Facebook Game target with 500 USD starting budget (16)
We are new to FB and this is our follow along.
Our decision came to pick a game offer and started out by collecting intel on demo that backs out for the advertiser. Our AM provided intel and we decided to target that age group.
These are our completed tasks:
- We brainstormed and came up with a plan to target fans of similar games.
- We then scraped uids in these interest groups, cleaned the list of dupes and made a custom audience and generated a lookalike audience from it as well.
- We used social ad ninja to take 5 images that had min running time of 2 weeks and we also made 5 of our own.
- We are going to run 5 ad sets each with 10 ads where the image is different but the text is the same.
- We are using
- We are bidding cpc and mid range of recommended. We are split testing 2 networks to see which 1 performs best.
That's it for now. What are we doing wrong? What should've we done differently?
In principle, nothing wrong.. however your data will tell you.
Post up first sets of ctrs, conv etc and then I'm sure people can help..
Looks good for now - looking forward to seeing your results!
Good plan so far, but one suggestion - split the images into their own ad sets unless you are testing ads consecutively.
If you have an ad set with 10 images in it, and they are all active, and with a budget of $10, your data will be spread too thin and Facebook will skew delivery to certain ads only.
I would opt for 5 campaigns (different age/gender/interests?), 10 ad sets in each, and inside each ad set have 2-3 duplicates of the same advert if running CPC.
Give each ad set a $5-10 budget. Let the duplicte ads run together and kill the ones that perform worse out of the gates.
Run only several ad sets at a time - just pick your favourite images first that you think will do better. E.g. 10 ad sets, run 2 of them to start.
Kill ad sets when the ads overall perform poorly. Rotate in another ad set to test. After several ad sets, if the campaign overall is performing crap, consider flicking the kill switch (be mindful of how lander/offer performance might be holding you back).
When running CPC bidding mode ads on FB it's important to completely segregate different ads, giving them their own budgets, run duplicates and cull ruthlessly.
We've already hit a hurdle: we were promoting a game offer, then FB banned the lander the CPA Networks provide, now we can't get anything approved. I checked and the WOT score went bad.
Do we look for a new game offer, we create an app and redirect, we get a domain and iframe their lander or something else? What is the solution to said problem?
We've also been doing some spying using Social Ad Ninja a lot, and noticed many people promoting this game link direct - they use special landers on their own domains that have a huge 300x300 button which say Play Now on Facebook.
How are they doing that? When asked, our cpa network said we can't get those landers and we cannot make our own that only has facebook on it.
Originally Posted by zeno
--
Another thing: Can someone shed some light on facebook apps, and iframe redirects, when and how to use them?
May be worth using a tool like WhereGoes to figure out which networks the other guys are running that offer on!
However, if the game offer itself is now RED on WOT, you're probably best looking for another offer unless you want to play chicken with the FB banhammer.
Originally Posted by jman86
What was the gaming offer? League of Angels with GTArcade or something? If it was, that game offer is syndicated to a number of other portals, e.g. Kongregate, Game321, R2 Games. If it's another offer, see if this is also the case - might be able to switch up the offer on the backend hence the URL and domain and accompanying MyWoT scores.
Facebook apps, iFrame redirects etc. are mainly when you want to get something non-compliant through or really want the social context that comes with a page ad. I'd recommend ignoring this approach to start with since gaming is generally clean and such measures are unnecessary.
We've redone the campaign using another game the way you've suggested (1 ad set, 3 exact same ads per ad set - 11 ad sets in total - same angle). Had 20 clicks, and 1 conversion.
Our audience is around 240,000, we're targeting people that like MMO strategy games. Based on due dilligence our age groups were 18-24,25-34,and35-45, male, and they liked strategy games and online MMOs.
Now the average CPC bid for our audience is 47 cents to 77 cents. The EPC at network for this offer is around 29 cents. We had 1 conversion and we're at 15 cents EPC atm.
Few questions
1) Despite setting it up with 11 different ads sets, 3 images each - we only got 20 clicks since yesterday 5PM - this morning 8AM. We're bidding 60 cents.
2) The cost of click is sky high, at this price we can't be profitable so how can we increase volume of clicks, and also lower the CPC so its below the EPC once optimized (lets say the click costs maximum 29 cents, if our optimized EPC is 30 cents)
Are you direct linking? What ad types are you using and what CTRs are they getting?
To get click costs down it is absolutely critical that you work on ad CTR.
Also, targeting plays a big role - you may want to try varying up the targeting, e.g. target just strategy games, and just MMOs. These audiences may perform quite differently. Also, try to use precise interests here, not broad categories.
What budget do your ad sets have? It is likely that the bids are too low and that is stifling traffic.
Lastly, don't expect every gaming offer and approach to work. $0.30 EPC is reasonable but you will need strong ads to make that work in the current climate on FB.
Develop a sound testing methodology, get data and get a feel for the ins and outs of the platform but don't be afraid to drop a campaign. In particular, certain age brackets can underperform on offers (e.g. angle/campaign might work on everything but 22-25 age group.
Originally Posted by zeno
adset08 25-35
http://i.imgur.com/10YNHg9.png
adset06
http://i.imgur.com/bDcJMtO.png
adset08 - 18-24
http://i.imgur.com/Vk8aW9i.png
Originally Posted by zeno
Adsets have a $10 daily budget
"Massively multiplayer online role-playing games" is likely a broad key interest group - they used to have # in front of them, or at least those were similar. The demographic will be huge and not very specific. You can identify these by the fact they get a heading in the interests list e.g. Entertainment > Games. Precise interests come under an "additional interests" heading.
I would remove this and target real MMORPG games instead.
Your problem is clearly ad CTR - they are performing very badly. You need to change up the targeting and try different ads. If an ad doesn't stick with at least a ~0.100% CTR it probably won't do to well.
Thanks for all the help so far...
We've ran a few hundred dollars worth of traffic for this one gaming offer we're testing.
We followed zeno's advice to segment our targeting in this format:
age-range-sex-special-interest-group-ad-design
In each ad set we put 3 of the exact same images
So far, we don't really have much statistical data to go by, we have some ad sets that generated 2 leads, but nothing more - because we've created so many ad sets, we've spent a lot but we don't have anything statistically significant.
so far we're about -60% ROI overall
what we're doing
-we are cutting ad sets once they reach offer payout x1 - no lead, or 2x offer payout if the ROI at that point is -30% or more.
what we're wondering
Experts, that test 100s of images to find the best one - do you create 100s of ad sets and in each put the same image 3 times?? It just sounds like SO much work .. or do you put 100s of images into the same ad set that are the same subject -- for example, spaceships
How are you creating ads? Use the power editor, duplicating ads is then easy. If you invest time in getting a spreadsheet template up you can also upload in bulk easily.
I wouldn't cut ad sets on offer payout x1, that's far too early. 2x offer payout and no leads is fine, as is 1x offer payout and no ad clicks. Cut ads if their CTR is pathetically below average.
How many ad sets are you running at the same time? You said you haven't got anything statistically significant, perhaps you are running too many in parallel?
Lastly, the 3x ads duplication is primarily for when CPC bidding. You don't have to do it, but it is wise. As I said, exact duplication is null work with the power editor.
We're running around 300 ad sets. The reason why so many is because of this:
- we looked for 10 similar MMORPG games
- we are running 3 age groups
- we are running different pictures per ad
- we're targeting 1 game per ad set only
The only thing constant is the ad copy.
So we're running 300 ad sets; clearly, our approach isn't sound at this point because we're not getting statistically significant data.
The question we have now is what process flow should we follow:
establish high CTR banner first? (what is a good CTR on facebook?) and then find the best audience?
or
find the best audience that responds the best to our offer and then work on increasing banner ctr?
I'd definitely recommend cutting unacceptably low CTR banners early: you can do that based on a few thousand impressions depending on your target CTR.
If you're running Right-Hand Side ads on FB you can use the "hurdles" I suggest in Part 7 of the Getting Started Guide as good guidelines for ad cutting: http://stmforum.com/forum/showthread...TION-AND-INDEX
I would recommend targeting similar games together to reduce the number of ads and increase the audience per ad. If you're splitting it so much that you get very small audiences (e.g. sub 25k) then the ads will burn out quickly even if successful.
Home > Facebook & Instagram > Follow-along Campaigns