How I Dealt With Banner Blindness When Buying 327,996,501 Impressions (11)
How I Dealt With Banner Blindness When Buying 327,996,501 Impressions
You launch a campaign, it gets a nice 1.5% CTR and then in 2 days it’s half that. “What is going on?”, you ask yourself. This is a phenomenon called “banner blindness” and it means that people have seen your banner so many times that they are “blind” to it. How can you stop this? I will explain the strategy I used on a high volume campaign that ran for a few months, til the offer got paused.
It’s not a question of if banner blindness hits, it’s just when. The best way to solve this is to deal with it before it occurs and has devastating effects. You will notice many posts about banner blindness if you google them, and how to avoid it by “mixing it up”. I will try to be more exact in the steps I took, which you can try to replicate for your campaigns. Let’s dig into it!
This campaign had 327,996,501 impressions on Go2Mobi, and kept running for several months. It is actually the same campaign I talk about in this case study.
The offer was a popular battery saver app and after just one day of good CTR on my best banners, second day just blasted me with only around 60% that CTR. My click price increased, and even with the same conversion rate, my campaign showed struggles. I tried renewing banners chaotically, and it helped, but this strategy had 2 big issues:
- I had no idea when to renew and what.
- It took almost 100% of my time so I couldn’t work on anything else.
This is the strategy that crystallized after a few weeks of chaos.
DAILY:
- Cut the worst performing banners til only the most successful is left. With so many daily impressions, it usually took 1 day to find the best banner in a batch. I used this tool to check probabilities of being the best and usually went for 90% probability.
- Create 5 variations of this winning banner and upload.
THURSDAY:
- Create 5 complete new designs with the same headline for the banners and add them to the campaign.
- Create 5 new designs with DIFFERENT headline for the banners and add them to the campaign.
You have to also mix in some new headlines and very different designs. Less often than smaller variations but YOU HAVE TO if you want your campaign to not die out in less than a month. You will find better ones than your initial profitable banners. My advice is to start out saying the same thing with different words. Creativity will start flowing from that point on.
FRIDAY/PRE-WEEKEND:
- Upload 10 variations of the best performing banner.
On the weekend, you cannot renew banners on Go2Mobi (similar on many other platforms) so you should have more than the usual amount in the test.
One variable you can play around with is the weight of each banner. What is the weight you ask? In Go2Mobi there’s the option to show a banner more frequently than another. This is called the “weight”. In some cases you may want to put a smaller weight for the very different banners, in case you are not there to cut them when you reach a statistically significant sample, with negative results for them. This means that failed tests will have less of a negative impact on the overall performance of your campaign.
The idea for my plan was the following: We know what works, so we want to play around with colours and design elements. We also want to test out new things once in awhile so if our angle and general design is also performing worse, we know we tested new things too. Finally, once you cannot renew daily, significantly increase the number of variations being tested.
Depending on the nature of your traffic source, maybe you have to renew less often. As long as you have a strategy prepared, you can wait before the first renewal, to see the fluctuation in the first 2-3 days once the campaign is profitable. Then start right away with a clear strategy.
I used to do this task myself. The banners took a significant portion of my time, even when having a plan, but I learned what exactly I would tell a designer to do. It was only towards the end of this campaign that I hired a great designer and now he follows this process for our banner campaigns. I’m glad to say he does a much better job than me at creating new designs and concepts. If most your traffic is from banners, a good designer who can pump out banners daily is a great investment.
While reading your post, I can't help myself but blame my lazy ass for not testing more banners when I was pushing the same battery app. Now I see I gave up too fast, with your approach I could of milked that campaign for way longer ... well, I do pay way more attention to banner burnout now than I used before 
On a side note, I'd love to know how much $ the owner actually paid for that battery app installs, pretty much any experienced affiliate I know has pushed it
Originally Posted by matuloo
On a side note, I'd love to know how much $ the owner actually paid for that battery app installs, pretty much any experienced affiliate I know has pushed it
Hey man, how is it going?
Couldn't meet you in AWA, I met Matuloo and a bunch of other people
Anyway, quick question. I am a little confused with your post here.
Did you say that you upload banners Daily, On Thursday and On Friday too? If you are uploading daily, why are you uploading on Thursday and Friday?
I am definitely not uploading enough banners, I sometimes only mass upload once in 2 weeks or so. So I am super interested in a less chaotic approach with uploading creatives and setting up tests continuosly.
Hey,
I was either on stage, in front of a camera or in the speaker's room during AWA 
To clear up the confusion - certain type of updates I uploaded every day, then bigger variations and quantities that deviate from daily rule. Thursday because no approvals happen on the weekend and Friday also more banners because weekend means more days without renewal.
And on Thursday I tested more concepts, and I needed at least 24 hours to see if they are better than previous banners, so I can upload 10 variations of that for the weekend.
Does that make it clear?
For me it's almost the only thing I focus on, finding good images and headlines to crush CTR's and get the lowest cpc's. I strongly believe this will beat the big media-agency-budget-bullies and the lazy-ass-spyers. In general finding good offers isn't a big problem if you have good connections with several trusted affiliate networks and when you know the product you are pushing well.
I have been wondering why this phenomena does not happen for pop impressions, why we can not talk about "Landing Page Blindness"?
I mean, I also noticed many times the quick decrease in CTR of banners pretty quickly, but the same does not happen with the LP ctr.
You can keep a LP up and running for weeks on a pop source and having the same LP ctr.
So my question is, what is different that the users get tired of seeing a banner but do not get annoyed of clicking on the same page that pops out over and over again?
I think that other than the simple "Banner Blindness" there is another factor that involves the display platforms them selves, which are not perfect in evaluating the performance of a banner and just start cutting amount of traffic too quickly.
This is more present in native platforms.
@Manu, thanks for the reply man. Yes, Matuloo told you might be near the stage but then I found Neil Patel on the way and was busy getting his signature and then kinda lost my thoughts and went somewhere else after.
I am also eagerly waiting for the AWA speaker videos, any idea when we will receive it?
Anyway thanks for the answer. Yes, its clear now. So I just realized I am being lazy by uploading once in two weeks. I can imagine how much you will be learning by testing a bunch of creatives each day compared to my slower approach. Thanks for the info!
Btw, I made a video of the AWA Bangkok event, check it out!
Originally Posted by ervin
Anyway, I also found that landing pages are more reusable. If you put a lot of work in your landing page it will stay for a longer time, compared to banner and offer (An offers longevity is more complex to predict because it depends on how things are working out for the advertiser and what he is doing at his end).
I am not sure about the reason, but this is my theory or opinion.
I think its because the banner is the initial point of contact with the customer, and it can easily get ignored once someone see it so many times. Banner is a distraction while watching content in the website, and our job is to make the distraction appealing enough so that someone clicks on it. On the other hand, a landing page is a website with no distractions where the sole purpose is to pitch the customer to sign up. If someone reach the landing page, he is already somewhat interested in the offer and his full attention will be on the landing page, so it takes longer to burnout.
The theory above might explain the drop in banner CTR but the drop in CR is not explained. Check out this article by Matuloo that explains that - https://stmforum.com/forum/showthrea...-down-and-down
That's a nice video!
I actually just had a chat with the team regarding speaker videos, interviews and such. They are working full time on it but can't give an estimate. I think you shouldn't expect them before the holidays and if some get released before it's a bonus.
There's so much footage to edit - 2 stages, post speech/panel interviews for both stages, countless separate interviews around the conference and side events. I know it's gonna be great but we do have to wait a little for everything.
Glad the post helped you!
high CTR is for sure very important to get low CPCs and backout. But the main point here is to operationalize. Come up with the a simple, scalable process that can be done without thinking. Creating banners when you need it is not the way to go. You have to get ahead of it with a regular workflow, that you can delegate over time
Originally Posted by manu_adefy
I didnt take notes during the speeches, so i already forgot some interesting stuff.
Home > Paid Traffic Sources > Mobile