Home > Paid Traffic Sources > Facebook & Instagram

---

SPLIT Testing landers on FB?! (20)

---

Are we able to split test landers on FB by letting traffic distribute between?

I've seen quite a few page builders like zipify pages offer A/B testing - but doesn't this risk triggers 'Circumventing systems'?

I always thought this was something we couldn't do. I was under the impression Facebook bots crawl the page for approval. If it turns out you serve a different lander from the one they approved, you risk triggering 'circumventing systems', as when they approved your ads, they're also approving your lander for that ad too.

Any insight would be appreicated.

---

Pretty much it.

You can do it, with server-side rendering or scripting e.g. PHP to include source page A or page B in the content served.

This would not be apparent as a redirect to the client, but its very obvious that you're serving multiple pieces of content given you could just reload the page or some other bot/person could check and see a different result.

It would go through approval fine in most cases (esp since you could serve one variation only then), but it is only a matter of time before FB catches it and the expected consequences...

---

OK thanks, I won't bother then.

Will continue to manually pause the ads and rotate landers to split test.

Strange that many landing page builders would offer this is a service when it has ban potential...

---

The thing is, people use all kinds of traffic sources and some are totally fine with rotating landers. So these tools offer this function in order to serve as broad customer base as possible. Push, POPs, adult... it's the standard thing to rotate landers with these traffic types. Or when sending out your own emails to existing customers or regular subscribers to your lists... it's very common to split test landers in such cases too. Or when you have your own organic traffic... But yes, with sources like FB or the large native networks, it's a no-no.

---

This is because many makers of tools do not actually run campaigns themselves. Or, they only have experience with other traffic types but not FB.

It WOULD make sense for them to at least do their research by collecting lists of best and worst practices, and some of them do, but the risk is always there. We need to be ever-vigilant in how we choose to use every tool.

I'm definitely with you and zeno on this one - showing different landers to FB would be asking for a ban.

One way to split-test: Set up multiple versions of the same adset, each going to a different landing page, and running them at the same time. For sure, the adsets would be targeting different audience pools, so the split-test wouldn't be 100% accurate. If you really want to be sure, repeating the test multiple times would increase your confidence in the results.


I'm wondering - what would this entail exactly?


Amy

---

Yeah, this is true although tools like Zipify are used mainly for ecommerce with FB being the most common source of traffic - which is why I was unsure if something had changed or not.

---

That's why I was surprised with Zipify particularly (and a few others, Shogun etc) as Ezra does run campaigns himself on FB. But I'm certainly not going to risk it, just not worth it.

For sure have tried this but may try again, thanks.

At midnight when conversions are low I will pause all ads, change the lander (same URL) and un-pause ads. I usually only do this once I have a stable CR day to day.

Thanks for input all. I will not be risking it for the sake of saving time!

---

Ezra also plays in a different pool that most of us. Just like ScottyG is finding out. Life is different at the top. Ezra almost certainly has a high level rep who is better able to tell him what is and is not allowed as well as help him recover any banned accounts.

There is quite a bit of conventional wisdom about Facebook that is just wrong. However, because it is such a black box, it is hard to be sure if the ban was for A or B.

---

This is very true. I think erring on the side of caution is the best way in this scenario when there are workarounds.

---

I would agree. It would make sense to be more conservative when you don't have access to those resources.

Personally, I would not split test landers on Facebook using any tool to rotate them. Instead I would use Facebook's built in split test or simply build similar adsets or even ads and split test the lander there.

Food for thought though. The last I saw, Clickfunnels is perfectly safe to use on Facebook. Just make sure you have a custom domain versus a subdomain assigned by Clickfunnels. If you've used Clickfunnels, you will notice that the Funnel URL is actually a redirect to the first page of the funnel. Even if the URL shown for a page is typically a redirect and you have to dig down to find the actual page URL.

So Facebook doesn't always seem to care about redirects. I suspect it is looking for behavior that isn't consistent in the redirect, I think some people still use bit.ly, and/or behavior that hints at cloaking.

---

But...how would this be different from simply rotating urls? Since pausing ads does not put them back into moderation.



Amy

---

Quite simply:

-When you have a rep, they can judge based on a case-by-case basis, based on judgment of the individual rep.

-When you don't have a rep, camps are at the mercy of the FB bots, and then "lower-level" reviewers that are conceivably just following a standard list of criteria. I'd imagine that both were designed to "err" on the side of caution.



Amy

---

You can try to test it in the following way: either by replacing the link in an already running campaign (via the tracker), or launch it from two different accounts. In gambling, it's a lot easier, you can conduct testing by simply replacing the deep link

---

I wouldn't suggest to replace the link in an already-running campaign.

For the same reasons already covered in the posts above.


Amy

---

It is bad enough you spam the forum, but that is absolutely horrible advice for Facebook. That is an easy way to get an account and maybe even a BM or profile banned.

---

No, you misunderstood. Of course if you launch a Facebook campaign, then stop it and change the links, your accounts will get banned pretty quickly. However, you can substitute the link for a landing in the tracker ensuring users get sent to your website without stopping your FB ad campaign

---

You have it completely backwards. Please stop giving bad advice.

You can edit a campaign in Facebook, the ad will then go back into moderation. Obviously if you are running BH or even GH, you may get rejected or even banned. But that is allowed, although not advised due to how the algorithm works.

But completely changing out a link on an active campaign by updating the lander in your tracker? That is very quick way to get banned.

---

Ok, let me try to explain in simpler terms so as to avoid any further misunderstandings.


Many trackers have their own cloaking services that are pretty easy to set up. Facebook has been triggering quite often lately, forcing people to use the latter. So what are we left with? White page - the page mods see, black page - the actual landing users are sent to. Experienced buyers have a list of white pages that are guaranteed to pass moderation. You can also swap out black pages while a Facebook campaign is running without needing to suspend it. What does this leave us with? After optimizing a Facebook campaign, you can test landing pages simply by changing them in the tracker as a black page and then scrutinize, analyze your findings. It may look a bit confusing but there are entire affiliate marketing teams that conduct tests exclusively this way. How about we don't flood this thread, we can continue in the DMs

---

I think we all understand what you have in mind. The thing is, we are discussing whether changing or rotating the actual LPs would equal to a possible suspension/ban, which is obviously the case. Doesn't matter what method you use, if FB detects a change in the LP, they will ban the account. The risk can be decreased with cloaking as you suggested, but it's a different field entirely and I believe it's not something that the OP was trying to find out.

So to put it simple, as long as we are talking about WHITEHAT campaigns, changing the LPs in a running FB campaign or rotating them is a NO-NO. I think we can all agree on that.

---

@clicklead If you want to advocate cloaking, then be transparent about it. This thread was never about trying to circumvent Facebook's system, but how to A/B test without circumventing and triggering a disapproval or ban. Someone less experienced in Facebook may have thought your advice was suited for a WH or even GH campaign.

---

Home > Paid Traffic Sources > Facebook & Instagram