Stacked properly they land nearer threefold than twelvefold, and anyone quoting you the product of the parts is selling arithmetic rather than results. Where we do publish a measured figure, the method is written down on our methodology page. If the tests come back clean and the only thing missing is the hours to act on them, that is when AI outbound sales becomes a question worth asking — and when the redial control comes back near zero, it is not, because a machine dialling dead numbers faster is still dialling dead numbers.

Contact rate is one stage of a pipeline: it feeds set rate, which feeds show rate, and each has its own diagnosis. The pipeline stages hub links out to the rest of them.

Frequently asked questions

What is a good contact rate for outbound calls?

There is no single number, because the metric has two legitimate definitions that differ by more than twice on the same campaign: by record and by attempt. State which you are using first. As a working rule from our own campaigns, we treat anything under about 20% by record on a recent, opted-in list as a process fault rather than a list fault, and we treat cold purchased data as a separate question, because the denominators are not comparable.

How many times should I call a lead before giving up?

More than once and across more than one day, which already puts you ahead of most files. There is a ceiling, though: the 2007 InsideSales.com/MIT Lead Response Management study by Dr James Oldroyd found that after about 20 hours from lead creation, each additional dial actually reduced the odds of contacting and qualifying that lead. It looked at North American web-form leads across six companies, more than 15,000 leads and over 100,000 call attempts, so treat it as evidence that the window is short rather than as a hard cut-off for your market.

Why do my calls show as spam or go unanswered?

Because unanswered is the default and blocking is now industrial. Hiya’s State of the Call 2026 report, surveying more than 12,000 consumers across six countries, found 86% of unknown calls go unanswered, and the ACMA reports that Australian telcos have blocked more than 2.8 billion scam calls since December 2020. Check your outbound number in a free caller-reputation lookup before assuming the list is at fault.

Does washing against the Do Not Call Register lower my contact rate?

It lowers the size of your dialable list, which changes your denominator, and it is not optional. The Do Not Call Register industry guidance states that a marketer who has washed its list against the register within the 30 days prior to the call, where the register did not show the number as listed, will not be in breach. Wash first, then measure — otherwise you are calculating contact rate against records you were never allowed to call. This is general information, not legal advice.

Can I call prospects on a Sunday in Australia?

Not for telemarketing. Under the Telecommunications (Telemarketing and Research Calls) Industry Standard 2017, telemarketing calls are permitted weekdays 9am–8pm and Saturdays 9am–5pm, and are not permitted on Sundays or public holidays. The practical consequence for diagnosis is that your legal window includes the late-afternoon block the research identifies as highest-yield, and most in-house teams are not using it.

Is my low contact rate a list problem or a calling problem?

Run the redial control before you spend anything: 200 records already marked no-answer, redialled on a fresh outbound number, in a different block of hours, with three more attempts across three days. If contact rate on those supposedly dead records comes back materially above zero, the list was never the problem and a new list would have reproduced the same failure on new data.