How to properly split-test landers and offers in Voluum?

[ppcnewbie]

Not sure if this has been discussed or not, but I want to know what is the best way to split-test offers and landers in Voluum.

Now they have that Flow entity which combines the offers with landers and in the campaign you just assign the Flow.

Anyway, let's say I have 3 landers and 3 offers and I want to split them equally.

There's one option to have:

1 path - 100 weight
3 landers - each with weight 100
3 offers - each with weight 100

But I can also have 9 paths wit a combination of one lander-one offer, each path weight 100.

Or I can also have 3 paths, one for each lander and each lander with the 3 offers.

I don't know if there's any difference in terms of reporting or anything. So what would be the best option?

 

[Dispply]

Hi!
What does it mean "weight 100"?

 

[ppcnewbie]

Paths – Customer Feedback & Support

If I have one lander with weight 100, it means it gets all the traffic.

If I have 2 landers with weight 100, they each get 50%.

If I have a lander with 200 weight and one with 100 weight, 1st lander gets 66% of the traffic, etc

 

[newlxx]

I don't know if there's any difference in terms of reporting or anything. So what would be the best option?

It doesn't make a difference at all.
To split test equally - add all offers and landing pages and put weight at as 100.

 

[ppcnewbie]

Hm... I would think that as well, bun I've read in Finch's Premium Posts that he always uses (and recommends) split testing like this:

if you have 2 landers and 2 offers, you'd have:

flow 1: lander a, offer 1
flow 2: lander a, offer 2
flow 3: lander b, offer 1
flow 4: lander b, offer 2

Why, I don't get it, he doesn't explain it, he just says:

"IMPORTANT: Use subids (or V1, V2 etc) to distinguish each duplicate offer
and landing page in your tracker. Otherwise if a conversion falls under
lander a, it won’t be clear which path it came from.
"

That's what confused me.

So, if I set it up as this, can I tell offer 2 converted from lander 2? i need to setup some special tokens for this?

 

[newlxx]

Hm... I would think that as well, bun I've read in Finch's Premium Posts that he always uses (and recommends) split testing like this:

if you have 2 landers and 2 offers, you'd have:

flow 1: lander a, offer 1
flow 2: lander a, offer 2
flow 3: lander b, offer 1
flow 4: lander b, offer 2

I'm pretty sure he was talking about this when setting up a matching funnel consisting from ads - LPs - offer

angle 1: banner a,b,c > lander 1> offer
angle 2: banner d,e,f > lander 2 > offer

In this case, you would need to set it up like that since you would have to redirect traffic according the banner which user clicks on to show them congruent LP and offer.

So, if I set it up as this, can I tell offer 2 converted from lander 2? i need to setup some special tokens for this?

If all your incoming traffic is coming from the same banner or it pops/redirect traffic. A simple setup is fine.
Your trackers reports will allow you to drill down specific stats for each combination.

 

[Dispply]

I agree, all the flows should be equal to have objective results.

Some more tips:

1. Run one test at a time
2. Test one variable at a time
3. Measure as far down the funnel as possible
4. Test at the same time
5. Make extra A/A - tests