Home > Paid Traffic Sources > Adult Traffic (NSFW)

---

Balancing "statistical significance" and "banner burnout"? (14)

---

So something I've been wondering lately when it comes to your banner optimizations for your campaigns is how do we balance letting a banner run until its "statistically significant" and with "banner burnout"?

For example I've read that people like to refresh their banners every 2-3 days, but that doesn't seem long enough to conclude whether a banner is good or not?

Say I launch 3 banners on a single placement, depending on the geo/offer etc. maybe I only get 5-10 conversions per banner. Then perhaps it takes 2-3 days to get a good amount of conversions to tell if I have a winner or not, but by then my CTR drops and the banners are already "burning out". Seems like I'd be running in circles.

Or am I missing something here?

Any advice would be appreciated.

---

Banners make the users interested to click through, landers warm the users up and prepare the conversions.
So for banners most important is that they are somehow related to what you are promoting and that they have a high CTR.
So you can just rotate banners in such short timeframe to keep high CTRs going.
It's just to get the users to your lander.
The lander is more important for the conversions.

---

I agree with twinaxe for the most part - except I've seen cases where banner+lander congruence can have significant effects on conversion rates. I've had banners that fetched very high CTRs that converted like shit - which made it all the more damaging on traffic platforms that naturally directed an increasingly larger percentage of the traffic to the highest-CTR creatives.

But in cases where banner burn out occurs at a very fast rate (e.g. adult dating at high volume traffic), twinaxe's advice will definitely come in handy. There ARE exceptions as I've stated above, but going with high-CTR banners may very well give you better results than waiting for stat sig.

Also, you can lower the confidence levels of whichever stats tools you're using, to find a balance between accuracy and efficiency. That way you wouldn't just solely rely on gut feeling or CTR, and yet you wouldn't be losing money to burnt out banners from waiting for stat sig at high confidence levels.

It would also help to figure out which banner ELEMENTS are giving you the most success, then test variations of those. e.g. Older models, bigger font, specific colors or banner layouts etc. That way you'd be able to churn out good banners in general, instead of having to rely on individual+random banners to do well.



Amy

---

I agree with it.
Don't get me wrong, I didn't mean to throw any shit on the platform and only try to get the highest CTR possible.
That's what I mean when I said it's important that the banners are somewhat related to what you promote.
I also tested lots of things and just as you say, you can have banners with extremely high CTR that don't convert at all.
Often that happens when you try really weird stuff that makes people click through just because it's weird.
And sure it can also help to have a similar visual style going from banner to lander to offer.
But without high enough banner CTR to make the ad spend worth it you won't even get to the point to start optimizing the other parts.

---

I LOL'ed at that!! It is exactly as you said. I'm remembering all the weird stuff I was testing that got high CTR but didn't convert.


It's a real honor having you as part of our community - you leave behind you a trail of gold.



Amy

---

Yes, stuff like midget sex, the woman with full beard, a guy with two dicks and so on.
Although most of that shit doesn´t convert at all it´s still much fun to come up with the weirdest stuff.
Especially in adult it´s a nice distraction from the 1000th pussy you see that day.
Thanks for the kind words, just trying to spill a bit of my sciolism here or there

---

Based on my own experience, banners play quite an important role in adult dating, not trying to diminish the importance of LPs of course, but I spend WAY more time on banners than on LPs. Banners burn out pretty fast, LPs always last way longer for me.

What I'm doing is paying attention to 2 things, let me quote from you guys as both have been already mentioned

1.

If the CTR isn't high enough, it's not worth messing with such banners for sure. The CPCs simply cannot be higher than "some" level.

2. This is what I do very often, I try to select the winner before it reaches FULL significance. So I'm not going by the gut feeling only, but I don't wait for high confidence levels either.

Using these 2 "points" usually works for me

---

Are you guys using rotating tags to display creatives?


Sent from my iPhone using STM Forums mobile app

---

I'm not sure I understand what you're asking for.

I'm rotating ads at the source and also using tokens to be able to track them properly. Is that what you're asking about?

---

Priceless LOL!! And you forgot women with 3 "girls" instead of the usual 2.

I don't remember much of what I was testing because it's been a while and I have the memory retention of a goldfish, but I remember instead of using the usual headlines, I was using quotes of women talking dirty, and matched those quotes with images of women in...uh...compromising positions. People clicked on them like crazy but nothing converted. I lost quite some money that day which is why I still remember those. Glad I set a budget on that campaign or it would have gotten way uglier.

You're fun to chat with!



Amy


Sent from my SM-G930W8 using STM Forums mobile app

---

Hey Matuulo thanks so much for mentioning this, yep i am interested in any solutions or guidelines for setting similar systems up to have js tags for rotating ads, that would be solid not needing too much maintenance, think this is a topic that has not been discussed much in the forum.


Sent from my iPhone using STM Forums

---

I think you should use what the networks you are buying traffic from are offering. These are called dynamic tracking tokens. Check this site to see what they are with many popular networks : https://trackingtokens.com/

You need to configure each of the traffic sources in your tracking system, not sure what you are using now ... otherwise it wont be working.

If you can provide more info about your setup, I can provide more guidance. But maybe it would be better to start a separate thread for it, so it doesn't get mixed with the topic of this one.

---

oh yes i was actually thinking about custom built solutions for rotating ads through iframing if traffic source allows it, i will try to formulate the question better and start a new thread, thanks!


Sent from my iPhone using STM Forums

---

This is a complicated one in the current climate ... google and their "pain in the ass" better ads initiative made publishers and traffic networks extra cautious when it comes to creative approvals ... without solid history and clean report, you'd have a hard time getting approved for iframe ad serving.

---

Home > Paid Traffic Sources > Adult Traffic (NSFW)