First campaign (14)
Hey guys, thanks for welcoming me to the STM community.
I launched my first campaign on facebook last night at 12:01am and have a couple questions. I have three campaigns separated my age demos, testing 4 different pictures; so 12 ads total. I'm starting with a daily budget of $10 per ad.
I saw an influx of clicks an hour after launch then it died down for 6 hours, had a quick hour spike, and for the past 8 hours has shown very little activity. Why is that? I was expecting a steady flow of impressions and clicks.
Also, the highest performing CTR have the smallest volume of impressions. How do I make sure that the higher performing CTR get feed more impressions?
My CTR are between .4%-3%. I have 280 unique visits to my lander, with 21 of those going to my offer page, and 2 conversions so far. I've spent $46 so far with sales of $72.
It's amazing that you got two conversions on 36$ offer with just 21 clicks 0_o
From what I've learned, you would need like 3k daily budget for testing offer with so high payout, but it might just be that you hit the goldmine with your 1st attempt. Seems so at least.
Thanks tomsko. How fast would you scale to a 3k type of budget? Does facebook cap you at how much you can spend per day?
Originally Posted by jblevine5
RHS or NF?
CPC or oCPM? What exactly are you bidding? What demographic? Sputtered spend like that suggests your bid is too low, though you got ~280 clicks so it might be the opposite where the ads did well so quickly that they overran the budget a bit.
If you want to scale to $3k a day you should do it slowly, build out more campaigns, increase budgets each night and see how it goes.
FB indeed caps you - you are likely on a sum billing where each say $100 they bill you. This will expand to $250, $500, $750 etc. and acts like a credit cap. I believe the daily limit on the account is actually 2x whatever this sum threshold is.
Zeno, first of all I want to thank you for your contribution to this forum as well as my ability to start in the first place. Your FB Guide is what got me started. I dont have anyone helping me in this particular space so I really depend on people like you. Thank you.
Mobile and Desktop NF.
My first test was using CPC. I bid .50, F50-65. I changed my bidding structure today to oCPM. What is your opinion on the two? It looks like so far that the CPC route got me more clicks and more impressions for my money. For example, on the ads I'm testing right now on oCPM, I am averaging around $17/oCPM. In my first test campaign, I ended up with $17 per 2k impressions on one of my better performing ads. Double the impressions. Are the impressions of different quality/value?
Re: FB cap. Where do I identify my cap?
When do I know if I have found something that is statistically significant? Because I have to start at a slow pace its hard to know how large of a data pool I need to make decisions. For example, my CTR is not congruent with my "Website Clicks" goal. I may have an ad that has half the CTR, but double the website clicks. At this point am I just looking for the ad that has the highest CTR or do you suggest a more holistic approach?
Also, how do you identify which ad generates the traffic that actually completes the sale? Is there a way to track this?
CPC and oCPM are different - the bidding modes result in you reaching slightly different audience segments, the delivery algorithms are different, and depending on the ad, one may do better than the other.
If you are starting a new campaign I'd recommend CPC just so that you pay for usable data (albeit x-xx% bled away by social actions), and this way you won't suffer too much if your ad CTR sucks.
After getting data you can test CPC vs oCPM. No single one is better, they are just different. So, for a specific ad, you can't really guess which one will give you higher volume/profit/ROI/etc. - you just need to test it.
As for the cap, I actually don't know - I use a PMD and it showed the cap in my account listings, not sure if there is anyway specifically to see it in the FB interfaces. The cap gets raised quickly so it's not much of a problem.
As for statistical significance - remember, this applies to comparisons in the data, not the raw numbers. I would look at cost-per-meaningful action and ultimately cost per conversion. If ad A has 5% CTR and you pay $0.50 per website click, and ad B has 3% CTR and you pay $0.37 per website click, chances are ad B is going to generate you higher ROI, though the social actions on A might in the long run benefit it. Ideally look at cost per conversion. Failing that, roll with cost per website click.
For figuring out which ads complete sales, you need to track. Tracking is a cornerstone of IM. Use a tracking system like
Zeno, really appreciate your guidance here. You don't know how meaningful this is for my trajectory in this space. Thanks again.
Re: Cap....there is a section in the ads manager under billing that states the account spending limit. Is this what you are referring to? It looks like I can change it to whatever I want.
As for the tracking, I'm hoping you can clarify a few things. I know I can track the NF ads to the offer page, and I know how many clicks I'm getting to the offer page...but I don't know how to retrieve information of exactly who is making the purchase on the offer page. I dl CPV Lab this past year but haven't put it to use. Would this software be able to tell me which particular ad that I was testing actually resulted in the sale? And if so, how do I do this knowing that the offer page is controlled by the advertiser and not myself?
Can you give me some insights into benchmarking an ad that pays out $36? What is your winning ads CTR, what is the average your paying per website click, out of the clicks how many are going to the offer page, and what is a good conversion% or EPC to shoot for?
Re: Billing. My account states "Unlimited" spend, with a $500 threshold, which according to Facebook, is just when I will be billed not how much I can spend.
Originally Posted by jblevine5
Whatever your limit is, you should focus on getting campaigns going first. There's no point thinking about account limits if you currently have zero traffic.
For tracking, you should use something like one of the trackers I mentioned, so CPV Lab would be fine. You need to configure your tracking system/campaigns to expect extra data in the ad URL, e.g. ...ad=something&gender=F&camp=ABC123 and so on.
This is quite fundamental, you should read Caurmen's tracking guide - http://stmforum.com/forum/showthread...-Read-This-Now
You will pass data to your tracker, it passes data to the affiliate network. When a conversion happens, data gets passed back to your tracker. You can also place FB conversion pixels with your affiliate network that might let FB see conversions as well, depending on what the advertiser supports.
You don't need to control the offer page as the affiliate network has tracking integrated with them.
As for ad benchmarking, I'm really not sure what you're looking for here. If an offer pays out $36, so what? It might convert at 20%, maybe at 0.02% for your traffic/angle/ads etc. There is no ad CTR to aim for or conversion rate, the only goal is for your revenue to be larger than your costs, and you simply can't evaluate this without testing. If you know what the average offer conversion rate is then you can make some predictions, but these will only be ballpark since they are not based on your data.
Great insights on the tracking! Thank you! I made a couple more sales last night. $300 total investment for $174 in revenue, so far.
Re: Ad benchmarking. I was looking for some guidance in how you determine if an ad is successful or when you decide to cut it. I have data from the testing I've been doing all week but its hard to tell when the data is statistically significant when I'm only investing $10-$40 per ad to test.
I was wondering if there was a scenario that you shoot for. For example, if you have a goal of website clicks on a $40 payout, then back into it...you want to average .40 per website click so that all you need to do is convert 1% to break even. Once your ad generates these numbers then you focus on testing the lander for optimization and continue to test and optimize on top of the winning ad.
If I am following you correctly, it sounds like the goal is just to get an ROI on the ad and to continue to improve upon the ROI until the ad stops making you money. Makes all the sense in the world.
Also, I noticed this morning that there were a few negative comments on my ad. Is there a way to block commenting on the ads? Or do I have to just patrol it closely?
With ad commenting, you can manually moderate these but you can also go to your page and add a profanity list that includes the top 1000 words in the English language. Note this isn't appropriate if it's an active page with user comments that you want... but will work fine if you are primarily running ads.
For benchmarking, I see what you're after.
The way I see it, you need an indicative earnings per click to provide guidance. Since you have made sales, you can now calculate an EPC. Say you earn $0.37 on average per click.
Now, if bidding CPC, you can look at how the CTR influences CPC for that specific demographic. From that, you can distill that you ideally need >2% CTR for the CPC to be below about $0.35, i.e. be profitable.
If bidding oCPM this is a bit harder as performance and costs vary day to day, but you can use your intuition here and look at average values over several days.
Now, if 2% CTR on a CPC basis is what you ideally need to be profitable - assuming no change in any other part of the system, then you can start to cut ads that perform well below this. E.g. if ads have <0.5% CTR and over 2,000 impressions, and CPC is above $0.55, then these are unlikely to be profitable in the long run and can be paused.
Be aware that this is not the most statistically sound approach because CTR does not make you money.
You can accelerate your testing, but it won't necessarily make it better, especially when you are optimising landers/sales pages and this raises the earnings per click, which would then weaken your initial CTR/CPC-based rules.
TL;DR: generate some rules once you have some data or insight into average EPC, then pause ads that are weak performers relative to this, but be cognizant of how statistically sound this approach is.
Thank you Zeno.
From our discussion and my initial testing I've come away with more learnings than I have in the last 3 months of just reading up on AM. I can conclude that affiliate marketing is not really a natural talent that someone has. At the beginning it's about patience, focused effort and critical thinking. When you weather the storm of losing half your initial investment, you look for the patterns of positive outcomes and scale on those outcomes. I assume the best AMs out there have created a system of testing, measuring, iterating, and scaling for all of their campaigns.
The key measurements I have found from my ongoing testing is website click cost weighed against EPC. If you can provide a spread there, then you have an ad that can make you money. You are right, CTR doesn't matter if you cannot render a low enough cost per website click compared to the EPC.
I now am further testing one ad that has performed the best. My total investment so far in the test has been $330 with sales of $144.
Thank you again Zeno for the guidance. I will no doubt pay it forward to other new members once I get my bearings. More to come...
It's been less than a week since I started promoting ads, and I just got a message saying my ads account has been disabled and will never be reactivated. Any advice going forward?
Oh dear.
This happens for several reasons:
1) Dancing with the devil and accidentally browsing FB through a VPN/proxy
2) Running in precarious verticals
3) Non-robust tracking setups that leak reviewers to affiliate links that geo-redirect to non-compliant offers.
I'm not sure if you mentioned what type of offer you were actually running?
Home > Facebook & Instagram > Follow-along Campaigns