An Experiment: CPM or oCPM When Targeting Facebook Fans With Ads?

Promoting Posts to Fans oCPM vs. CPM

A phrase I repeat regularly when it comes to succeeding with Facebook ads is “Never assume anything.” If you assume you know, you don’t. As a result, you may either waste money unknowingly or miss an opportunity.

This is why I recommend starting most campaigns broadly in terms of targeting and placement. If you assume mobile is going to be most cost effective and sidebar won’t work, for example, you may be surprised by the actual results.

Since I’m always second guessing, experimenting and trying new things, I recently decided to test a long-held advertising habit when targeting Fans only.

My Typical Bidding Behavior

First of all, a very quick overview of your bidding options when creating Facebook ads:

Cost Per Click (CPC): The maximum you’ll pay per click on your ad. Since distribution is based on an auction format, the amount you ultimately pay per click will depend upon the competition.

Cost Per 1,000 Impressions (CPM): The maximum you’ll pay per 1,000 impressions of your ad. Once again, the amount you pay will depend upon competition for the same audience and placement.

Optimized Cost Per 1,000 Impressions (oCPM): Facebook optimizes your ad by showing it to the people most likely to perform your desired action (within your target). Additionally, bidding is automated. Your bid will change dynamically based on competition, assuring that you’ll reach your desired audience.

Since oCPM is “optimized,” the final CPM price is typically much higher than CPC or regular CPM. But — also because it’s optimized — I’ve found it’s almost always most efficient.

While I used to split test CPM vs. oCPM in particular, I rarely do anymore. I’ve simply found repeatedly that oCPM gives me the best Cost Per Desired Action (Page Like, Link Click, Conversion, etc.).

This includes targeting Fans, whether it be promoting a post or selling a product.

Second Guessing My Advertising Habits

However, I started wondering recently if using oCPM made sense when targeting my Fans only. Why, for example, do I need Facebook to optimize my audience? By reaching my Fans, I’m already reaching a naturally optimized group of people.

And if CPM costs less than oCPM (typically significantly less), might I save money by going with it instead? Or is oCPM so effective that the higher price per impression doesn’t matter?

This is what I wanted to find out. So I ran a test…

CPM vs. oCPM for Fans Test: Link Clicks

I promoted four different posts last week. In each case, I targeted Fans and email subscribers (who aren’t current Fans) only in the News Feed (desktop and mobile). I optimized for Link Clicks for each one.

These are organic posts that I simply promoted to reach more of my relevant audience in an effort to increase website traffic. Here are the four posts:

Jon Loomer Digital Promoted Posts 12-9 to 12-12

I promoted each for a very short period of time because I prefer to only have posts promoted while I don’t have another new blog post shared in my Fans’ News Feeds. They all ran from late morning when the post was originally shared to 6:00am the following morning.

In each case, I created separate campaigns for CPM and oCPM, with two different ads in each (one targeted at Fans and one at newsletter subscribers who aren’t Fans). Each campaign had a lifetime budget of $5, and I set a maximum CPM bid of $2.50 (reminder: oCPM is set dynamically).

To summarize, here are the campaigns that were created:

  • Reach Rant – Fans and Subscribers – oCPM
  • Reach Rant – Fans and Subscribers – CPM
  • Increase Reach – Fans and Subscribers – oCPM
  • Increase Reach – Fans and Subscribers – CPM
  • News Feed Test – Fans and Subscribers – oCPM
  • News Feed Test – Fans and Subscribers – CPM
  • Facebook Offer ROI – Fans and Subscribers – oCPM
  • Facebook Offer ROI – Fans and Subscribers – CPM

Since the impact of the ads targeted at the newsletter subscribers was so minimal (it was a small group and Facebook preferred targeting Fans), I will lump that ad in from this point forward and focus only on the campaign results.

CPM vs. oCPM: The Results

Here’s a comparison of results on a campaign vs. campaign basis:

Campaign Spend Link Clicks Cost Per CPM
Reach Rant – oCPM $5.00 44 $0.11 $5.83
Reach Rant – CPM $1.67 10 $0.17 $1.38
News Feed Test – oCPM $5.00 59 $0.08 $4.35
News Feed Test – CPM $1.58 18 $0.09 $1.38
Increase Reach – oCPM $5.00 97 $0.05 $5.95
Increase Reach – CPM $1.64 32 $0.05 $1.40
Facebook Offer ROI – oCPM $5.00 27 $0.19 $5.17
Facebook Offer ROI – CPM $2.12 11 $0.19 $1.33

A couple of things should immediately jump out at you:

  1. The cost per 1,000 impressions is significantly higher (3-4X) when using oCPM
  2. The budget was never reached (not even close) when using CPM

Now let’s take a look at the overall results, lumping all similar campaigns together to compare CPM vs. oCPM:

Campaign Bidding Method Spend Link Clicks Cost Per CPM
oCPM $20.00 227 $0.09 $5.24
CPM $7.01 71 $0.10 $1.37

Okay, now let’s start breaking this down:

  • Cost Per 1,000 Impressions nearly 4X higher for oCPM
  • oCPM reached full budget, while CPM reached only 35% of it
  • Cost Per Link Click nearly the same
  • oCPM resulted in more than 3X the Link Clicks

Also, here’s another important stat that I haven’t covered yet: Total Impressions…

  • CPM: 5,124
  • oCPM: 3,817

Even though the ads are being shown to 74% of the audience, oCPM is resulting in three times the link clicks.

So what we find here is that while the Cost Per Link Click is nearly the same, oCPM brings more results. In order for this to happen, oCPM must actually be optimized — Facebook must be showing my ads to Fans most likely to click on a link in order to counter the significantly higher Cost Per 1,000 Impressions.

A Note on Bidding

Since I didn’t come close to reaching my budget using CPM, it’s important to note my maximum bid of $2.50. I could not bid higher than that due to my $5 daily budget.

The budget could have held the campaigns back some. However, keep in mind that the overall Cost Per 1,000 Impressions for my CPM campaigns was $1.37. Since I didn’t get particularly close to that $2.50 maximum, I question whether it would have made much of a difference.

That said, I have to recognize it as a potential limitation. I plan to try this again with a $10 budget and $5 maximum bid (if not $20 and $10).

In Conclusion

This study reaffirms my faith in oCPM. While using the CPM bidding method may seem like the most cost effective method on paper, oCPM is so well optimized that I still end up getting better results.

Now, keep in mind that these results are for my Page only. It’s a small sample size. While I trust the results for me, you should always test to see what works for you.

How about you? Do you tend to use CPM or oCPM when you target Fans only?

Let me know your approach in the comments below!

  • Jason HJH

    Hey Jon,

    Thanks for raising this. I remember experimenting with both CPMs and oCPMs for my client’s FB page and obtaining similar results. When I micro-observed the individual ad campaigns, I noticed something unusual and I’m not too sure if this is the main reason why oCPMs cost so much more for CPMs.

    I observed those ad campaigns quite closely nearing the end of every day. To illustrate my point, suppose I run 2 ad campaigns with CPMs and oCPMs respectively from 0000 Monday to 0000 Tuesday with a daily budget of $10. Close to the end of the day around 2200 to 2300, I notice that both campaigns do not differ much in terms of CPM. However, once the day/period ends, sometimes I see stark differences like you did – higher costs for oCPMs than CPMs. This happens when the total spent for the day an hour before 0000 is way below $10. I suspect the oCPMs raises your bid prices significantly in order to show your ads to your target audience nearing the end of the day, while this is not possible under CPMs because of your bid ceiling.

    • Jon Loomer

      I haven’t observed this specifically, though I have seen something similar when starting an oCPM campaign late in the day. And it sounds like you know exactly what’s happening!

      • Jason HJH

        Thanks Jon. Now that you brought it up, I’ll try it again and see if the same thing occurs. I’ll update you again if I discover anything.

  • Grant Perry

    Thanks Jon, another great specific article. Honestly, I’ve never tested CPM v oCPM but I’ve tested a manual maximum action within oCPM and v CPC.

    My goals are almost always conversion oriented – for e-letter or paid subscribers.

    Generally I find oCPM works great to start a campaign but often (usually after a few days) I notice the CPA increase significantly and I need to switch to manual bidding for a maximum cost per conversion (action). I find the actual cost for that action to be inconsistent though. I know FB says they can’t guarantee that cost but it can be so varied it gets frustrating. If I find the CPA still doesn’t give me the required ROI I tend to move to CPC.

    So far this strategy is proving very effective but across the board I’ve noticed a significant increase in competition and therefore cost.

    • Guest

      I can’t say too much but someone told me a little behind those Ad stuff. I’d like to suggest that you try duplicating many (>20) of those ads when you launch a campaign, and you’ll find pretty extreme variances in how some cost per action can be really LOW. Hope you succeed.

  • Stan Dubin

    You mentioned that you optimized for link clicks in promoting your four posts. If you could do a blog post on how we can still do this (optimize for link clicks), that would be GREAT! I have been pulling out my hair trying to figure out how to accomplish this with the new FB ad layout.

    • Jon Loomer

      Hey, Stan! Actually, this is done very easily in one of two ways:

      1) Use the Website Clicks objective
      2) Switch to “Old Ad Types” in Power Editor and make sure the Conversion Spec is set for Link Clicks.

      • Stan Dubin

        That helped. I now can go back to the conversion specs area and work with it.

        Question: Does your current Power Editor course cover conversion specs in a way that even the dumbest of us could use it expertly? Sometimes I think I have just the right code in there to optimize for a link click and it doesn’t fly.

        • Jon Loomer

          Hah! I think so. But really, the new Objective flow solves a lot of that. Assuming they have an Objective that matches what you want to do, it replaces the need for Conversion Specs.

  • Nitish Dhiman

    Hey Jon Thank you so much !! Because few days before i asked you about this particular topic cpm vs ocpm. You can with a bang through this article. now every thing is clear to me.

  • William Thomas

    I think a lot of advertisers don’t give FB credit for how smart they are. Their goal is to maximize CPM and they have Terabytes of data at their disposal about user behaviour. That goal + that data = targeting precision like the average marketer could never have on their own. It truly is optimized.

    • Jon Loomer

      Indeed, William!

  • Gonzalo

    I’ve used oCPM with offers when targeting fans and non-fans, and sales EXPLODED. CPM or CPC… not so much.

  • Alex

    I have seen pretty much the same results. In some cases cost per action was a bit lower in CPM than oCPM, but the actions were minimal (th actual number) and it would take a week to reach a day’s oCPM results.

    My guess (totally guess, not a single fact) is that Facebook in CPM might show the ad in a very low or intermediate position in the newsfeed, thus counting as an impression, but the user never reaching and seeing it. Total random thought though.

    One more great post, thanks Jon!

  • Ashish Batra

    Dear Jon,

    Very well written article. Yes, I too noticed oCPM is always better than CPM when we want to achieve certain goals. Showing the ads to optimised users is one reason for higher CTRs. But the other big reason is inventory used in oCPM ads are better than in CPM ads. for example oCPM and CPC ads use mostly the top slots while CPM ends up appearing in 3-4 positions or even below. Though it is counted as impression the chances of clicks gets reduced drastically.
    One thing to remember here Ad gets clicked when the Ad gets noticed.

    Ashish Batra

    • José

      Very helpful. I have been wondering why I had such lower CTR with CPM. I liked your explanation.

  • Max Weber

    Great experiment Jon, I tested oCPM vs CPC with a promoted post. Target: link clicks. CPC gave me less cost per click than oCPM.

    • Jon Loomer

      Cost Per Click or Cost Per Link Click? Two different things.

      Either way, this is way you always test to find what works for you!

  • Gregor Koprivnik

    Hey Jon & Co! I’m a retailer, so I’m interested in conversions, not just clicks. So I have 2 questions:

    1) What about oCPM + conversion pixel? How could FB possibly know which audience to show my ads for max. conversions?

    2) Plus, Jon … Your test only show you can get clicks that are kind of the same in terms of CPC. But, don’t you think that folks that “click everything out there” are going to be converting at a much lower CR?

    What I would suggest is you to try a test like this that has a sale as the end goal > and then calculate CPO and see if oCPM is **bottom-line** more effective or not.

    • Clara.Freder

      when you have some free tíme use ít to get a little extra money onlìne… visit the link lísted on my profile for more informatìon

  • Brad H

    An interesting study but what was the conversion ratio on site between the different campaigns? Wouldn’t that confirm more accurately the optimisation?

  • EtienneJ

    Awesome data, love it. I’m going to try oCPM now, as you suggested on Facebook.

  • Erik Spalding

    We have found that our ROI is better when we use CPC/CPM and conduct our own optimization a series of A|B test. We have been using and have seen amazing results so far.

  • ronmartin05

    CRO (Conversion Rate Optimization) saved our campaign on Adwords. Conversion tracking involves placing a cookie on a user’s computer when he/she clicks on an ad. Then, if the user clicks on your ad and reaches one of your conversion pages, the user’s browser sends the cookie to a Google server, and a small conversion tracking image is displayed on your site. When such a match is made, Google records a successful conversion for you. Over a period of 6 weeks, Simon’s team at RDM lowered my CPA (cost-per-lead) by 27.6% simply through creative split-testing of new ad copy. It’s often overlooked but is still a very important part of the overall success of your PPC marketing with Adwords. Simon’s team will do an audit of your account for free, you can call him at 256-398-3835.

  • Wichet Amjuang

    I’m new with FB Ads, I start a few campaign to sale T-Shirt. first campaign (custom audience 18K) is CTW and 2 more campaign use Dark Post (PE) with (custom audience 6K & 13K), All use oCPM at the starting point. What I found is after keep run all for 2 – 3 Day, my ads didn’t show any impression. After that I test buy changed from oCPM tot CPC (max bid 1$) them my Ads is working.

    I can’t understand why oCPM can’t work with me. Please help to explain and how to improve/solve the issue.

    Thanks for you help


  • Paul

    Hi Jon,

    I do choose oCPM for what appears to be better results based around receiving clicks to my website. However as of late I am not so trusting of the results that Facebook states for my campaigns. As the end of the campaign draws near then the ads do appear even more optimised especially if there is still budget left on that campaign and more results always seem to appear during that last hour or so.

    However, looking at the analytics for my site through Google and also ClickTale the results never match and are always very different to each other. I am talking about an 80% difference in what Facebook says I have received as clicks direct through to my site. So for every 5 clicks that I have been charged on Facebook only 1 will be apparent through my other analytics.

    This is not down to the oCPM being allowed to charge for any actions on my advert which even when choosing Clicks to Website the Power Editor ‘always’ has a cost in this box. I always make sure I manually set the limits so that a cost is ONLY present for a website click. Yet, Facebook seems to be charging me for traffic I am clearly not receiving based on the results I am getting.

    Therefore my question is this. How much trust do you or should I have in the results that Facebook gives me for my campaigns?

  • Gary Peterson

    I was recently introduced to Jon’s page. What really impresses me is not only the quality of Jon’s comments, but the quality of everyone else’s comments. There are a lot of very talented and smart people

  • Aerohio Skydiving

    I was about to conduct this same test, thank you so much for doing the leg work.

  • divakar

    Really nice article I have this same kind of view. I have also written post about my facebook ad campaighn “” Which you may guys want to check out.

  • Tom

    Your popups are really annoying.

  • guiavila

    Very cool. Thanks for such a valuable information!

  • michaela salcido

    When I multiply the amount of ‘Cost per’ click with the number of ‘Link Clicks, why don’t the amounts add up to the total CPMs in either campaign?

  • laddu

    I tried oCPM: My cost for video views was 19 times higher than the CPM bids. For people who are patient and willing to bid lower over a longer period I would bet CPM is much better bet.

  • George

    oCPM is evil like google automatic cpc. In real auction u will never agree to hire same guy to bid for you and your oponent… Its like to hire same lawyer as oposite side… its nonsense.

  • Daniel

    I am advertising this game on FB. So far CPM hasn’t resulted so good. But now based on your advise, I’m going to try oCPM. Hope it works.

    BTW, its a FREE cricket game with very cool and innovative gameplay.

  • CT

    What is the recommended audience size for oCPM for Website Link Click?