This is not a TrafficJunky Follow along (yet) (14)
If banner blindness is kicking in too fast, try lowering the frequency cap. Make your banner headlines simple, and easy to read. Use bright colors.
If you have a winning banner style then just keep adding more images to test on it. Find out which ad -> lp combo is converting.
Sometimes 1 ad performs better with 1 lp and then the other ads might perform better with other lps. Then you should split these combos into separate campaigns and test them.
Since you are using
Sympathy - that's a frustrating situation to be in.
How different are your landing pages?
Unfortunately I couldn't read your LP data - could you upload to an external service and then link here? But in the kind of situation you're describing, my usual hunch would be that you need to test wider.
Is there anything at all that you always do the same every time? Similar copy, similar images, similar designs? Are you always running rules landers?
One thing you might want to try is running your banners enough to cut on ROI rather than CTR. I usually only use CTR as a gateway - killing banners with really poor CTR - rather than cutting and choosing solely based on CTR. It's all too common for a high-CTR banner to also have low CVR.
@philme Yes, I am now using the drilldown reports in
This is one of the features I miss the most from Prosper. I know the
I am now seeing a drop in clicks from the traffic source. I might need to increase my CPM bid, which is currently at around $0.40. This might push me in to red number territory...
@caurmen
I'm using three, pretty different LPs, design-wise. But in the end they are mostly the same: rules and pre-qualification questions, but modified to fit my angles.
Damn, I didn't test the images when I uploaded. Here are the reports so far:
Banner reports
LP reports
I am cutting images based on ROI actually. I used to have these images where I could get like a 3% CTR but they didn't convert at all.
What do you mean by testing wider? Testing very different LP concepts?
Thanks!
Time for another update. Unfortunately I couldn't update during the weekend because I often need to travel to other cities because of my job (that's my other immediate goal, quitting my $300/mo engineer job!)
Here's a day by day update of the progress:
21/3
Cost: $19.35
Rev: $16
P/L: $-3.35
22/3
Cost: $20
Rev: $12.8
P/L: $-7.2
23/3
Cost: $7 (ran out of funds!)
Rev: $6.4
P/L: $-1.4
I am now back in the red. I am getting to the point where my campaigns usually start failing. And this is usually the point where I start running around like a headless chicken trying to save the campaign:
There is one ad that is bringing in most of my conversions, so I see what kind of ad it is and started adding variations of the ad, unfortunately they are not working quite as well. Then I add more banners. I should be killing those that are not working but for some reason I feel as if I don't have enough data, so I don't. Now I've completely messed up my campaign and I've got a bunch of banners, angles and LPs and I've managed to fuck it up once again.
During all this time running campaign after campaign I have not managed to come up with a system, so I hope that I can work something out during this follow-along.
In my mind I should be doing this (and please add to this process and make suggestions):
-Find a couple of offers that I've asked my AM is working.
-Come up with 2 angles.
-Make 10 banners: 5 for each angle. (I still need to work more on this. A lot more.)
-Rotate 2 LPs initially. Copy and modify for each angle, so I end up with 4 LPs, 2 for each angle.
-Start sending traffic.
-Start culling banners and LPs once I have enough data, using Caurmen's kickass tips.
-Start culling angles once I have enough data. Come up with new angle to throw in the rotation.
-???
-Profit!
Usually, the part where I have a problem is where I keep the winning LPs and banners and start making profit. Right now my winning ad should start dying because of banner blindness.
I'm very tempted to pull the plug on this campaign and start all over again.
I've got a couple of questions:
-When you test multiple angles on a single ad zone, do you create a separate campaign for each angle or would that make you compete against yourself? Right now I'm adding all my angles on the same ad zone.
- Sometimes I've got banners with a really shitty CTR but it converts. For example, my third best banner (conversion-wise) right now has a 0.15% CTR . I know now not to go by just banner CTR. The only way to avoid this is to hit a certain number of impressions. Should I cull according to impressions or banner ad spend? How many impressions are enough?
Thanks everyone for your help!
I funded my account yesterday and could not get any data for the day. Someone increased their bids and put me out of the game. Today I noticed and increased my bid.
I started to see some traffic coming in, I even got a conversion!
But I just got the dreaded "The advertiser has requested you stop sending traffic to this offer" email. 
Fortunately it's not so bad. I was using the SOI version of an offer that also has a DOI offer with a slightly higher payou, so I'll just move to DOI.
Couple of questions:
- What's your opinion of SOI offers? Am I more likely to get optimized off the offer due to the fact that these types of offers probably don't convert well to paying members for the advertiser?
-My best performing ad was an IM style ad. What's your opinion of these ads? High conversion, bad quality? I'm thinking that maybe guys sign up expecting to chat with the girls on the site , but if that's not real it turns to bad leads for the advertiser?
Cull by ad spend, every time. If a banner's making more than it's costing, that's all that matters. I'd let the spend get to at least 2x payout, then start cutting on ROI using the confidence interval calculator I talk about in Part 10 of the Getting Started Guide.
Do all your angles work equally well with your lander? I normally find that you need a specific angle-targeted lander as well as angled banners: that preserves the user experience through the entire thing and makes sure you're never bait-and-switching on him. Multiple angles to a single LP tends to be a rapid path to a rather generic, uninteresting lander.
Thanks caurmen!
I've been trying to keep my angles tied to the LP. Generally , this involves modifying the text on the LP to match whatever angle I'm promoting. Still, I almost always end up with the same rules or pre-qualification questions LP. I am aware that's one of my weak points and I need to improve that.
I will start culling ads using that calculator. Unfortunately I think I now need to collect more data because I had to switch my offer.
Alright so now it seems that I may be forced to start this follow along all over.
The lack of traffic yesterday wasn't because someone outbid me. I emailed my traffic source today and they said that the adspot is down and they are not receiving any traffic there and will not receive any until further notice.
This is the first time that something like this has happened.
But I'm gonna take this opportunity to start a new , more organized campaign, following the system I laid out above. Stay tuned...
Ok, so last night I started a whole new, more organized campaign, since the traffic source pretty much forced me to.
I took my two angles, made 5 banners for each angle and put them in the same campaign. Took two LP models, modified them for each angle and now have 10 banners and 4 LPs:
-5 Banners for Angle A going to 2 LPs modified for Angle A
-5 Banners for Angle B going to 2 LPs modified for Angle B
While I was doing this last night something started to nag me:
Should I be making 5 completely different banner concepts for each angle?
Basically what I did was make a banner template for Angle A, found 5 different images and used the same template, different images. I'm not sure what I'm testing there besides the efectiveness of each image and the angle. Maybe I should not work with a template , but instead try out different fonts, colors,etc?
But then I would be adding more factors to the equation. I could complicate things and end up with 15 different factors to test.
Anyone care to share their thoughts on this?
My usual approach is to test image first and other things second. It has been my experience, and is also generally received wisdom, that in dating and adult verticals the image makes the most difference to the banner's performance, at least CTR-wise.
I've finally got some good news:
The good news is that I'm in the green already on the first day of the campaign!
28/3/
Cost: $1.93
Rev: $12.77
P/L: $10.84
The bad news is: I'm the top bidder on the site and I'm barely getting any traffic. It's converting (maybe it was a first day fluke) but the traffic is so low I'm thinking there might not be much of a future in this campaign, at least on this site. But hey, a profit is a profit!
Maybe I can use this data to copy this to another traffic source? We'll see.
Right now I'm thinking ahead. I'm kinda worried that I might get optimized off the offer again. I'm going to be incorporating more "pre-qualification" conditions in to my ads and LPs, like adding a "25+ year old only" notice to my creatives. Maybe once I have a proven converter I can keep the good offer for the 30+ leads and look for a "burner" offer for 18-29 leads.
I'm gonna repost a couple of questions I previously had, see if I can get some input from fellow STMers:
- What's your opinion of SOI offers? Am I more likely to get optimized off the offer sooner or later, due to the fact that these types of offers probably don't convert well to paying members for the advertiser?
-My best performing ad was an IM style ad. What's your opinion of these ads? High conversion, bad quality? I'm thinking that maybe guys sign up expecting to chat with the girls on the site , but if that's not real it turns to bad leads for the advertiser?
As you say, a profit is a profit! And you never know - sometimes these low-volume campaigns can just sit there dripping cash for months.
Does the site have any other placements, or are there any similar sites you could scale to?
I've never seen any strong relationship between SOI / DOI and chances of getting optimised off the offer, but I'd be interested to hear what other people think on that.
AFAIK chat style creatives garner high CTR but generally lower backend CVR rates, both on the lander and offer. If they were working for you then sure, run em - I have no idea if they tend to bring low quality leads!
Home > Adult Traffic (NSFW) > Follow-along Campaigns