conversion rate

Every agency quotes a conversion rate and almost nobody defines it. The numerator is rarely the problem. A sale is a sale. The denominator is where the number is made, and two teams using different denominators are not measuring the same thing even when they print the same figure. Before you compare anything, write down what you divided by.

What counts as a conversation?

In practice, a thread where the fan answered at least once. But nothing forces that choice, and the same week of work produces very different percentages depending on which of these you count:

Denominator	Counts	Effect on the number
Every fan messaged	Including people who never replied	Lowest, and the most honest about reach
Conversations with at least one reply	Real two-way threads	The usual working definition
Conversations where a PPV was actually sent	Only pitched threads	Highest: measures the pitch, not the funnel
Fans, not conversations	A repeat buyer counted once	A different metric entirely

None of these is wrong. Publishing one without saying which is. If a rate arrives without its denominator attached, it tells you nothing, and the gap between the first and third rows is wide enough that two teams doing identical work can publish very different numbers.

Why are comparisons usually meaningless?

Because the denominator is only the first difference. Three more make cross-team comparison close to useless:

• Traffic source. Fans arriving from a free page and fans who paid to subscribe do not convert at the same rate.
• The hour. A team staffing evenings and a team staffing nights are working different populations, not the same one with different skill.
• The window. Sales counted the same day and sales counted over two weeks (with a follow-up in between) produce different rates from identical work.

The comparison worth making is your own rate against your own rate, same creator, same denominator, same hours of the week.

What did our data actually measure?

Across 42,852 conversations, 1.5 million messages and 38,879 sales, four things moved the rate reliably.

1. When you pitch. Pitching before the sixth message drops conversion by roughly a third. The optimum sits after about ten real exchanges.
2. How you close the sentence. Ending a sales message with an ellipsis is the worst-performing closer we have measured.
3. What kind of question you ask at the close. A closed question asked at the moment of closing costs several points of conversion.
4. The hour. Between 2am and 6am conversion collapses; evenings are the peak.

There is a fifth result, and it is the one nobody predicts: a personal callback, bringing up something the fan told you earlier, placed at the exact moment of the pitch costs points, even though it feels like it should help. It works in the conversation. It does not work in the pitch: personalising a sales message explains why.

What do you do with the number?

Read it next to average order value, never alone. Rate and order size trade against each other: drop your price tiers and the rate rises while revenue does not. A team optimising conversion on its own will find the cheapest thing to sell and sell it.

And treat a sudden jump as a definition change until proven otherwise. Most of them are.