Plenty of Fish Case Study – CPM Bid Effects – Results
Posted by Riley Pool | Posted in CPM, Case Study, Plenty of Fish | Posted on April 11th, 2010
21
Earlier this week I made a post about doing a case study on the effects of different CPM bids on Plenty of Fish. I literally gave away an exact campaign, including ad copy, images, targeting and bids. I tested different CPM bids of 15c, 25c, 35c and 45c. I ran this campaign from Sunday – Wednesday.
|
Campaign with 15c bids
|
Campaign with 25c bids
|
|||
| Impressions | 39,728 | Impressions | 65,592 | |
| Clicks | 41 | Clicks | 57 | |
| Avg. CTR | 0.103% | Avg. CTR | 0.087% | |
| Total Costs | $5.96 | Total Costs | $16.40 | |
| Leads | 4 | Leads | 6 | |
| Conv. Ratio | 9.76% | Conv. Ratio | 10.17% | |
| Revenue | $11.60 | Revenue | $17.40 | |
| Profit/Loss | $5.64 | Profit/Loss | $1.00 | |
|
Campaign with 35c bids
|
Campaign with 45c bids
|
|||
| Impressions | 278,539 | Impressions* | 447,061 | |
| Clicks | 283 | Clicks | 413 | |
| Avg. CTR | 0.102% | Avg. CTR | 0.092% | |
| Total Costs | $97.49 | Total Costs | $201.18 | |
| Leads | 21 | Leads | 22 | |
| Conv. Ratio | 7.98% | Conv. Ratio | 5.58% | |
| Revenue | $60.90 | Revenue | $63.80 | |
| Profit/Loss | ($36.59) | Profit/Loss | ($137.38) | |
* This campaign hit it’s $50 daily spending limit everyday it was active. If this limit would’ve been higher I would’ve received more impressions.
I was pleasantly relieved to see that my assessment of the Plenty of Fish self-serve advertising platform was wrong. They absolutely do favor the higher bids and will provide more impressions to higher bids. Ben from Plenty of Fish noticed my last post and if you didn’t happen to read the comments, he provided some valuable insight to their platform. His quote is below.
Your CPM will have an affect on your CTR because we serve ads based on CPM price. The highest bids get their ads shown to the user first, followed by the next highest and so forth. So if your bid is too low, then the impressions you receive will be from people who have already seen 20 ads (thus the likelihood of clicking and converting are low).
Also, your CPM will have an affect on your traffic flow because the highest bids get their ad quota fillled first, followed by the next highest and so forth. So if your bid is too low, you might not get the priority to start getting a lot of traffic to your site. This explains why CPM is so critical in highly targeted campaigns.
After reading the quote from Ben above and assessing my stats from the 15c and 25c campaigns, I’m wondering if I could use the Session Depth targeting criteria to my advantage. I realize that my stats on those two campaigns aren’t statistically significant, but it got me thinking and made me realize something. When I first login to Facebook, I’ll be checking messages, notifications and status updates first thing and won’t pay attention to the ads at first. Then as I start just dicking around on Facebook, I notice myself paying more attention to the ads.
So this coming week will see me doing a case study that split tests only the session depth.
Related posts:
lol read the thread markus already told us this
[...] This post was mentioned on Twitter by Riley Pool. Riley Pool said: [BLOG]: Plenty of Fish Case Study – CPM Bid Effects – Results – http://bit.ly/9IKu4X [...]
@Ja Ja Do you take everybody’s word as the truth?
Great analysis Riley. I should do more testing on POF – last time I ran volume there, I just couldn’t crack it
Nice case study. I applaud your attempt to spit testing using POF, but a few things I’ve noticed that hopefully will help you in your testing.
Time of day can have huge swings on CTR so make sure to launch your ads at the same time and maybe have them show evenly instead of running through budget.
Also, when you launch multiple ads that are identical (targetting, copy, image) you’ll notice that they can often have drastic differences in performance. I suggest you launch at least 3 identical ads and average out their results for a more accurate view of how that ad would perform.
Keep it real.
Awesome CS Riley. So what would you recommend to make this profitable? The above results are pretty consistent with my ROI on all my POF campaigns. It’s interesting how your CTR pretty much stays the same but your conversion rate steadily drops.
You’re right regarding session depth, time for me to do some tests myself using that parameter.
Ben suggested playing around with login counts too, any thoughts on that?
@GoogleSearchSux CTR wasn’t part of my case study. I was focused solely on CPM vs. Number of Impressions. However, I do completely agree with you on the CTR vs. Time of Day and having multiple ads to get a true feeling for CTR.
@DH I would recommend using your best images and ad copy. Using that strategy, I’ve made this exact offer profitable.
@Kang No clue on the login counts. Never tried testing that. It’s definitely something I can try out at a later date though.
Well this is interesting. Isn’t what Ben said a complete contradiction of your results? Your results show that the conversion rate drops as your bids increase whereas Ben claims higher bids lead to ads being shown to users earlier, which should either have a positive or no effect at all on conversions.
Yea, pretty much. Real world test results often contradict theory. Therefore, the only way to find out how something truly performs is to test it out.
Yes login count is very effective, especially for dating ads.
With a higher CPM and impression flow, you’ll have to tighten up the targeting to ensure that your $50 daily spend is going to a qualified audience – I believe the targeting is all the same in this case study.
I’m very interested in the Session Depth targeting case study that’s coming up. I have my own suspicions on how that works but it’d great to have them confirmed/denied
@Ryan, agreed.
Ask yourself, when do you think ad blindness kicks in? I think Riley’s got it right w/ session depth (pending the upcoming case study) but what about log-in count? Are brand new users more prone to ad blindness or users that have been POF members for years? Or are brand new users extra-prone to ads vs. older members?
Just keep in mind that if you put a low log-in count ceiling on your campaign, it makes it impossible to scale.
Nice Riley, very stimulating case study.
With a higher CPM and impression flow, you’ll have to tighten up the targeting to ensure that your $50 daily spend is going to a qualified audience – I believe the targeting is all the same in this case study.
I’m very interested in the Session Depth targeting case study that’s coming up. I have my own suspicions on how that works but it’d great to have them confirmed/denied
@Ryan, agreed.
Ask yourself, when do you think ad blindness kicks in? I think Riley’s got it right w/ session depth (pending the upcoming case study) but what about log-in count? Are brand new users more prone to ad blindness or users that have been POF members for years? Or are brand new users extra-prone to ads vs. older members?
Just keep in mind that if you put a low log-in count ceiling on your campaign, it makes it impossible to scale.
[...] 6 – Stats Recap and Going Forward How I Generated $1,700,000 in Auto Sales Despite a Weak Economy Plenty of Fish Case Study – CPM Bid Effects – Results How I made $7,144.00 using TrafficVance. NEW Facebook Ads Plentyoffish Self Serve [...]
[...] 6 – Stats Recap and Going Forward How I Generated $1,700,000 in Auto Sales Despite a Weak Economy Plenty of Fish Case Study – CPM Bid Effects – Results How I made $7,144.00 using TrafficVance. NEW Facebook Ads Plentyoffish Self Serve [...]
Can you explain what Session Depth means?
Thanks.
Every time a user loads a page on POF, their Session Depth increases by one. Hopefully that helps.
[...] 6 – Stats Recap and Going Forward How I Generated $1,700,000 in Auto Sales Despite a Weak Economy Plenty of Fish Case Study – CPM Bid Effects – Results How I made $7,144.00 using TrafficVance. NEW Facebook Ads Plentyoffish Self Serve Advertising. How [...]
[...] my previous Plenty of Fish Case Study, I tested CPM bid vs. Number of Impressions and received results that shattered my original assumption and the results turned out as they [...]
Great stuff! I never thought of using plentyoffish. I use mainly CPV traffic like mt, tv & linksador