i would suggest using google analytics content experiments to track all your urls and you an split test them where half the traffic gets sent to one version of the squeeze page and the other half gets sent to another and then see which one converts better than create another variation to get a better conversion etc.

Hope this helps you

Media buying and testing traffic is quite hard with a small budget.
Usually you will burn quite some money, until you find the traffic source that brings profit or at least pays for itself.
On Adfly there is a lot of bot traffic and autoclicking. Better try something of higher quality.
At the same time keep improving your funnel.

They are, yes.

And that's why it wouldn't be a good way to split-test squeeze-pages: you wouldn't learn what you need to know.

When you split-test something, to see what performs best according to the parameter(s) you're monitoring, you need to do so with traffic from the same sources that you're going to be using after you're selected something on the basis of the test you've done.

There are a number of classic mistakes that marketers make, in this context, which obstruct their intended objectives.

One of them is exactly the kind you're asking about here.

Another is the assumption that anything can be learned by testing the conversion-rate of an affiliate sales-page without list-building, when one's going to be sending subscriber traffic to it, later.

A third is the assumption that "the largest list you can build from the traffic" is going to be the same thing as "the list that produces the largest income that you can build from the traffic". This is very rarely the case, and there are good and reliable reasons for that, one of which is that different traffic demographics opt in selectively under different circumstances. (I earn significantly more, across a range of niches, by opting in 15% of the visitors to a content-rich opt-in page than I do by opting in 45% of the visitors to a squeeze page: http://www.warriorforum.com/main-int...ml#post7939758 ).

You can. But unfortunately you can't draw any statistically valid conclusions from their behavior.

Yes; it is. But not just "bad". "Different", too.


.

Can I ask what is your goal?

If the goal is to optimize the conversion rate, then is the traffic from adfly applicable? Or should you be getting targeted traffic for your niche? Else, the numbers you get from adfly traffic don't matter.

Here's what I'm currently doing to split-test my squeeze page via ClickMeter (you can do it via Google Analytics as well):
1. Create 2 versions of the SP I'd like to test (keep the changes minimal - at least 90% controlled)
2. Create 2 Tracking links & place the 2 conversion pixels in my 'Thank You' page (where subscribers are directed to after opt-in)
3. Create a rotator link to distribute the link evenly (insert the 2 tracking links)
4. Send traffic to the rotator link
5. Track & measure results of the 2 tracking links

This is called 'A/B' testing. If you're starting with a small amount of traffic, I'd suggest you trying A/A/B testing first: test the same squeeze page with the same traffic source over 2 sessions .

Issue with a small sample size is that the results are not definitive and conclusive The difference in results can be attributed to randomness and noises - which you have to factor in your A/B test.

Just my 2 cents.

Alvin