Home > Paid Traffic Sources > Facebook & Instagram

---

Please explain the wacky performance of this FB ad. (11)

---

I've been having a hard time understanding what is happening with my ads. I've asked friends, posted on here, and yet when I took a look at this one ads performance below I realized there is so much I don't know. In a previous I noted that my ads have a very high CTR and high CPCs that don't drop as time goes on, however I've noticed there is a lot more that doesn't make sense. For example, with the ad below, I noticed that the CPC went up as I increased the budget. The CTR also went up as I increased the Budget. Very confusing.

I hope I'm not the only one that feels this way after looking at the graph below. We have some brilliant people on here, and I'd appreciate it if some of you would take a moment to help out. I've asked a bunch of questions below. Answer as many as you can or have knowledge about.

People have given some great insights on a previous post where I wondered about similar problems: http://stmforum.com/forum/showthread...around-.17c-!-!

This time I'm providing as much information as possible about one specific ad, with questions after.
Demo: USA, 50+, Male, Single, Interested in Women
Pricing: Bid set at .90cent CPC (Manually bid). Never changed
Note that the info graph goes in reverse chronological order.



1) Are these CPC's considered high for the demo?

2) Why didn't CPCs drop as time went on and the budget and CTRs increased. The data shows that they did the opposite. The higher the CTR and budget, the more i was paying per click.

3) Why the slow start CTR-wise the first couple of days considering how well it did later on?

4) Are the Budget and CTR correlated? If so, why did CTR increase when the budget was increased?

5) Are the Budget and CPC correlated? If so why did the CPC increase when the budget increased?

6) Would changing the bid as the campaign ran have made a difference? How?

7) Assuming the CTR's were the same, would the ad have cost less if I used a CPM model?

8) Would the CTRs have been the same using CPM?

9) What correlations are there, if any, between CPM and CTR, and CPM and Budget?

---

1) Older demos are more expensive in general
2) I'd say it's there's probably alot of competition in this demo and your bidding for position... higher bid, higher up the page, higher ctr......
3) Shit budget?
4) ^^
5) Again higher bid, higher positions
6)
7) Why don't you test yourself? In general CPM ads don't perform as well but get the data yourself!
8) Again test yourself!
9)

---

Some more stats to look at - Reach and Frequency. Can you see how they changed as well? Higher freq at low budget may contribute to the lower CTR and vice versa at high budget.

Additionally, there is one factor that will effect your campaigns that you can't get data on: Competition. It's highly possible the CPCs are floating high because there are a lot of people bidding for the traffic with profitable campaigns going, thus the premium on CPC isn't diminishing. For all we know, in competitive demos the FB algorithm may show less favour to higher CTR ads and instead focus on delivering more advertisers ads with higher charged CPCs. Whatever earns them more $$ per impression in the end. It's even possible that someone has swiped your ads and they are a direct source of competition.

Competition aside, I reckon your increasing CTR over time could be the ads sticking better and moving up the delivery column -> seen first -> CTR picks up. Lower down in the column they may get lower CTR but cost less per click, not sure if FB differentiates costs based on position but I imagine they do. You could try gaming this with bids - dupe some campaigns and let the ads fly at $0.90 and stick hopefully, then throttle bids, e.g. $0.75, 0.60, 0.45, 0.30. You may be able to get them to a sweet spot where they get decent CTR and for whatever reason actually get lower CPCs. Maybe all the high bidders dominate the top 3-4 ads and if you focus on the bottom of the stack you might get things cheaper without losing too much traction. It's hard to say how it will go without actually testing, FB is far too fickle.

Also, your CTRs with CPM would probably be horrible. I imagine the demo is so competitive that CPM ads get horrid placement in comparison to the high CPC bidders. Maybe oCPM would stand a chance but I wouldn't waste money on it.

Edit: hahaha sniped, that's what I get for leaving a tab open for ages.

---

Wow, I just had about 10 realizations since reading this. I have follow-up questions to ask, but they will have to wait till tomorrow. But here is one that immediately comes to mind. How do you test for successful images? At POF I learned a simple way to test. You simply don't need that much data on POF to make decisions. My general rule of thumb was to submit a batch of 10-40 ads. Then I would wait till the ads had reached 10k impressions each. A CTR of .100% was the baseline. Underneath that, an image/ad wasn't worth running, although I did check prosper analytics to see if an ad had done very well despite its bad CTR. I would run ads that had a low CTR but a great conversion rate. Didn't happen very often though. Using this system, I could typically test an image using $1.50 to $2.

So in essence, it was easy and cheap to test images on POF. FB is very different. The CTRs go up as your spends go up, so that testing images would be very expensive. My ad above is a perfect example. Low CTRs until I increased the budget. If you're submitting a batch of pics for testing that could be very expensive, because it seems every ad needs at least $40 in budget to determine how well it will do. So do you guys even test on FB? Do you just look for a pic that you know will do well? What am I missing here?

---

CTR is important, but it isn't really the be all, end all metric to go by on FB. You're better off going by what ad/headline/lander is giving you the best CR and EPC and making it work with the CPCs you can manage to get in any given country.

---

So I budget $5 per image. Maybe this is too expensive, I don't know. I tend to test 4 images at a time, 1 image duped 4 times in a single campaign. But the bottom line is if I can get a reasonable CPC/EPC ratio and decent volume, I'll keep it in the mix. 'Decent' volume is completely subjective of course, because as you've mentioned, if you bid for more volume, your costs may make that particular image unprofitable.

---

Well here is some bad news for me. I took 4 images that had done well on POF, and gave them $150 budgets in both 40-49 and 50+ demos, for a total of 8 campaigns. I ended up spending $1085 and earning $804 for a $280 loss. I assumed that by giving them big budgets, their CTRs would be higher, but most of them fell around .050%. I still don't understand how to test images on FB efficiently, but considering my results, I'll go back to a $4-$5 budget per image to test. I still don't know then how to reconcile a .084% CTR at a $10 budget versus a .507% CTR with a $400 budget. How do you test then on FB?

---

This is what i'm facing too, looking for answers

---

By increasing your budget, it doesn't mean your CTR will increase.

Whats weird is that what your doing is right, thats a great way to test images to see which perform the best. The wacky CTR rates you're receiving shouldnt be happening. There's no other factors coming into play?

---

I don't know why it's happening. I ran 3 ads a few months ago, they received normal CTRs, and then I paused them after a week or so. I unpaused them a few months later, and all 3 had CTRs over 1%. Things have been wacky ever since. What other factors might be at work? Anything specific I should be looking for? I'm a relatively noobish user of FB, in case that wasn't obvious lol, so please tell me exactly what to look for.

And budget does not increase CTR? It was written above that budget improves position on page which in turn increases CTR. Can you elaborate?

---

Are you referring to the way I tested at POF, or FB?

---

Home > Paid Traffic Sources > Facebook & Instagram