For years, when I’ve compared safelists, the number I care about most has always been conversions.
Traffic matters too, of course. You need people to see an offer before they can respond to it, and reaching a larger audience can give you more opportunities to generate those conversions.
But traffic alone has never been enough for me.
That’s actually one of the reasons I started the Safelist Attention Project.
I wanted to look at something that normal hit counts and conversion stats don’t show very well:
When people click an ad and reach the page, are they actually paying attention to what they’re looking at?
So I created an experiment that wasn’t trying to sell anything or collect an email address.
I gave people one simple thing to do: notice the page and deliberately tell me they noticed it.
And after looking at more than 8,000 visits, one of the most interesting things I discovered was that traffic volume and attention didn’t always move together.
Some safelists were especially good at generating reach.
Others sent fewer visits, but a much larger percentage of those visitors deliberately interacted with the page.
That doesn’t make one type of traffic automatically better than the other.
But it does tell me something I couldn’t see from the hit counter alone.
What Safelist Traffic Adds to the Picture
Traffic still matters.
Every visit is an opportunity for somebody to see the offer and eventually take whatever action I’m hoping they’ll take.
But the Attention Project was designed to look at the space between the visit and the conversion.
What happens after the page loads?
Does the person actually notice what they’re looking at?
Normally, a hit counter can’t answer that question. And if somebody doesn’t ultimately opt in or buy, a conversion report doesn’t tell me much about what happened in between either.
That’s the layer I wanted to explore.
This Experiment Gave Me a Way to Measure That
If you read my first article in this series, I Tested 45 Safelists: Which Ones Sent Visitors Who Actually Paid Attention?, you already know the basic setup.
I sent the same email to 45 safelists.
Every safelist sent people to the same experiment page, with its own tracking source so I knew where the visit came from.
The page wasn’t selling anything.
Instead, it asked one simple question:
Did you notice this page?
And there was one button:
Yes… I Noticed This Page
A visit told me the page loaded.
A click on that button told me the person was engaged enough to deliberately respond.
It still wasn’t a sale or an opt-in. But it gave me a measurable signal of attention.
After cleaning the final data, the experiment produced 8,392 named-source, human-likely visits, 143 attention clicks, and 94 distinct tracked clickers.
That works out to an overall attention rate of 1.70%.
Then I started looking at how those visits and clicks were distributed.
That’s where things got interesting.
SendCircle Didn’t Need to Send the Most Traffic to Stand Out
The clearest example was SendCircle.
It produced:
174 analyzed visits
32 attention clicks
20 distinct tracked clickers
That gave SendCircle an 18.39% attention rate.
For comparison, the overall experiment rate was 1.70%.
But here’s the part that really caught my attention.
SendCircle accounted for only about 2.1% of all the analyzed visits in the experiment.
Yet it generated about 22.4% of all the attention clicks.
A relatively small slice of the traffic generated almost a quarter of the measured interaction.
And those 32 clicks came from 20 different tracked visitors, so this wasn’t simply one person clicking over and over.
SendCircle didn’t need to send the most traffic to stand out.
Its strength in this particular experiment was how often the traffic it did send deliberately interacted with the page.
That’s something I wouldn’t have seen by looking at traffic volume alone.
But Volume Still Matters
I don’t want to swing too far in the other direction.
The lesson here isn’t that fewer hits are somehow better.
My Daily Mailer is a good example.
It produced:
1,005 analyzed visits
30 attention clicks
19 distinct tracked clickers
That works out to a 2.99% attention rate.
Obviously, 2.99% isn’t close to SendCircle’s 18.39%.
But My Daily Mailer produced only two fewer total attention clicks.
So we had two very different paths to almost the same amount of interaction.
SendCircle had relatively modest reach with an unusually high response per visit.
My Daily Mailer generated substantially more reach with a lower—but still above-average—attention rate.
Both results are useful.
They just tell me different things.
Full disclosure: I own My Daily Mailer, so its members may already be more familiar with me, my writing, or the kinds of pages I promote. That familiarity could have influenced the results.
Still, the comparison illustrates an important distinction:
Attention rate and total attention are not the same measurement.
A smaller source can have a remarkable response rate.
A larger source can have a lower rate while still generating a lot of total interaction.
You need both numbers to understand what happened.
Click Leverage Found a Nice Middle Ground
Click Leverage generated 342 analyzed visits, 10 attention clicks, and 8 distinct tracked clickers.
Its attention rate was 2.92%.
It didn’t have SendCircle’s unusually high percentage or My Daily Mailer’s traffic volume.
Instead, it combined a meaningful amount of reach with attention comfortably above the experiment average.
I like having Click Leverage in this discussion because I don’t own it.
It shows that the patterns weren’t limited to my own sites or audiences that might already know who I am.
Mister Safelist Showed Me Another Side of Attention
Mister Safelist produced 668 analyzed visits and 14 attention clicks, giving it a 2.10% attention rate.
But what interested me most was that those 14 clicks came from 14 different tracked visitors.
There were no repeat clickers contributing multiple interactions.
That matters because two safelists can generate the same number of clicks while the behavior behind those clicks is very different.
One might have a smaller group returning and interacting multiple times.
Another might spread its responses across a wider group.
Again, I own Mister Safelist, so familiarity could have influenced the result there too.
I’m going to save the deeper repeat-participation patterns for another article.
For now, this showed me there was one more measurement worth watching.
The Three Things I Want to Know Now
After digging through the results, I think there are three useful ways to look at a traffic source.
Reach
How many opportunities did my ad get to be seen?
More visits mean more chances for something to happen.
Attention Rate
How often did a visit produce the response I was measuring?
For this experiment:
Attention clicks ÷ analyzed visits
Unique Attention
How many different tracked visitors responded?
That helps distinguish broad participation from repeated participation.
And I still want to know the total number of interactions.
Because a percentage without volume can be just as incomplete as volume without a percentage.
A 10% response from 50 visits gives you five interactions.
A 3% response from 1,000 visits gives you 30.
Which is better?
That depends on what you’re trying to accomplish.
What Am I Trying to Get Those Hits to Do?
I think that’s the real takeaway for me.
Traffic, attention, and conversions are connected—but they’re not the same thing.
Traffic gets somebody to the page.
Attention tells me they noticed and responded to something on it.
Conversion tells me whether they ultimately did what I wanted them to do.
If I’m running a lead capture page, I care about opt-ins.
If I’m promoting a paid offer, I care about sales.
And in this experiment, I specifically wanted to know whether people were paying attention.
That’s why I don’t think any one number should make the entire argument.
I’ve written before about how most people don’t need more traffic if the traffic they already have isn’t producing the desired result.
I’ve also been thinking a lot about why some ads make you curious enough to click.
This experiment fits right between those ideas.
Getting somebody to the page is one step.
Getting their attention is another.
Getting them to convert is another.
Different Strengths Showed Up in the Data
Here’s a simple snapshot of the examples I’ve talked about:
| Safelist | Visits | Attention Clicks | Unique Clickers | Attention Rate | What Stood Out |
|---|---|---|---|---|---|
| SendCircle | 174 | 32 | 20 | 18.39% | High attention rate |
| My Daily Mailer | 1,005 | 30 | 19 | 2.99% | Reach plus interaction |
| Click Leverage | 342 | 10 | 8 | 2.92% | Balanced performance |
| Mister Safelist | 668 | 14 | 14 | 2.10% | Broad unique attention |
I’m deliberately not ranking these four.
The interesting part is that they stood out in different ways.
Some platforms were stronger at delivering reach, while others stood out for interaction.
Traffic volume and attention turned out to be different strengths.
What This Experiment Does Not Prove
A lower attention rate does not mean a safelist sent bad traffic.
A high attention rate does not mean that source would necessarily produce the most leads or sales.
A different email, page, offer, or audience could produce different results.
I also wasn’t working from identical membership levels or mailing reach across every safelist, and familiarity may have influenced results on the two sites I own.
Finally, the distinct visitor figures are based primarily on persistent browser IDs, not guaranteed unique physical people.
This was one experiment measuring one specific kind of deliberate interaction.
That’s enough to learn from it without pretending it proves more than it does.
Traffic Still Matters. It Just Doesn’t Tell Me Everything.
I’m still going to track hits.
I’m still going to care about reach.
And conversions will still be the number I ultimately care about most when I’m running an actual marketing campaign.
But now I have another layer to think about between the traffic and the conversion.
Someone reached the page.
Did they actually notice it?
And did enough of them notice it to make that traffic source interesting for what I was trying to accomplish?
So rather than asking only:
Which safelist sends the most hits?
I think the better question is:
What am I trying to get those hits to do?
Because this experiment showed me there wasn’t only one way for a safelist to produce an interesting result.
Some stood out for reach.
Some stood out for attention rate.
Some showed broad unique engagement.
And some found a strong balance between them.
That’s what I’m going to look at next.
