Contact rate calculator

The four ratios, and how they constrain each other

Connect rate is connects over dials: how often an attempt produced a live human. Right-party contact rate is right-party contacts over connects: how often that human was the one on the record. Multiply the two and you get right-party contacts per dial, which is the whole campaign in one fraction -- and dials per right-party contact, the third output, is simply that number turned upside down. They cannot move independently, which is what makes the pair diagnostic rather than decorative.

That is the useful part. A campaign with a healthy connect rate and a poor right-party rate is reaching households and missing people -- old numbers, shared lines, a list where the contact name and the phone number came from different places. A campaign with the opposite shape is dialling good numbers at times nobody is home. Both look identical if you only track one blended figure, and both get "the agents need more training" as a diagnosis, which fixes neither.

The tool also says so when the counts do not nest. Every right-party contact is a connect and every connect is a dial, so a report where connects exceed dials is two periods or two campaigns glued together. It still shows the arithmetic -- hiding it would just look broken -- but it names the offending pair above the results, so a connect rate over one hundred percent reads as the counting problem it is rather than as a number worth acting on.

Count them the strict way, or do not compare them

A contact rate is only comparable against another period counted the same way, and the definitions are where most comparisons quietly break. A dial is one attempt, so three attempts on one record are three dials, not one. A connect is a live human on the line; a voicemail greeting that got classified as an answer is not one, and every dialer will hand you some of those. A right-party contact is a connect that reached the person the record is about -- not a spouse taking a message, not a receptionist, not a wrong number who was polite about it.

Then hold the window still. Contact rates move with the hour of the day, the day of the week, the age of the record and how many times that record has already been called, so a Tuesday-morning fresh-lead number and a Friday-evening backlog number are two different measurements wearing the same label. If you only ever produce one figure a month, produce it over the same window each time and write down which window it was.

What list penetration is not

The fourth output is dials against records: how many attempts the campaign spent per record in the list. It is not the share of the list that has been reached, and it is not the share that has been touched at least once. Those are different questions and neither is answerable from these four inputs -- both need a count of distinct records dialled, which most reports will give you but which is a fifth number, not a rearrangement of the first four.

Read it as pace. Below one, the list is bigger than the effort being spent on it and there are records nobody has called at all. Well above one, the campaign is re-dialling the same records, which may be exactly right -- the attempt that reaches someone is frequently not the first -- or may mean a small list is being burned while fresh records sit in a spreadsheet. The ratio cannot tell those apart. Put it next to the distinct-records number and it can.

Why there is no benchmark on this page

The obvious thing to put beside a calculator like this is a target: a percentage you should be beating. We have not, because a defensible one does not exist here yet. The outbound contact-rate figures that would come from our own floor are scaffolded in this repository with no value attached, and they stay that way until they are measured properly rather than remembered. The percentages that circulate in outbound sales content mostly trace back to vendor marketing, and repeating one would make this page worth less than the arithmetic it already does.

There is one measured outbound dataset published on this site, and it is a shape rather than a level: which dial attempt reaches a person most often, and how much that matters next to the hour you call. Its own write-up explains why the height of that curve is a property of the classifier that produced it and must not be quoted as "the" outbound contact rate. Read it as a direction to test against your own list, which is the only benchmark that ever applies to you.

Worked example, using the numbers the page loads with

The tool opens with four thousand dials, five hundred twenty connects, a hundred ninety right-party contacts and a list of fifteen hundred records. Connects over dials is a thirteen percent connect rate: most attempts do not reach a live human at all. Of the humans reached, right-party contacts over connects is about thirty-seven percent -- so roughly two in three live answers are not the person on the record, which is a data problem rather than a dialing one.

Put the two together and it takes about twenty-one dials for every right-party contact. List penetration -- dials against the fifteen hundred records -- comes out to about two hundred sixty-seven percent, meaning the campaign is dialing each record roughly two and two-thirds times over on average. Read that next to the connect rate: heavy re-dialing of a small list and a low connect rate together usually mean the list itself has gone stale, not that the agents need more attempts.

Where this fits

• Outbound dial cadence benchmark -- our published study of which attempt and which hour actually reach a person, including the parts that do not flatter us.
• Outbound call center -- what a managed outbound campaign involves once the ratios above say the problem is capacity rather than data.
• Roofing lead generation -- the follow-up argument in full: most lead problems are contact problems one layer down.
• Call center cost estimator -- once you know your dials per right-party contact, this turns it into what a conversation costs.