Scaling Facebook Campaigns - tips/best practices? (22)
I have 2 FB campaigns on fire now and I'm thinking about how to scale them (my first time doing this).
Was spending $100/day on one and bumped it to $1K/day today... interesting results, yesterday was doing close to 300% ROI and ended up slightly above break-even today and was getting burned with my CPC shooting up ($0.12 yesterday and seeing around $0.47 today). I could've handled this a lot better, but my first go at it so I'm still learning. And the campaign still has gas left in it, so I'm not worried either. I had it on oCPM and just switched it to CPC and set a much lower max bid to keep it in check, but it seems like ad spend hasn't been going up.
Very true I think what Zeno says here about FB throwing the ads up on ROI killing spots if you scale abruptly maybe: http://stmforum.com/forum/showthread...l=1#post147845
So my big question to you guys is, once you've spotted a winning campaign...
How do you go from spending $100/day to spending $5K or $10K/day on it on FB? What's the best way to do it while maintaining your ROI and low CPC?
Any specific FB metrics you keep your eyes on?
I've been hawking my Reach and Frequency on my ad as I've increased to $1K/day, to make sure the spend isn't being wasted. Target audience is 220K and my campaign's current reach is 32K, 1.35 frequency, 4.871% click-through rate. I'm expanding outside of FB too to scale this offer, bought banner ads on a niche site yesterday and they're totally crushing it so far.
I'd like to see how far I can take this though, how can I spend $5-10K on this on FB without CPC shooting up? Sorry if this is a n00b question, but would really help me out.
Very eager to read these responses!
I'm not anywhere close to knowing really how Facebook works myself since I'm a PPV kinda guy. But what if you were to simply copy the winning campaign of the same exact thing 10 times? And keep each one of them to 100$ budget per day?
Wouldn't that fix the problem of ramping up only just 1 campaign up to 1k, and killing yourself with the higher CPC's?
While there will surely be answers to come that deal with the problem in general, my observation is that your demo size is far too small.
You want to spend $1,000. FB will let you.
You want to spend $1,000 at 12c CPC, with such a small demo? I got some bad news for you bro...
$1,000 / 12c clicks = 8333 clicks.
At 5% Unique CTR, that is 166,666 people that need to see the ad.
220k demo size, maybe 100k log into FB every day. Not enough people. Add to this the fact that FB will serve ads in an uneven manner. If you want absolutely the top ad slot every time for each person in your demo for the day, you will be OUTPAYING every advertiser who is also targeting them with a News Feed ad (easy to get 5% on those). Money wins. The FB ad auction is described in the API, you can't simply expect cheap clicks if you desperately need the ad to be seen by every person (which you do).
When setting to CPC to cap the bids, you will not be able to spend your budget. You need to find the balance.
I know with my niches and ads, that for every 100,000 people, I will only budget $[amount] per day. Any more than this and I will see bids skyrocket like you have.
While i agree to above post. But fear not. Try this approach:
Campaign mode:
OPCM. Set budget $50 on your ad.
Check, you stats when you are getting max clicks.
On 2nd day exactly 1 hour before you Peek CTR time, clone ad 5x times.
Now, just wait and let the stats settle. ( You must use OPCM. give altleast 5 hours to per campaign ).
Keep repeating it and keep increasing clone X factor.
I milk out profit till campaign works. Then move on.
Cloning campaign at peek time, after 5 hours you will get a report of CTR. You still can't trust these CTR/CPC values. But i delete campaign which are giving me high CPC values. ( above average )
Now, its a very simple approach. But CPC will further slim out after 2-3 days. So, campaign will profit more if you run campaign for long stretches. My audience is usually broad, never faced any problem with CTR dying before 20-30 days.
Originally Posted by redrummr
Good to know, man. This has been a great learning lesson for me so far. Reminds me of this post by Smaxor before on targeting big demos (22M females): http://stmforum.com/forum/showthread...et-on-Facebook
@davidkingstonn @richierich Thanks, I was thinking of doing that too. Will do next time... like the idea of cloning at the peak CTR time too.
This campaign's going to run for a while, but yeah, I can't have wheels go off the cart so to speak by scaling so much with my small demo size. I'm toning it down for this offer on FB and instead focusing on banner ads on this other website, keep a healthy ROI and keep it going and explore other opportunities.
Richierich's suggestion is good, but oCPM does not take into account 'starting CTR' or anything like that. It's only an issue with CPC bidding, and even then it doesn't matter if you have 1 ad per campaign.
Having multiple ads can be good, up to a certain point. Such a small demo size does not need multiple ads. In fact it can increase CPC, as the ads will enter the auction together if they all target the same demo. When using multiple ads, or multiple accounts to serve the same demo (I do both), you should only do it in extreme cases. 220k does not warrant it, you will only read too much into things.
Keep a good thing going and expand to other offers/demos.
Redrummr, do you change oCpm default values or you just tweak budgets/day to get a good cpc for small demo sides?
oCPM bid values should remain high.
Do you mean all them: clicks, reach, social and actions?
Or just high values for clicks and $0 for the others?
Leave them as they were when you started.
Depending on what you are optimising for you will have $0 for everything but that selected action (E.g. 0/0/0/30 or 5/0/0/0)
So if I'm looking for clicks, should I leave the defaults, or should I set manual bids for clicks (say 5 dollars) and let the remaining values at 0?
In your experience, does it help to set bids manually or should you always stick to ocpm defaults?
I stick to oCPM defaults and these change based on the campaign-level optimisation goal.
You can change these manually if you want i.e. you have a campaign for clicks to website but want to weight it more toward engagement/actions.
Just make a campaign with whatever objective, make an oCPM ad with default bids, then note down how these are weighted for that specific objective (e.g. for clicks it's $5 to clicks and $0 to everything else). Then you can just set these manually in the future if need be.
I'm still trying to figure out how to scale up a budget in a very large demo. I had a nice ad running with a demo of about 6 million and quadrupled the budget overnight after 12 from 500 to 2k and everything went to hell. I'm not sure if that was just bad luck/variance or if you really have to nudge these things up gradually...even in huge demos.
Scaling campaigns on Facebook is art not science. So, good luck with your creativity.
Does anybody know where to find information about what hours of the day are the peak times for each campaign? I know I can drill down by day, but I haven't figured out how to see the stats in a 24 hr time period?
Originally Posted by escondido
Originally Posted by escondido
You can get that information if you have access to their API.
We record the data every hour from each update and use that to build out hourly reports for ourselves.
Reading the way they write about hourly stats in their developer section, seems like they have no interest in building it into their reports.
Facebook is a bit different now vs when i was taming the beast.. but back when i ruled the kingdom it was all about CTR. My main trick was to hire photoshoppers +do it myself, and find the weirdest craziest images. I was ALWAYS working trying to increase my CTR to get the most traffic.
Facebook doesn't provide hourly data, you need to get to this via a PMD (i.e. their API) or by your own tracking - though the later can't record frontend performance.
Zeno's Guide is awesome.
Find the SWEET SPOT i would say!
I had similar situation for my campaigns. My campaign was set to spend $200 everyday ..i was paying 20c CPC..adcopies were receiving 2% CTR for a 2 million targeted audience and the overall ROI was 150%. I was all happy and wanted to scale it BIG so increased daily spend from $200 to $1000.
Now..FB started charging me 45c CPC..obviously ROI fell on face so quickly. Since I had a targeted 2 million audience (all women), I thought CPC will stay normal.
******************************************
If you are running FB newsfeed ads...
Good converting offer --> perfect interest/targets (1 million is awesome) --> Good CTR adcopies --> And that's pretty much it!
When the campaign is going well, find the daily budget "sweet spot" and keep it going.
When CTR drops..find new adcopies or add few more relavant interests to the interest list and that should do.
******************************************
....so for this campaign, I started dropping daily budget each day and when it reached $500/day I started getting 25-30c CPC. Campaign was profitable and it continued to be profitable for around more than 2 months.
We ended up making around 55K USD in revenues from one single offer (all from FB NewsFeed traffic yes)!
The best part was..LP had 85% CTR after 150,000 clicks.
PROOF:
Home > Facebook & Instagram >