Home > Facebook & Instagram > Follow-along Campaigns

---

Multiple FB Campaigns to (hopefully) find a good one ! (27)

---

Hello guys,

To share or not to share, that's always the question but as the chances to find a killer campaign are close to zero let's share everything and see how it goes.

I'll try to keep things short.

Strategy :

- Multiple tiny campaigns
- Drop them quickly if they suck (e.g. after spending 20€)
- FB Mobile primarily (and transfer to other source once I have a winner)
- White Hat App Install & Games

Campaign 1 (started today morning)
--------------------------------------------

Offer :

Trivago Hotel Comparison

Research / Angles :

Research 1 = common complaints with hotels are : dirty / nasty / rude staff / old-outdate / noisy / bugs / stealing / privacy issues, etc...

Angle 1 : Avoid dirty hotels

Research 2 = target country FR, preferred cities abroad for french peoples = Venice, London, Amsterdam

Angle 2 : "be the best in a subcategory", in my case present the comparator as "Number 1 in Venise|London|Amsterdam..etc"

Creatives :

For angle 1 : picked some of the ugliest hotel rooms ever and added a text "Not precisely as advertised/as in the brochure"



> 3 Ad Sets > Each one with 3x the same image duplicated

For angle 2 : picked a few pictures representing the mentioned cities and added a "Number 1 in ..." text :



> 3 more Ad Sets > Each one with 3x the same image duplicated

Copy :

FB Page Name :

"Clever Traveling" (in FR of course)

Body :

Sick of dirty hotel rooms ? Compare hotels with our FREE APP. (in FR)

Headline :

Dirty hotels ? Avoid the TRAPS with this FREE App!

News Feed Link Description :

Install NOW

Note to self : for mobile the bloody Headline is very very short... my text got cut

Targetting :

- Women / 25-44 y.o.
- Interested in : Air France (airline company)
- Mobile NF
- Android only
- Size : 230 000

Bidding :

- Optimize for clicks > Suggested bid: €0.13 EUR(€0.08–€0.17) > Bid : 0.4€
- 3€ for every Ad Set

Campaign 2 (started this afternoon)
--------------------------------------------


Offer :

ShowRoomPrive

Research / Angles :

Research : ShowRoomPrive is mostly private auction for designer clothes but they have a special on Harry Potter's behind the scene in London

Angle :

Exploit the Harry Potter deal for Harry Potter fans, and of course profit of Equadox awesome copy tip ;-)

Creatives :



> 7 Ad Sets > Each one with 3x the same image duplicated

Copy :

FB Page Name :

"Harry Potter Fans" (in FR of course)

Body :

Don't be a MUGGLE

Take advantage of the exclusive HARRY POTTER London deal from ShowRoomPrivé !

Headline :

DON'T BE A MUGGLE

News Feed Link Description :

Look at Harry Potter's Special on Showroomprivé App

Targetting :

- Women / 18-34 y.o.
- Interested in : bloody Potter
- Mobile NF
- iOS
- Size : 150 000

Bidding :

- Optimize for clicks > Suggested bid: €0.13 EUR(€0.08–€0.17) > Bid : 0.3€
- 3€ for every Ad Set

Various Questions at this point :

- How do you deal with the very small amount of text available for the headline/copy on mobile ? (especially with "non-concise" languages...)
- The DOs and DON'Ts of USING UPPERCASE IN THE AD COPY... what's your take on this ? For me it's more or less yelling so I try not to use it too much but not sure if I have the right balance right now...
- Regarding the Hotel Comparison App : I find it very difficult to find a really "spot on" interest (especially trying to find Pages with affinity > 10 is almost impossible in this GEO)

Cheers,

Ben

---

Interesting angles mate, will subscribe to see your progress.

However the Harry Potter ads are very very bland, they don't tell me anything at all i'm afraid. I think they need to be spiced up

---

Incredible pic you found ! Yeah despite photoshop mine aren't very recognizable / don't pop much... for some reason I thought it would be good to show picture of the actual show but who cares...

---

Update 1 :

- Monitoring profit/loss is a nightmare with FB. No place where everything is consolidated... so I've added the average CPC from FB to CPVLab "CPC" value to have an idea of what is going on...
- I got the "feeling" that the visitors might not "get" the link between my angle... and the final offer page (which is an App Store...) so I'll add a lander in between.

e.g. :

Don't be a Muggle go see Potter in London => Lander say "grab the offer from XYZ available in their App" => App Store

Good / Bad ?

Cheers,

PS : If I do a lander in Muse with a width of 320px, will it look like crap on the bigger smartphone or when displayed in landscape , I might go the purecss route if that's the case...

---

Yes I agree you need a lander because I had no idea what it was, even now I don't know what the offer is

---

Are you using power editor to upload your ads? If not, use it and it, it actually allows you to write a much longer headline.

Also, for your offer #1 you might want to test targeting the "frequent traveler" behavior with no interests.

---

I'm using the Power Editor. The problem is not the length of the headline per se... but the lenght of the headline "optimized for mobile" => if it's too long, it gets cut on 320px width screens and in the preview window... a bit like the Title Meta tag in the serps for the SEO guys...

So with or without the PE, you need to keep your titles short for mobile Ads or it will get cut...

Yeah I might try the "frequent traveler", Zeno mentioned being not a big fan of these broad categories but it might be interesting to test.

I'll post an update later.

Cheers,

---

LMAO... both offers got paused...

I had just finished a polished LP for the Harry Potter stuff... arghhhh

---

Update : so these 2 offers got closed... end results :

ShowRoomPrive iOS App Install with the Harry Potter angle : -82%.

5 conversions for 77 clicks (clicks taken from Aff network... FB gives a lot more clicks)

I'm sure it would have worked better with my new Images & the LP but I didn't have a chance to test it. Not too bad considering the weird angle and the fact it's an iOS App (I've read they dont convert as much as Android)

Ads got a CTR around 1,6% with a max Ad CTR at 2,4%. Quite good considering my poor images imo.

Got some Likes & Comments, Harry Potter is social friendly...

Trivago Android : -92%

2 conversion for 63 clicks (clicks taken from Aff network... FB gives a lot more clicks)

Ads got a CTR around 0,78% with a max Ad CTR at 1,13%. That's a pity as I had crazy images, I guess good images can't do much if your targeting s*cks.

I think this one is a looser due to my very generic audience targeting... using an Airline company don't mean you use hotels...

I started this campaign because I had an angle idea... and got a bit stuck when I had to find the targeting to work with this angle...

Any tips / idea on what should have been done ?

Thanks,

Cheers,

---

Hey Guys,

So I've started a few other campaigns a few days back... be warned the results are absolutely UGLY (well that's a good point for you as I guess there is a ton of room for improvements).

Campaign 3 (started monday morning): HEX

Offer :

HEX Game, Geo : FR

Research / Angles :

It's an Online Trading Card Game / online CCG game which seem to have real "trading" abilities (others don't) as well as characteristics which makes it more MMORPG gamer friendly. It seems like a good competitor to HearthStone, which seems to be the #1 online CCG right now.

So angle 1 is : The Favorite Online CCG for MMORPG Gamers

And targeting is Male 15-24 interest : MMORPG

Creatives :

Well CCG aren't very good for creatives as most in-game screenshots looks like crap, so I've made some based on HEX artworks. I tried to do them quite different.



DAY1 stats :

views : 1,743
clicks :7
spent €1.43
CTR : 0.40%

Hell that's the worst CTR I've ever had on Facebook, I've done better with stock pictures

(Note: in the meantime these Ads are still running and now got 1983 views and 7 clicks... so 0,53% CTR)

So I've quickly modified the images (it was late monday) and added the same images with a red border in 5 new Adsets (I let the first 5 run) and they started showing on DAY2, stats for these since DAY2 :



views :957
clicks: 3
cost: 0,65
CTR: 0,31%

Wooohooo, I actually succeeded in doing even WORSE than the original images...

So it looks like either my images are pure crap, or my interest targeting sucks.

Had an interesting chat with Greg in the meantime which told me that "maybe" this interest has been hammered to death by everyone promoting any king of MMO/MMORPG game on earth.

So yesterday / DAY3 (18 hours ago) I used the images with the red borders (yeah I know I should have taken the others, somehow the stats weren't clear at that time) and created 5 new Adsets targeting "hearthstone" fans, changed the copy along the lines of "HearthStone fans love this Online CCG. What's your take on that ?"

These new sets started on DAY3 yesterday and so far the stats look very bad too :

views : 358
Clicks : 0

Of course that's not enough volume, but it surely doesn't look like a big improvement anyway.

What's your take on this campaign ?

Thanks,

Cheers,


PS : Campaign 4 started yesterday, will post once I have some data to give.

---

OK so Campaign 3 is still running, but of course my CTR is so poor that FB doesn't give me lots of impressions. I've upped a bit my bid (FB recommend 0.13-0.33, I've set it at 0.3). I think with such a CTR it just doesn't make any sense to continue...

=> what do you think ?

Campaign 4 (started yesterday): SMITE

Offer :

SMITE Game, Geo : FR

Research / Angles :

One of the character of the Game is Thor, so I thought about using it and target fans of the Thor movie (audience size: 270K) & The Avengers (audience size 350K) movie using a copy along the lines of "Become Thor".

As my previous gaming campaign has a crappy CTR I decided to bid directly on the upper side of the range to get more data fast. Suggested range : €0.12–€0.27, My bid : 0.3

Images :



Results after 1 day :

983 views Total
8 clicks Total
€2.30 Total Spent
0.81% CTR
€0.29 Per Clic

Again an extremely low CTR.

I don't really understand why this one perform so badly, I real thought it would be a good angle / target.

Any tips ?

I'll change my bid to 0.5 to see of something happens.

Thanks,

Cheers,

---

Hi guys,

Looks like Follow-Along aren't much followed outside the Mobile section

Hopefully someone will get inspired and chime in with some tips (especially : wth is my CTR so low?).

Thanks,

Cheers,

---

I think you really need to put yourself more in the mind of the consumer. People don't just book hotels on a whim like people download games. There are much more costs, logistics, coordination, planning and complications involved. You are using a screwdriver to hammer a nail right now.

---

Thanks but this hotel campaign got paused days ago. I was more interested in tips on the latest 2 campaigns. (Also I had done quite some research as to what people don't like when booking hotels, but anyway this one is paused for good)

Cheers,

---

For these last 2 campaigns, are you doing newsfeed or right sidebar ads?

Here's a tip from me - go on Fiverr and order one of these "hold a sign" gigs (I was using this one ) and got decent results (up to 10-13% CTR in some campaigns desktop/mobile newsfeed).
Split test your 4-5 images,find the winner and then just crop the 'hold a sign pic' and you got yourself a new image that'll probably do better than the last one.

---

Thanks for the tips Arlind

Newsfeed only. I think RHS Ads are almost dead in 2015 (?)

Not sure what you mean with this one : I am already using 4-5 images (all with terrible CTR), do you mean to test the best of mine against the Fiverr one ?

Cheers,

PS : No clue what to ask the girl to put on the sign, "When I play HEX, I'm an Orc, what are you ?". I'll still have to deal with the 20% text rule I guess ? (or is FB more flexible when it's handwritten on a sign ? Might be the case actually...)

---

Find out the winner image of yours and then just include the "hold a sign" pic in that winner image.
As for the text on the sign - just come up with something that'll catch attention a bit.

---

Update :

So the SMITE is performing horrible. I've stopped it.

FB CPC was already high at 35 cents, but when checking within CPVLab to get the real number of website clicks it's a LOT higher, around 64 cents...

Note to the noobs : the FB CPC doesn't mean shit... as they count likes and other interaction stuff as clicks... needless to say that the "clicflow" leading to a like has nothing to do with the clickflow leading to a Game install...

CTR stayed at a miserable 0.5%.

After getting a crappy CTR when targeting fans of the Thor & Avengers movies with the Angle "become Thor", I created Ad-Sets targeting MOBA games and it performed even worth...


HEX

At first I thought this one was almost break-even / slightly in the green, but then I found out the huge discrepancy between the "CPC reported in Facebook" and the "real CPC" that you compute by dividing your FB costs with CPVLab's click count.

FB CPC says 35 cts but actually a "real" (CPVLab) website clicks cost 50 cts !

Despite trying new Ads following Arlind (girl holding a sign) advice my CTR is still abysmal... around 1% for the Ads getting the more volume. Had a chat with Arlind, he said sometimes a country just perform really badly compared to others (e.g. countries doing 10x ROI and for some reason a few countries ROI=0).

I have tested quite a lot of creatives but still my CTR is still bad, I've tested multiple Interests as well... I'm clueless.

Actually the only thing I haven't tested is a "no buttons - no borders - no text - just game creative" image.

I'm pretty sure it could be slightly in the green but my CTR is just way too low to be positive.

Question : it seems like when running duplicates of Ads in an Ad-Set (CPC bidding), Facebook just "give up" on every Ad that don't get any click after around 200 views. Is this a "trustworthy" way to decide whether an Ad is crap or not ? Or how many FB views do you want to validate an Ad is crap ?

And the really bad news : offer payout got a 15% decrease, needless to say that there is no way I'm gonna make this one a winner unless I fix this CTR.

I'm testing oCPM since this early morning, we'll see if it does better than CPC bidding.

Any tip is really welcome ;-)

Cheers,

---

Update :

Yesterday :

- Created a nice and simple LP : background image + fixed foreground image + Button CTA "Download now and play for free", just to see if it does any good.
- Got my best (well I should say least-horrible) performing Ad to test it in UK... unfortunately it has been disapproved as "too much text" (weirdly the same Ads got accepted maybe 10 times in the FR geo...)

Note : again it's been disapproved after being first approved, I guess the review process is really like what Wildfing described : bots review almost immediately > Auto approval > humans review a few hours laters

Today :

- Well looks like trying oCPM with a crappy CTR wasn't a good idea (that's no surprise, just wanted to test it...) as it got me FB clicks at 95 and 48 cents. Note to self : NEVER go oCPM without an acceptable CTR (doing it with a CTR <1% is a no-go)
- This campaign is slowly getting on my nerve, as it's really not in the green at all after modifying the calculated CPC => killed almost everything
- I've just kept the 3 least horrible Ad-Sets and trying to do Day Parting, as it seems like there is clearly a pattern. No doubt holidays will bust the pattern and screw me lol

As the UK Ads weren't approved yesterday I guess I'll submit my second best Ads to see if another GEO gets a different CTR... unfortunately for me the second best is the "race" angle which got disapproved a few times before... or I'll try to submit the disapproved one with a slight change to lower the text %.

Cheers,

---

Haha, love this line:

A good lesson to learn early on!

As for your gaming campaigns, the most glaring issue to me is that you, as cmdeal alluded to, are not getting into the mind of your consumer.

For Campaign #3 you targeted a wide age range - Males 15-24, which captures people in very different stages of their life. Come on, some are still teens in high school, others may be in well-paying jobs and be married with kids. Don't mix teens with older groups where possible. Secondly, you are targeting MMORPGs. I'm not sure what interests you mean specifically, or if it was just the broad interest. This is a terrible idea as your audience will contain anyone and everyone.

This is akin to advertising Nike FreeDiving shoes to anyone on Facebook who likes "footwear".

A better approach would be targeting people who like other TCGs.

For Campaign #4 you did something even worse - targeted an extremely branded term with a large % of the audience likely being poor prospects. How many people do you know who would have went to the cinema to see The Avengers, but who would NOT love to play DotA, LoL, etc? Tween 15 y/o girls would have gone, and they are likely to prefer Candy Crush over Smite. Don't expect movie interests to work out well at all for games unless a) the movie is very niche or game-related, or b) the movie demographic is ideal and the game has very broad appeal.

MOBAs are a much better target for this campaign but then that audience is very large and heterogenous, as well as very strongly advertised to - because every other advertiser at least knows how to Google top MOBAs and can find DotA and LoL, and away they go. Same with World of Warcraft for example. Secondly, because these games are so big (LoL is the most widely played PC game in the world), the audience will consist of a large % of gamers who aren't actually that interested in the game or playing similar games. They like it because they are super casual players.

What I would do there is split-test LoL, DotA, and other MOBAs as completely separate interest groups and break down the ages properly.

If an ad gets no clicks after 200 views in the news feed then chances are FB will reduce ad delivery to it because it's performance is well below the average for that demo. Tough luck, make new ads. If an ad isn't going to get clicks and the offer has a low payout, the economics are against you so you should move on to other ads.

Also, FYI, the ad text rule doesn't effect text that is part of actual photos of real life objects. So you should be fine with the signs without worrying about the rule too much. This is at their discretion though - they may consider it as a clear attempt to find a loophole.

---

Thanks for the tips Zeno, much appreciated !

Actually I didn't post everything as it didn't look like people were actually following.

So a few details :

- For campaign 3: Yes I've tried targeting people into CCG/TCG, as well as HearthStone fans (main competitor of HEX), as well as RPG (not "online") fans, no luck with any of this targeting. There was a good reason to target MMORPG fans : the fact this TCG does have an MMO component to it, I didn't pick just some random interest.

But you're right regarding splitting the age range : young people seem to be "Like" happy but not really "Click and Convert" happy... i.e. the worst audience for CPC bidding...

- For campaign 4 SMITE: I have also targeted precise MOBA games (dota1/2-lol), but again it didn't work either. I included all these interests together, which might not be the way to go... but I can't run 75 Ad-Sets at the same time to split test Images / Interests / Age Ranges in parallel, or should I do it another way ?

Question : Of course I could split everything possible... but if the Ad/Image has a poor CTR is there really a point in splitting ages/interests to the upmost granularity BEFORE having fixed the Ad/Image problem ? Can a poor CTR be only explained by a sub-optimal interest targeting ? e.g. Do you think a good Ad/Image with a CTR of let's say 3% used with e.g. Dota interest would dive to 0.5% used with LoL interest ?

To be honest I was convinced my problem was 90% the Ad/Image, and NOT the interest / age range, that's why instead of splitting interests/ages a lot, I continued testing Ads/Images... no clue if that's the right/wrong way to do.

=> Your remarks make me think my problem probably wasn't bad images, am I right ?

An important point to re-iterate : I was in a "medium" size GEO, so MMORPG/LOL isn't that big of an audience but I recon it probably doesn't contain 100% die hard fans, I see your point here.

SMITE campaign ended up with 9 Ad-Sets and 27 Ads

HEX campaign ended up with 31 Ad-Sets and 85 Ads

Thanks,

Cheers,

---

At the end of the day, on FB there are 3 and only 3 levers you have.

1. Bidding strategy
2. Creative strategy
3. Targeting strategy

Good news is that these are all there is to optimise.

Bad news is that you can't just focus on 1 and ignore the others and thin everything will magically fall in line.

---

Well, I wouldn't consider that a very big connection. MMO is a very broad term. LoL is technically an MMO, so is World of Warcraft, so is Eve Online -- but they are very different games. The MMO term itself has grown less specific over time. So, someone who is interested in MMOs may still be interested in one of half a dozen genres, TCGs being a very small portion of those.

Yep. Younger demos tend to be more trigger happy. Cheaper CPMs/CPCs and lower value to advertisers in most situations. Never equate cheap clicks with low CPAs, sometimes older gamers may cost 3x as much per click but convert 10x as readily.

Yes, absolutely - changing interests can easily cause order of magnitude changes in CTRs.

If you are in a small geo you will need to sacrifice volume and audience size to maintain good targeting practices. It's tempting to try to keep audiences big like you can get in the US with precise interests, but altering targeting to aim for some arbitrary audience size is not logical in the context of ad performance.

The image and ad copy is important, for sure, but the market always comes first. You can create interesting ads that arouse curiosity and draw in users, but you can't create these prospects in the first place. Ask any experienced marketer what they would pick:

a) a market thirsty for a certain type of product
or
b) a great product

The demanding market will take solutions put in front of it, the great product is nothing without a market to consume it.

---

I would split that age range in to at least 3 groups like zeno said. It also seems like you're trying to keep your costs low but you need to overbid what facebook recommends on cpc at the start. If you start out with low bids in a competitive niche you're setting yourself up to get a lower ctr from the start. Facebook will show you to higher value people if you bid higher. I tend to bid double the high end of the recommended bid and use low budgets. Then lower bid, increase budget, lower bid, increase budget until I reach a reasonable price. Your CTRs should be higher this way.

I've never ran gaming on fb so take the rest of this with a grain of salt:

Not sure what copy you've tested but, off the top of my head, I would try something along the lines of "Try the NEW Game that has blizzard shaking in their boots. Play For FREE" and target hearthstone fans.

I would test new images too I think. I can imagine that the average hearthstone fan has seen about a million images like the ones above. Try more wtf style images. Maybe try an extreme close up of a popular hearthstone character's face. Maybe the hearthstonbe character is the one who's afraid, not blizzard.

Spend a couple hours watching some popular hearthstone streamers on twitch.tv and learn some of the lingo. I've found newsfeed ads perform much better when your copy sounds natural and conversational.

---

Thanks guys.

Yeah but what would be a good one ? Maybe "teenagers" (13-18), "students/yound adults" (18-24) and older (25-35 or 25-44) ? Audience Insights don't give any numbers below 18 y.o. but it says
18-24 = 46%
25-34 = 41%
35 + = 13%

No i'm not really trying to keep my cost low, and I'm actually always bidding above the recommended range / at the upper limit, and always got some traffic despite my poor CTR. But yeah maybe at the start I could have bid more, but with my poor CTR I would have ended up paying my max bid no matter what...

Now I tried the same campaign in UK and Jesuuuusss, after 12 hours the recommended CPC range went from around 0.4 to : Suggested bid: €0.90 EUR(€0.49–€1.27) !!!

There is just no way I can make this work at that CPC level, even with the best LP/Offer page conversion ever (for my payout, around 1.7).

Another thing I might not do the right way : The "Fanpage Name" used to post the Dark Posts : do you use a "kinda official" page-name (e.g. "HEX Online") or a more "broad" like "Gamers Reviews"

I used a kinda official and it may smell too much like Ads in the newsfeed...

Thanks,

Cheers,

---

"Now I tried the same campaign in UK and Jesuuuusss, after 12 hours the recommended CPC range went from around 0.4 to : Suggested bid: €0.90 EUR(€0.49–€1.27) !!!

There is just no way I can make this work at that CPC level, even with the best LP/Offer page conversion ever (for my payout, around 1.7)."

I believe that if the CPC shot that high is because your CTR was pretty crappy. With a good CTR, your CPC will drop and you'll end paying much less than the suggested.

---

Yep I think so.

Unfortunately if I can't get a decent CTR targeting TCG fans for a TCG game (which apparently I can't) then it's game over for this offer.

I've tested multiple images and none of them get anything close to a decent CTR. The best one got around 1-1.5% CTR (for FR Geo) but then in UK the same image doesn't perform well at all. I suspect FB finds my bid way to low so it just sends crappy traffic.

I guess I'll stop this campaign (I'll maybe continue to run it day parting in the FR Geo just to see how this work).

The thing is I haven't found a "turnkey plan" explained here to test FB Ads so I'll need to build one myself. (not that I expected to find one but I could have been lucky )

Apparently I actually need a crap ton of budget to test FB.

e.g. let's say my test setup is the following :

- Pick 5 images
- Pick 2 interests
- Pick 3 age ranges

=> That's already 30 Adsets... at 5€ each that's 150€ daily budget, not sure at all that after 1 day I get enough data to make any conclusion.

Also I don't really see how the maths could work :

- From 100 FB "payed clicks" on average 66 make it to the tracker/offer (due to page likes and discrepancy unique clicks / total clicks)
- With a 15% CR this would be 10 conversions

Then, well to be positive with a 1.7€ payout offer that would mean I should get a CPC below 17 cents... doesn't seem very feasible (?)

Am I missing something / wrong maths ? Or is it clearly not feasible to promote offers with such a payout in a GEO like FR/UK ?

Cheers,

---

Home > Facebook & Instagram > Follow-along Campaigns