BillBelew.com
Email Marketing

Sometimes I don't know anything abt email2822

Forum Owner Belew private msg quote post Address this user
I did an A/B test on an email yesterday.

I told my buddy that chances are they will tie or the difference be so insignificant that it doesn't matter.

4 hrs in to the results one email title was far outperforming the other ... by 8% or so.

It was a question - Why is .... vs an answer 3 Reasons Why ....

Both of us knew the 3 Reasons why ... would win.

It didn't. The 3 Reasons why got skunked.

Now ... don't stop reading yet.

I had the option to stop the A/B test after 30 mins, 1 hr .. 4 hrs or whatever time.

Had I stopped at 4 hrs ... the remaining batch of emails would have reverted to the winning title. Good deal, right? Who wouldn't want an 8% overall open rate?

But ... I set it to one day.

I checked just a few moments ago ...

Question in the title or Answer in the title - Results.

Drum roll plz.

Why is ... 42.86%

vs

3 Reasons .... 43.21%

The difference is nearly insignificant ... unless you have 1,000,000 emails ... and I don't.

I don't know anything.


Post 1 • IP   flag post
Teacher Rev private msg quote post Address this user
You know how it's said if you buy a red car suddenly you start seeing a lot of red cars? I don't know why I thought of that but when I did, the other day, I was sitting on the porch watching traffic go buy. I decided to test the statement a little and see if red cars were actually more numerous than other colors. So I started counting red cars and total cars passing by on the street in front of me. I thought of you, Bill, as I was doing this.

Anyway, of the first dozen cars that went by, 9 were red. Wow! That's huge. If I had stopped there I would have had numbers in support of a very specific theory. But I kept counting.

Of the first 15, 10 were red.

Of the first 23, 12 were red.

Percentages were slipping. By the time I got to 100, only 18 red cars had been seen. The range went from 3 of 4 in the first few minutes, to less than 1 in 5 after a half-hour of counting. Why had the numbers been so skewed in the beginning? Did I create my own universe? Did my belief waver as I went forward? Is 1 in 5 the real number or are there other factors to consider? Would it continue to change over time? Was my sampling to small to even be relevant? Do there just happen to be more, or less, red car owners in my neighborhood than average? Is red even an available color in some models? Do socio-economic factors play into car color choices? What does any of this matter anyway?

Good luck with your email testing. Just remember... Some will. Some won't. So what?

Perhaps time would be better spent sending out more emails instead of tracking A/B tests. Bean counting might tell you how many beans you've got but I'd rather spend my time making soup.
Post 2 • IP   flag post
Forum Owner Belew private msg quote post Address this user
What's a porch? 8-)

Part of the bean counting on my end is to determine how best and how often to send out emails.

What I have learned so far is that the more I know, the more I don't know. At least I know that now.
Post 3 • IP   flag post
Forum Owner Belew private msg quote post Address this user
More numbers that ask more questions.

The question - Why is ... got opened (so far) by 44.49%
The answer - 3 Reasons... got opened (so far) by 43.62%.

But ...

the first one got reopened by 75.92%
the second one got reopened by 64.61%.

I guess the first bunch needed to read the thing again.

I don't know anything.
Post 4 • IP   flag post
Teacher SusanDay private msg quote post Address this user
@Belew Perhaps the topic was so compelling the 3 reasons or the why was insignificant?
Post 5 • IP   flag post
Forum Owner Belew private msg quote post Address this user
@SusanDay

But the Why out performed the 3 Reasons significantly. I thought that was interesting.

Still the open rates in total was good.
Post 6 • IP   flag post
Teacher SusanDay private msg quote post Address this user
They were excellent
Post 7 • IP   flag post
29614 7 7
Log in or sign up to compose a reply.
destitute