Home > Paid Traffic Sources > Facebook & Instagram

---

Mobile Facebook (17)

---

Facebook generally speaking can be quite expensive cf. sources like mobile display, PPV, etc.

But, it can offer very high quality traffic due to the precision of demographic targeting that the aforementioned sources simply cannot compete with - not even close.

When it comes to gaming apps for example, FB on average gives similar CPIs to display but the cost per loyal/engaged user is lower - in other words, FB traffic is often of high quality for advertisers. This isn't always true but at least for games this is well known from huge amounts of data.

You can do $0.30 app installs just fine as long as you hit the right countries and have very good ads - but this is true for all campaigns.

1) Yes. But it can take a lot of testing, good angles/ads, and testing of CPC/oCPM/CPA bidding. Virility would help a lot.

2) CPC so that you get actionable backend data for your dollar.

3) Not possible to avoid. FB charges you for actions users perform on the ads, e.g. likes, shares, comments, and link clicks. If you bid CPM/oCPM then you pay per impression - so don't get double charged for a user performing multiple actions. However, it's still sensible to start with CPC and test oCPM later, especially if ads have a high % of social actions.

4) No, not really. Mobile = NF only.

5) I'd recommend starting with the default and going from there. If it's a low CPM country your payouts are low you can drop the bid but it's really not worth playing with much at first. If the country has cheap traffic the CPM costs will drop anyway, so having $5 there doesn't mean your cost per conversion will be anywhere near that. See my FB guide and the Venn diagram on oCPM. If you bid on actions only, you will only deliver ads to a segment of your target audience. If you bid toward multiple ones you are really not taking advantage of this oCPM audience segmentation. Better IMO to split test bidding on actions only, clicks only, etc. in separate campaigns.

6) There really is no best anything but app installs make sense over lead gen, games offer much higher payouts usually, so games/utilities/misc apps make sense. Ask your AM for recommendations on what offers are doing volume and converting well. If something is averaging 0.5% conversion rate network wide, you might be able to do 10-20% on FB.

7) FB data isn't real time and sometimes the stats in different areas are out of sync, at least at first. The CPC will of course fall down over time as the ads get traction - the CPC bidding algorithm/system is not "you pay exactly what you bid per click".

---

Your lack of conversions has nothing to do with bot traffic.

Ignore fixed price, that's not applicable to your current campaigns. It won't be an advantage to you and isn't for fixed click pricing. If it was, everyone would be using it and it would lose FB money.

---

I've found what was wrong. Checked all the links and settings. Simply no postback at affiliate network. Don't slap me, I already did. So there are conversions in affiliate network, but not all of them are from Facebook, as there was a little traffic from ad networks, which Im stopping to work with. Lets see what happens tomorrow.
As for now, instead of stopping that "0 conversions" campaigns, I increased the daily budget for them twice.
Funny thing is that I see conversions even within default ads. I mean I have only 1 ad set in each campaign, and only 1 ad in each ad set. All the ads are the same for all campaigns, no angles, just a general decription of offer and a general image.
Maybe it is time to make more ads in ad sets, if I see a winner?
As for now, I'm making a dark post (unpublished on my page). When a user likes this dark post, do his friends see these post in their NF? Or they don't because its unpublished?
Im reading STM Facebook posts now to faster my learning and get faster results. And I thought why to make dark (unpublished) posts, if you can make it published, so that other users (not from paid ads) could visit a page and interact with my post? And I found this:
Is it true, that you can't target only mobile, when promoting a published (not dark) post?
Thanks

---

Makes sense. Now I'm thinking of making a 1 post for every offer in English in my page (like for everyone), all the other GEO posts in different languages could be hidden. Or another thought is to make a similar page for every geo I work with and put 1 post for every offer to that special geo in their language. But not sure if its a good idea to make sooo many pages now, wheather Facebook will like it or will it be worth of time. Maybe its too early to think of such things on the 4th day of working with Facebook.

I guess I see a winner "geo-offer" now, and you know what - it was that "0-conversions-bot-Facebook-traffic" from above. From rags to riches.
So now I gotta milk it as fast as possible, according to STM) I'm increasing my daily budget gently 10-> 20-> 50-> so on. And I increase it among 00am-1am. Not sure why its best, but Ive read it somewhere here in threads and Im sticking to it. I dont want to make big jumps in daily budget, not to fail if smth goes wrong. Right now im at $50 a day for that campaign. Going to increase it today, when time comes (00am).

Wanted to know what is the best way to scale more: just daily budget increase or adding more ads? To make new campaigns? To make new ad sets in the existing campaigns? Or to load more ads to the existing ad sets? The last thing I want - is to break the working campaign (right now its 1 ad for an ad set, and 1 ad set for a 1 campaign). I tried several biddings (cpc, ocpm both manual and deafault) and seems like default ocpm works best for my ads: more traffic and cheaper cpc (but not much). Surely not enough testing, but I will continue.
Thanks

---

Your budget increase tactic is bang on.

For scaling, budgets are set at the ad set level so just a) increase ad set budgets and/or b) make new ad sets within the same campaign for new ad variations.

If you want to try something different e.g. different targeting, I would make a new campaign.

---

So, I increased the daily budget. CPC campaigns went negative ROI (they were +ROI) first 2 days. OCPM campaigns CR felt a bit, but is still ok and positive. It sure cant be banner blindness as I've ran not more than 30k clicks total.
Maybe it is the difference between weekends and weekdays.

Several questions to master Facebook:
1) Do users who interacted with my Ad Post (click/like) see this post again? And if yes, how to avoid it?
2) When adding a new Ad Set to a campaign, does this new Ad Set connects somehow to other Ad Sets? Maybe it gets the "cost per click" from nearby Ad Sets for tha start? Maybe this new Ad Set will not be translated to users, who have already seen the ads from nearby Ad Sets from this very campaign? Or it begins its separate new life and starts to show the ads to everyone who is in target of the campaign?
3) Is there a way (when making a new campaign or an Ad Set) to exclude users, who already interacted (clicked/liked) with my other dark posts from other campaigns or form the campaign Im adding a new Ad set?
4) Does the "cpc" depends on targeting (when im bidding ocpm), if we are talking about 1 geo? Or it does not matter whether I target broad or narrow audiences, the only things that matter are: competition and CTR for my ads?
Thanks

---

1) Yes, of course. You can avoid this is by tagging people into custom audiences when they click through and relaunching campaigns regularly with that audience added as an exclusion. The day these go dynamic will be a joyful one for sure! You can also exclude people connected to your pages pretty easily. Pretty sure that function is dynamic - so target people who are NOT connected to your page.

2) To the best of my knowledge it starts fresh with no interaction with any other ad sets, other than potential account-wide effects. Aged accounts often don't have to bid as high to get traffic, but I can only really speculate on anything further to that.

3) See #1.

4) Facebook has a tendency to increase pricing as you get more and more specific... but CTRs may jump up concomitantly and lower pricing so it's a give and take. The CPMs depend on competition and what other users are bidding so it's a mix of things. My gut however tells me that FB will charge a premium for tight targeting giving CPMs higher than you would get for the same users being hit with a broad campaign.

---

here share my ideas of you questions you mentioned above,
it's really difficult to run a low payout campaign like battery. but it's possible to be a winning campaign if you have great creatives and enough experience. in my experience, I always start from ocpm, because it can have a optimize for the "like" . The proportion of "like " is different in various countries. so you could use ocpm to reduce your cost of "like". and I suggest you begin with a popular offer with a high CR, maybe the utility type is a good choice.
it's my experience for your reference.

---

So, my profitable GEO CR dropped. Not sure why that happened, I did not change anything in the campaign, but its about breaking even now (yesterday was -$50). Now Im making new ads and testing other geos. Spending like 2k per day.
I decided that I want to track conversions in Facebook reports. I've read the guide, but still have some questions:
1) Is there a difference what to choose when making a campaign: "Clicks to websites" or "Website conversions"? I've made campaigns with "Clicks to websites" and I still can see, that I can "create pixel" when making ads. So whats the point in "Website conversions" type?
2) If I choose "Website conversions" and do not use tracking pixel, that means that Facebook OCPM algorithm will be smashed? As it should try to bring more conversions (as an objective), and if it will not see any conversions - it should be stopping the traffic from some audiences (even though they could be profitable for me). But Facebook will not be able to define (ocpm) good audience for me and spread traffic to everyone, trying to see any conversions. Or am I wrong?
3) I asked my AM if I could track conversions with them. She told me to send the conversion code from FB. But what if I have already 20 campaigns. What exactly tracking pixel should I send to my affiliate network? Or it is all the same for all my campaigns (I doubt it).
Thanks!

---

1) The website conversions objective just sort of locks you in to using a pixel if using the native interface or power editor without spreadsheet uploads.

It seems they have also split pixel usage into a conversion tracking and optimisation pixel but honestly I wouldn't worry. Just use either. It won't change ad performance.

2) The oCPM algorithm will initially work exactly the same. However, over time (i.e. a week and longer), it won't expand your audience to reach users similar to those who have converted in an attempt to lower your CPA. How well this works depends on the campaign. Quite often us performance marketers are testing new ads and improving them so rapidly that waiting several weeks for an oCPM-induced CPA decrease is not a priority. So, without the conversion pixel your audience will stay similar throughout, with unknown consequences.

3) Make a new conversion pixel for a new offer. Conversion pixels are unique for each one you make. Custom website audience tracking codes are unique for your account and can be used everywhere (since they filter by URL rules), but you don't use these for conversion tracking.

---

Impressive amount of detailed answers. Thanks Zeno

---

Thanks for Zeno's deep reply~

For conversion tracking pixel with affiliate offers, it seemed like 202 is now workable directly for FB pixcel(image pixel and smart postback). but for imobitrax and Voluum, they could not possibly solve it so far, right ?

---

Tracking with Facebook and a tracking system are, and should be kept, completely separate.

Use server-to-server tracking where possible with your tracking system as it is the most reliable and widely adopted by advertisers and affiliate networks.

Then, if the advertiser allows it, place your Facebook conversion pixel. This must be client-side so there is little point in connecting it with your tracking system and using client-side tracking there, just so that you can load a Facebook conversion tracking code at the same time (using iFrame/JS pixels from the tracking system).

I repeat, sacrificing accurate server-to-server tracking in favour of a JS/iFrame code just so that it loads the FB code in parallel is not an approach I would advise. Only do this if the advertiser doesn't support S2S...

Either you can place the FB code separately at your network, or send it through to your AM/the advertiser directly, or they won't support client-side tracking = FB pixel not usable.

---

That would be a foolish thing to do.

---

Hi!
I still could not find out how to send conversions to Imobitrax. I don't know if it is possible to make a global pixel (code) in Facebook for an offer. I see I can make Facebook pixel under every "ad name". So I did several pixels for different ads (for 1 offer), inserted them in my Imobitrax (in different ways: img, link in Get, that I took from Facebook code), but It did not work(
As for the campaigns:
As long as I increase my daily budget for the campaigns and create similar campaigns for this offer - my ROI dercreases so much. Now its about $3k spend per day and ROI falls down(
What is the best strategy of scaling? Seems like both increasing daily budgets and creating similar campaigns are not working (ROI drops). In the begining ROI was like 35%, 2 days ago it was 17%, yesterday it was less than 10%. Today seems like I be at negative ROI( I have like 10-15 geos, 4 of them are looking promising, but can't scale them. I add more creatives to that 4 promising geos, but its not giving me profits. I thought maybe I've already reached the audiencies at some geos (1 of my dark post reached 3kk people), but the average Frequency is not more than 1.3, so it should not be the issue.
Could you suggest how to scale best?
Thanks!

---

You're still misunderstanding some fundamental things about tracking.

1. Make one conversion pixel at Facebook for an offer. You use the same one for all ads, you don't make different ones for each ad.

2. Place the Facebook pixel at the affiliate network. Don't try to integrate iMobiTrax and Facebook. Use the normal server to server tracking with iMobiTrax (where you pass click IDs) and then treat the Facebook tracking is something completely separate.

How to scale is something you need to test for your specific campaigns since you will have different audiences sizes, ad performances, geos, etc.

Either raise budgets for your ad sets at the start of the day or duplicate them so that you are running multiples of the same ad in different ad sets with their own budgets.

---

That is the problem! I can't find where to make this global pixel for an offer. I can only find how to create a pixel for an Ad. Can you give me a screenshot of where to make this global pixel for an offer?
Thanks

---

Home > Paid Traffic Sources > Facebook & Instagram