Ad Delivery Problems (23)
Hey guys,
I've been advertising with Facebook for years and I seem to have stumbled on to an issue that I haven't had before.
On low volume, I'm getting a .9% CTR. Historically, that means that it will level out between .5 and .7 after the dust settles. Unfortunately, I haven't been able to scale the operation this time. I've been using the same method of bidding and scaling for years without running into this. (Skip below if you don't want the boring details)
- Create 4 campaigns each containing 5 ads.
- Set daily budget to $5 on all campaigns.
- Launch in the early morning so that FB doesn't kill my CTRs by delivering them all in a 10 minute span.
- One ad generally gets more love in each campaign, I pause the rest.
- Next day, I keep the two best ads running and scale them in small increments throughout the day.
If I haven't lost you yet, the suggested bid is .04-.09. I've been bidding .25 to try and get some traction with no luck. So, here's a few questions I'd like to discuss.
1) Is anybody else having this problem? Or better yet, had this problem and solved it? This is happening in every demo on every offer I'm promoting at the moment.
2) What is your experience with letting FB optimize CPC/CPM prices for you? It seems to be the only way I can get any sort of volume, but the quality seems to be low despite my bid.
3) Does adjusting your bid and budget too much during an ad's infancy screw things up?
4) Any tips on getting the ball rolling based off what you read?
5) Can putting too many ads in a single campaign potentially water down results?
Sorry for the long post. Thank you for any help!
Quick Update: Walked away for awhile to let things be. The CTRs shot up to 2% on most of them. The click prices are way above suggested but perhaps they'll level out tomorrow. I may have been editing the bids and budget too much. Patience isn't my strong suit.
Thanks for the detailed summary! Interesting question.
What's happening when you try your standard approach at the moment? Are CTRs plummeting, or is something else happening?
The standard approach didn't seem to be resulting in the volume I expected with the CTRs I was getting. This morning I upped the bids/budgets and paused every ad except for the one in each campaign that's delivering the most while maintaining a good CTR. I plan on letting it ride out through the day without micromanaging and changing the bids or budget. Hopefully hey settle into a nice groove.
Each ad is in their own campaign.
AD ONE - CTR: 2.7
Suggested Bid: .04-.08
CPC: .07
Bid: .14
AD TWO - CTR: 3.1
Suggested Bid .03-.07
CPC: .21
Bid: .25
AD THREE - CTR 2.5
Suggested Bid .03-.08
CPC: .15
Bid .25
AD FOUR - CTR - 3.2
Suggested Bid: .03 - .08
CPC: .23
Bid: .25
Ad one is the only one that's making decent progress as far as impressions go. It's in a different niche than the other three and is also one day older. Generally in that niche I can get CPCs down to .02 with a CTR that high. Fingers crossed.
As for the other three ads, I'm really hoping they pick up momentum if I leave them be. They only have a few clicks a piece at this point and I couldn't even spend the $7.50 budget per campaign last night. I can't walk away from CTRs that high though.
One other note - I let FB auto-optimize impression bids yesterday. I split test all of these campaigns in separate auto-optimized campaigns. I bid $1 on the CPM. All of the ads dropped to around .020 CTR. I was only paying .02-.04 despite my high bid. For some reason it seems like letting them optimize resulted in the cheapest/worst quality placement and traffic, though they did spend my entire budget on the optimized CPM campaigns. Keep in mind that those were the same ads that are doing great CTR-wise with a manual bid. I thought that was interesting.
I'll be back later with another update!
Thanks for doing this detailed test - I think it'll be invaluable for figuring out what's going on!
Quick Update. The three ads besides the first one bombed on me after changing the budget. I'm talking 3% CTR to .030 at the snap of a finger. Despite the low testing volume, I'm pretty confident that these ads are winners based on my experience with FB over the years. From the few days of testing so far, I've found a couple variables that seem to be significant.
Coddling these ads during their infancy seems to be pivotal to their success. Changing bids and budgets seem to be pulling them out of the ad rotation for a long period of time (way longer than it used to be.) When they've eventually served again, they seem to receive less than ideal placement and the stats take a hit that they can't recover from.
The one successful ad was a high bid that's over double the suggested. I didn't edit it since pressing play. I'm still bidding .14 and receiving .06 clicks at a 3% CTR. The suggested is down to .02 and .04 at this point so I am confident that it's not competition keeping my CPC higher than usual. I'd like to lower my bid and scale but I'm afraid of the repercussions.
I'm beginning to suspect that the first day or two requires steady traffic with no changes to the bidding. Starting them as early as possible (midnight) is still a good idea of course. I'm going to relaunch the other three campaigns and use these methods as a rule of thumb. Time will tell if I'm making any progress.
When using CPC, it seems that campaigns with more than a few ads are less likely to take off. Split testing is working better for me with 2 ads spread over way more campaigns. I'm not sure why that is but it's been consistently happening during these odd times.
Making duplicate campaigns in times like these is a solid idea as well. You can try different methods concerning your scaling approach. If you're confident that your ad should be working and you know the niche well enough to draw that conclusion, having paused duplicate campaigns on standby seems smart.
I've never had problems like this before so I'll continue to report what I find and hopefully I can figure out and share a winning method with you guys. Generally the big struggle is targeting and ad copy but this is all scaling and delivery related. Any insight and ideas are appreciated!
Hmm - I wonder if Facebook's techs are attempting to discourage people from gaming their bidding system? Advice about adjusting bids rapidly in the first 24 hours has been out there for a while now, and we know FB people keep tabs on the affiliate community: perhaps this is another move akin to the "soft ban" maneuver.
I was beginning to think along those lines. I generally don't touch my ads on day one (unless I'm pausing the immediate losers) but the complete lack of traction after half a day made me think I needed to up the bid more. All of those huge CTR ads fell to .0's the next day. I'm waiting on approvals and I'll definitely report back. The one ad that I didn't touch dropped another two cents to .04 so if these ads end up delivering like the stats suggest they should in a day, I'd say your theory could be a strong a possibility.
Howdy doo, my rowdy crew. It's a cloudy day but I have bright news for you.
After split testing HOW to split test nowadays, all of the ads that I knew were good ended up killing it.
1) One image per campaign. Two winners in one isn't going to pay for dinner, hun. Test out different ad copys in a campaign, but if you have a good feeling about an image, don't muddle it up with other images.
2) I must forbid just one type of bid, kid! If you're comfy with manual bidding CPC or CPM, it's worth doing a small test and letting FB optimize it for you. I ended up optimizing for likes after trying the other methods and I'm getting two likes on the penny for all of my ads.
3) Paying for likes is slick, but promoting clicks is the trick. I'm getting the same CTR to my lander for a fraction of the cost. Just because I'm bidding for likes doesn't mean I'm going to change my pitch to "like this page and ____." I'm getting those sweet, sweet, likes you guys know and love while still urging them to click. My ROI is huge with the cheaper the clicks and I'm getting more fans to monetize. At least give this method a try.
4) A fan page may not be all the rage, but they still work great in this day and age. Just because you have one doesn't mean that you have to update it with relevant content. A lot of people will like your page/offer in passing and never think about it again. A fan page is a great FREE way to promote the same offer again, or similar offers, directly to the people who wanted it in the first place. It's very little effort to milk a lot more money out of the users you paid for. They take like five minutes to set up so give it a try!
5) No need for news feed leads! Disable them before you proceed and you'll still succeed. The CTRs are through the roof and the results can be great for the news feed ads but the price is just far more expensive for the same results. I know a lot of people who accidentally forget to disable them in the ad making process and that's VERY easy to do. Unfortunately, the option to disable them once the ad is submitted doesn't exist. So at the very least, split test them if you want to use them.
In closing, the same campaigns I had high hopes for at the beginning of the thread are all chugging along at 1% + ROI and penny clicks and under. Keep testing and don't give up if you have a feeling something should be working. Every batch of ads you submit and learn from gets you one step closer to success.
Fantastic tips trip from your fingertips! A Facebook ad may have made us sad, but with gems like these I concede the need to plead for CPCs to drop will stop - or at least slow, as our profits grow. 
What did you end up concluding on the bid change vs CPC issue?
I forbid you kids from changing your bids! This bumble bee guarantees lower CPCs that will have you buzzing with glee. These results made me slap my knee, wet my tee in pee and scream I PAPI! ... ahem.
I replicated those campaigns with the same the bids and they settled in at a cozy .04 after a day of leveling out. That's pretty much where I wanted to be. When I was testing that out, I realized that I should probably try out optimizing for page likes since that's such a huge part of my business model. A page like for .01-.03 is worth way more to me than a high quality click for .04. That's assuming there actually is a difference in quality and placement though.
Also, I wanted to clarify what I meant by running one creative per campaign. Running a bunch of creatives in a campaign to see what works the best is a good way to weed out the losers, but one constant I've found is that replicating the winner in a different campaign by itself and giving it a fresh start yields better results. Maybe it's because the shit ads that you paused still have stats attached to the campaign and affect the quality score in the algorithm? That's how I'm spinning it to myself so that it makes sense at the very least.
The one problem that I'm running into is actually maintaining CTRs for more than a couple days. My campaigns combined barely reach 10% of the available audience at a 2.0 frequency before running out of steam. I've checked my insights and they're not getting negative feedback so I'm having trouble sorting this out. Starting a new campaigns every few days is the only way I've been able to stay where I want to be. That's my current struggle. As always, I'll report back with any findings.
interesting... so using a page likes campaign sending them to your fan page - or an optin form inside your fan page.... but getting cheap clicks via the like ad.
you cant use fb conversion tracking with this method however....
You couldn't throw it in the iframe HTML?
when they ask you what your goal is... if you say page likes, it doesn't give you the option to use fb conversion tracking....
but that might be a trick to get cheap clicks.
i just setup a likes ad.
destination is the woobox optin page on my fan page.
guess you gotta track using your own tracking system.
Understood. The static iframe tab has some pretty good free tracking in place so if you're direct linking from the tab you could use that alongside the network data if you know and trust who you work with.
Why hello there. Gozzo here with another update to what has turned into a minor follow along. We left off with me being unable to keep my engagement cost at the same level for more than a few days. With a few helpful STM tips and a lot of split testing, I've managed to come out on the positive end of this situation.
- I took the creatives that were already running in the other campaign and made them again. I used three duplicate images per campaign with three duplicate pitches. The same ad three times. There's a huge thread on this method that's still going strong in this section of the forums.
- I created 4 different campaigns using each image as well. I wanted to split up the age demos and see if I was targeting too broad by going 18+ on everything. So 4 images, 4 campaigns per image, and 3 duplicates per campaign. Each campaign with a budget of 2.50.
- I had seen people having success start ads up around 4-6PM as well as 12:01 at the beginning of a new a day. When my ads got approved, I couldn't be patient so I pressed play on 4 campaigns. They were all the same ad image with a different age group.
- Each campaign had one ad get a majority of impressions, higher CTR, and lower engagement cost immediately. So I deleted the 2 shitty duplicates in each campaign.
- After pressing play at 6PM on the ads, I was already down to 4 campaigns and 4 ads. So I started another creative with 4 age groups and 4 campaigns.
- By 7PM I had repeated the last step. I now had 8 of the new ads running. One duplicate per campaign outperformed the rest. So I did it one more time!
- I threw two campaigns up from another image I used. Same thing happened but the click cost wasn't as good so I stopped there.
- You know what's interesting? As I spent more money over the course of the night, my older campaigns slowly dropped to day one costs and CTRs! You'd think the new ads targeting the same fans with the same offer would make the old perform worse, but that's not the case. I originally said I'd have to strip the welcomes during the approval process but instead, I made a duplicate fan page so I could keep traffic moving at all times.
- After the dust settled, my stats looked like this:
- .00 cent CPC on US traffic on my best ad? I'm getting a little over 2 clicks per penny for that one! That's amazing! My older ads continued to tear it up as well. So in conclusion, here is the biggest variables to take away from this post:
1) 3 duplicates per campaign is KEY once you've dwindled yourself down to the best images working for you.
2) Create more than one campaign per image if the demo is large enough! Split the ages up a bit more! More impressions without burning out your ads nearly as quick!
3) This morning, the trends are continuing. My older ads are getting cheaper and higher CTR. It's as if Facebbok is rewarding me for scaling. It makes sense. Why hook somebody up with the best placements and price if they're not spending more to scale the quality?
Any questions? Any more tips or suggestions from you guys? I'm going to jump into another demo that I don't know as well and post results here after applying everything I've learned/tweaked over the past week or two. It's a campaign that I was sure I'd make work but didn't manage to get the cost down on. This entire thread was based off of my gut instinct that I knew I was on the right track. So lets put it to the test again!
Thanks for reading!
Quick little update: I'm scaling up campaigns one by one. The higher the budget, the cheaper the clicks seem to be getting. All CTRs are between .900 and 3.0 on volume. My ROI is 793.8%. Any FB users should read what mistakes I've made and testing I've done in this thread. With a unique angle, good creatives, and solid strategy, you can make it happen.
EDIT: I haven't monetized any of the fans that I've acquired, which overtime will easily quadruple the ROI.
gozzo.. good stuff man.
so when you're starting a new campaign.
how are you setting that up?
same image, different ad copy? same ad copy, different images?
really curious.
Batch 1) I used the same ad copy with different images in each campaign. Round one was finding the right images (testing different images, borders, contrast, and whatnot.)
Batch 2) I used the winning images with four different ad copies in each campaign. The same ad copy won in every campaign.
Batch 3) I made campaigns with three duplicates per campaign. I also fleshed it out and split up the ages I was targeting for each campaign so I could scale them where they were performing the best.
Every campaign was $2.50 a day budget to start. That was enough for me to get the data I needed. So I have 16 new ads running. 4 different images in 4 age groups with the one ad copy that was outperforming the others by a huge margin.
Also like I had said, my ads for my last campaign are being rewarded with better placement. The CTRs are skyrocketing and the CPCs are plummeting. Those ads are the same ones as this batch going to a duplicate fan page. The difference is that I didn't use duplicate ads in the same campaign in the last batch and I didn't split up the demos.
Smooth sailing so far. The ROI is only going to get bigger when I promote the offer to my fans. A large majority of users click "like" when they see the ad but never look at the content the page is promoting.
I'll be sure to update with how my day turned out later on tonight.
what kind of ads are these?
(non power editor?)
when fb asks for your goal, what are you putting in?
so $2.50 per day to start.
same ad copy - 4 different images in 1 campaign
just a little confused... trying to make sense of what you said so i can run a lil test .... LOL :-)
If you already know what ad copy and creatives are good, you wouldn't need the majority of what I said. I was just trying to explain the process of split testing I went through.
I'm bidding for likes but encouraging the click in my ad copy.
One Creative, One Ad Copy, Three of the same exact ad per campaign.
As soon as the campaign is live, keep your eyes on the stats. One duplicate will outperform the other two in each campaign.
Delete the other two. I'm not pausing and leaving them in there. I don't want the two paused ads to effect how the algorithm scores a campaign.
As of now...
Best of luck! Update me on your results.
gozzo.
thanks dude. i have the basic ad copy and images down... haven't tested enough different ad copies yet, but i have good metrics to go by.
curious how you're tracking all this... the like ads send ppl to your fan page (or optin tab inside the fan page)
but how are you tracking which ad is making sales?
since fb is only showing likes as the metric.
with the other ads, you're able to use their tracking pixel.
I'm going at it blind. I've had different iframes setup for different ads before and used the subids on the network for quality checking but this demo generally pans out evenly across the spectrum.
My main concern is getting as many likes for as little as possible right now. The profit from the ad is just a cherry on top of viral facebook post profit.
Home > Paid Traffic Sources > Facebook & Instagram