When To Drop A Landing Page

At this point, what should I do?

  • [a] Only run LP 1 since it has the highest conversion rate

    Votes: 2 16.7%

  • [b] Run LP's 1-3 (highest converting) and drop LP's 4-6 (lowest converting)

    Votes: 3 25.0%

  • [c] Run LP's 1-5 (1-3 are highest converting; 4-5 needs more data) and drop LP 6 (lowest converting)

    Votes: 4 33.3%

  • [d] Continue split testing all the landing pages because that is not enough data

    Votes: 3 25.0%
  • Total voters
    12
makethatgreen

makethatgreen

The Freeway Killer
Mar 1, 2009
1,608
18
0
bewbs
I have a quick question about analyzing data, but first, how about a little
bit of encouragement? :evil_laughter:

Advanced Thank You For Question: ( . )( . )'s
(I opened a seperate thread because I thought boobs were not allowed in
this section)


So, here is the deal:

I've been running a campaign for about 30 days now, and I am at the point
where it is time to ramp the traffic up. Throughout the month I have split-
tested and optimized several landing pages. My question is do I have
enough data to determine which landing page(s) I should be running at this
point?

Here are some stats from my campaign:
(ctr is irrelevant because the conversion rate is based on ad clicks, NOT
click throughs.

Landing Page 1: 72 leads, 1.52% conversion rate
Landing Page 2: 73 leads, 1.47% conversion rate
Landing Page 3: 66 leads, 1.48% conversion rate
Landing Page 4: 37 leads, 1.34% conversion rate
Landing Page 5: 21 leads, 1.25% conversion rate
Landing Page 6: 63 leads, 1.20% conversion rate

When analyzing the data be aware that the offer promoted is a higher
paying offer. ($30+)

At this point, what should I do?

[a] Only run LP 1 since it has the highest conversion rate

Run LP's 1-3 (highest converting) and drop LP's 4-6 (lowest converting)

[c] Run LP's 1-5 (1-3 are highest converting; 4-5 needs more data) and
drop LP 6 (lowest converting)

[d] Continue split testing all the landing pages because that is not enough
data.


IM has been really dry/slow for me lately and this is the first profitable
campaign I've had in over 6 months. So far it is pulling a pretty solid ROI.

I'd really appreciate any input you guys might have related to what I should
be doing at this point.
Thanks.
 


jackfitz said:
I'm no pro but 1.5% seems pretty low conversion rate for a $30 range payout offer.
Are you running this on search?
Click to expand...

Actually, I guess you'd call 1.5% the sign-up ratio.

What I was trying to say is that out of all the clicks to that landing page
x.x% converted, NOT the number of clicks to the offer. (of course the
conversion rate is much higher if we are talkin landing page click
throughs to the offer)
 
When you say ramp the traffic up, are you going to be using the same source? If you are going to be adding sources, test all LPs as the traffic from each may convert differently.
 
powell1 said:
When you say ramp the traffic up, are you going to be using the same source? If you are going to be adding sources, test all LPs as the traffic from each may convert differently.
Click to expand...

Yeah, it will be the same traffic source, just with a larger budget.
 
If you are having all the traffic go to those landing pages equally in rotation (which it sounds like you do, 50+ leads to a LP is respectable (It should be around 1k spent or more for 50+ leads), some may argue that that's not enough but 50+ is reasonable especially since you have that across like 6 separate LPs) to say that LP #1 is the best then drop the others (I would probably keep the top 3) and have them all go to those top 3 (or just LP #1 if you want).

After you have your top landing page or your top 3, I would then try to improve on your #1 and adding that to your landers to test. Seriously, after you test like 5 or so landers and you find the top lander, try to improve on it and put it up against your top lander and see how it does. Don't just settle for your top lander at the time, keep trying to improve on it and then split test your edited lander to your top one, if you know what you are doing you can keep this up and you can continue to improve through multiple cycles until you have a really good one :)

Of course by the time you are through a couple landers you got to expect other people are going to copy you and rip your landers (This happens all the time) but just take it as a compliment because that just means you are doing something right, doing better then everyone else.
 
Unarmed Gunman said:
Run LP 1 and drop the other 5. Now build a few more LP's to split test against LP 1. Let Darwin decide.
Click to expand...

That's what i was gonna say. Ideally you should never stop split testing.
Take the winner, make 2 more and run them against the winner.

Rinse repeat.
 
I'd test 1-3 a little further, see if they even out or even swap positions. Test them against another 3. Do you have a specific ROI you're aiming for? What conversion rate do you need to hit with this traffic source to get it?
 
TylerL said:
I'd test 1-3 a little further, see if they even out or even swap positions. Test them against another 3. Do you have a specific ROI you're aiming for? What conversion rate do you need to hit with this traffic source to get it?
Click to expand...

Bad advice. Even if the top 3 do "even out", how does that help? You need to take the top one and split test against completely new variations. The other 4 have proven that they aren't any better so you'd be wasting time to continue testing them. New blood, now.
 
Unarmed Gunman said:
Bad advice. Even if the top 3 do "even out", how does that help? You need to take the top one and split test against completely new variations. The other 4 have proven that they aren't any better so you'd be wasting time to continue testing them. New blood, now.
Click to expand...

Depends what the landing pages are. They're all within half of a percent of each other. If they're all review landers, yeah drop everything but the top one, but if they're different styles then I'd keep the 3 top contenders and split test variations of them against themselves. 1.52% versus 1.48% conversion on 1000 clicks at $30 per lead is a difference of ten dollars, which I'd say would be worth it to figure out wether or not a whole style of LP isn't effective.
 
TylerL said:
Depends what the landing pages are. They're all within half of a percent of each other. If they're all review landers, yeah drop everything but the top one, but if they're different styles then I'd keep the 3 top contenders and split test variations of them against themselves. 1.52% versus 1.48% conversion on 1000 clicks at $30 per lead is a difference of ten dollars, which I'd say would be worth it to figure out wether or not a whole style of LP isn't effective.
Click to expand...

The reason I say it's a waste of time, is because you only have a finite number of clicks to rotate to the LP's. Why keep testing the same LP's that he has over the last 30 days? You take the best converting one, and split test new variations that might actually be better. He already has 30 days of data showing that they are worse, or slightly worse. It's time to take the best one and test new ones with it. If you keep 3, then it will take longer to properly test the new LP's, because they will have fewer clicks than they could if you just keep the best one.
 
Alright, a couple hours ago I decided to eliminate LP's 4-6. I'll keep LP's 1-3 in rotation
and in the mean time I am going to be working on modified versions of the three current
LP's.

I guess for a start I'll change up the colors, then layout and text. I know that you
don't want to change more that one or two elements at a time, correct? I guess a
work around is split testing a lot more LP's at the same time.
 
You must log in or register to reply here.
Top Bottom
Back