[The Grind] Nutra: From Struggling To Prosperity (15)
I'm a Blackhat affiliate new to paid traffic and new to Facebook/Nutra. My past endeavors have included BH Adsense, BH Amazon, acting, modeling, and more shifts as a waiter/bartender/caterer than anyone would ever ever want to do 
.
I decided to enter the Nutra space because I know that the ROI and scalability potential is MASSIVE.
This Follow Along will serve as a place for me to connect with this amazing community. I love how many smart and generous people there are on STM and I would love to become more acquainted with all of you.
The Beginning
I retained FB Queen a couple months ago to teach me the ins and outs of account procurement, tech setup, warming, and all of that nuts and bolts stuff that's needed to make BH facebook sustainable. It was really hard wrapping my head around everything, but after 2 of my 3 calls and many nights working until 5AM, I was finally ready to launch.
My first day was 10/13. I ran a bit of traffic through 4 landers to a non-US Skin offer and saw mixed results. Decently cheap clicks with an ad relevance of 9, great lander CTRs, but no conversions from 12 offer clicks.
The next day I ran more traffic, tested more landers - and saw the same result: great lander CTRs, but still no conversions..
After a few days of this I had 75 offer clicks and no conversions! I switched affiliate networks and tried a new ad creative, and BOOM it started to work. 150% ROI the first day of these changes.
The next few days I made the Noob mistake of raising the adset budget and saw my ROI disappear right away. I realized my mistake, kept tweaking and split testing elements of winning landers, and boy I was really off to the races!
My landers were solidly 4.5+% CTR, offer was converting well, and my step 2 take rate was healthy. 100% ROI on a daily basis, ahhh good ol' Nutra/FB. Then of course, the account got banned. Well no problem, I had a backup that was 2-3 days from being ready.
That account performed similarly, but I made warming/scaling mistakes in my haste to get back on top, and burned the account early. I probably about broke even, but lesson learned.
I hadn't anticipated this, and was really caught with my pants down. I started in on another account and it got insta-banned. Likely due to my US card trying to run on a foreign account. Lesson learned.
I overhauled my account procurement process and got to a place where now I have plenty. I started warming again, building that quality score.
And finally.. salivating at the mouth, I was READY to start running again! Decided to go with TWO accounts this time, with plans to reinvest profits and start in on 2 new accounts every single week, expanding Geos and verticals, dreaming every night of $XX,XXX profit days..
Loaded up those proven landers (11/12/16) and some proven ad creatives into both accounts... and... what?
For some reason I was barely profitable in one account, and had zero conversions in the other. And those very same landers that were performing at 4-6% CTR were now about 3%. And in the days since they've declined to 1.5% - 2%. I had 2 proven ad creatives running on the accounts but still could only get conversions out of Account A. And what was barely profitable or break even has now suffered 3 days in a row of seriously negative ROI.
I have NO idea what could have caused my lander CTR to drop so drastically. They're similar landers to what long running/ high reach ads are using in spy tools - and I checked with my AM who reported no major spike in volume for my Vert/Geo. So it's not like the market is suddenly super saturated and people developed debilitating lander-blindness in a matter of 3 weeks.
But all the same, I started testing variations of winning landers again. New headlines, new images, new domain names/CNAMEs. Still performing at around 2%. Which in Q4, obviously, really doesn't foster good ROI.
In my Aff network I'm noticing a few dropped clicks (they register in my tracker but not in the network). And this week - I've sent 41 clicks according to my tracker, 33 clicks according to my Network's tracker, have only 2 conversions, and according to my AM - 13 partials. That's just a little funky.. Seems like something there is off to me.
One possible answer to my lander CTR dropping so drastically is some sort of back end tech issue. I'm on Rackspace and haven't adjusted any settings there, but who knows..? I'm going to optimize my landers today and see if that helps. I realize now that I should have done this from day 1.
I'm also going to try another offer, on the chance that as well as lander CTR and clicks dropping, the offer could be having some sort of difficulty.. though this seems unlikely - my AM said my offer is still performing well in the network.
So - lots of uncertainty. Lots of stress. My mind is fixed upon this though and I'm going to keep testing, grinding, and trouble shooting where I can. I've had a taste of "It works!!", so like Matuloo says in his awesome "How to keep from giving up" thread, I know it works. I know that I probably got very lucky at first. But I also have faith that if I keep working at this and investing money, it will eventually work out. I want to be able to look back on this day in a year and be proud to have stuck it out - and be in a position where I am indeed running multiple accounts in multiple verts/geos.
Today
-Traffic paused.
-Switch out current offer (A) with new offer (B)
-Optimize landers using iAmAtilla's guide: http://iamattila.com/iamteacher/must...more-money.php
-Restart traffic around midnight so as to give tomorrow a fresh start.
Thanks to everyone reading - I look forward to sharing this journey!
I am having the same issue. I have multiple set ups as well just like yours. However all set ups yield the same result....very low CTR to the offer.
So what i did and you can too is install a recorder on the landing pages. Now I watch on video each clicker's behavior on my landers. Videos tell me people aren't interested in what you have to offer, they're only there because the ad seemed interesting and just want to read the story. For my case I run a celeb angle, so the ad seems interesting, they click, they read the story but as soon as the story starts slowly integrating the offer, they bounce.
Originally Posted by _kelso
Random thoughts...
- did you rip the landers? Any chance you have a thief script embedded?
- not sure on your volumes, have you had enough traffic to ensure your not just see day or week parting fluctuations?
Originally Posted by clickwork7
Thanks for your thoughts clickwork -
Please see the "Next Step" portion of my last post. I and my dev both looked over the landers very carefully - I've also done the ol' refresh trick. While it's possible that there's a very advanced thief script that would be defeated by my dev's and my eyes, knows to thwart multiple refreshes, and was inactive until I started these most recent accounts.. but it seems very unlikely that all of those conditions would hold true. All the same I'm ripping new landers to try.
As for volume, no 300 clicks isn't a whole lot. but being that CTR after testing speed-optimized landers with a new offer was down 3.5-6% from what it once was, I'm confident that it's not just a day fluctuation. As for total volume on this account, there's more than enough to be significant of a startlingly change.
Currently
Working on 2 new landers which I'll be testing with 2 new offers.
Updates?
Originally Posted by davidatias
So I have some very frustrating news, though it will hopefully alleviate everything that's been going on.
While testing each step of my traffic flow, I noticed that the link to my central traffic node in Funnel Flux was often throwing 504 bad gateway errors until I refreshed the page, then it would serve a lander. I assumed this was likely a security feature to obfuscate access to the landers if the user hadn't come from the ad.. but as mentioned in a previous post, I did email them asking them to check.
They got back to me yesterday morning and told me that it's NOT a security feature, but that there's something wrong with my server. They're having their Sys Admin look into my server. This would explain exactly why clicks are registering but CTR dropped so drastically.
I love the features and UX of Funnel Flux, but this is of course, very disconcerting. They handled the installation and optimization on my Vultr server, I haven't touched anything since. I can't imagine they're about to reimburse me for the $$$$ in traffic I've thrown away in the past week - but now I know that this is a part of my traffic flow I can't necessarily count on 100%. Perhaps there's a solution to monitoring my tracking server without having to pay for a 24 hour Sys Admin.
So fingers crossed that they get back to me soon and that this resolves everything and I'm back on my way..
For Now
- Waiting on FF to get back to me.
- Working with Miro @ Banners and Landers, cleaning/optimizing my current landers further, and implementing stealing code of my own. He's very helpful, SMART, and quick to respond. He and his team are expert coders and are familiar with the needs of affiliates - way better than working with a cheapo dev on Upwork. Very very worth the slightly more expensive cost. They're even implementing a custom stealing code solution for me based on Geo redirects
Update
Got the fix from Funnel Flux, turned traffic back on, and lander CTR is back to a healthy 3.5-7%.
But now for some reason out of 23 clicks I have 0 conversions. My hypothesis is that In running traffic with such poor CTR and subsequently such low CVR - I may have trained the pixel to avoid otherwise profitable demographics. I'm not sure if this accounts for 0/23, but it could speak to it somewhat. I started a fresh adset which I believe will start the pixel's optimization from scratch.
To Come
Waiting on Banners and Landers to come through with 4-5 new templates. Once they do, I'll be iterating versions of the landers, let's say a total of 7 versions. I'll then be testing 4 product pairs across 3 networks. So: 7x4x3 = 84 different possible conversion paths. Thankfully FunnelFlux can track conversion paths and has an integrated Bayes calculator.
11/21 Data
Total clicks: 1094
Total Cost: $330
Avg CPC: $0.30 (not bad for q4!)
My offer: $40 Step 1 / $40 Step 2
Offer clicks registered in network: 55
Step 1 conversions: 4
Step 2 opt-ins: 2
CVR: 7.27%
EPC: $4.36
Total spend: $330
Total revenue: $240
ROI: -27%
Adsets
Adset 1 / Angle 1 / Image 1
$110 spent
1 conversion, 1 step two
Revenue: $80
ROI: -27%
Ad1 Relevance 10
Age group 1 ($31) .46/4.5%
Age group 2 ($21) .30/6.8%
Ad2 Relevance 10
Age group 1 ($21) .38/6.5%
Age group 2 ($21) .41/6%
Ad3 Relevance 10
Age group 1 ($7) .46/4% - 2 cvr @ $3.73 each
Age group 2 ($8) .33/6.5%
ROI: 433%
Adset 2 / Angle 1 / Image 2
$220 spent
3 conversions, 1 step two
Revenue: $160
ROI: -27%
Ad1 Relevance 10
Age group 1 ($67) .39/6% - 2 conversions, 1 step two @ $22 each
Age group 2 ($75) .25/9.5%
ROI: -15.5%
Ad2 Relevance 9
Age group 1 ($11) .47/5%
Age group 2 ($15) .29/9%
Ad3 Relevance 9
Age group 1 ($25) .54/4.2% - 1 conversion, no step two's @ $25
Age group 2 ($7) .30/7.8%
ROI: 25%
Landers
Keep in mind - these click stats are higher than reported in my aff network. My hypothesis is this is a clicksteal code issue. Resolved shortly when Banners & Landers gets me my cleaned up landers. To the lucky bastard stealing my clicks, F-you and you're welcome for me previously using an incompetent dev to screen for such things.
*Step two opt-ins are counted as a conversion.
**Conversion rate based on views
Lander K:
Total clicks: 277
CTR: 3.6%
Conversions*: 3
CRV**: 1.08%
Revenue: $120
Bayes chance of being best: 44%
Lander Wee:
Total clicks: 226
CTR: 7.5%
Conversions*: 2
CRV**: 0.88%
Revenue: $80
Bayes chance of being best: 36%
Lander Wee (URL Variation):
Total clicks: 194
CTR: 8.8%
Conversions*: 0
CRV**: 0
Revenue: 0
Bayes chance of being best: 0%
Lander Pl:
Total clicks: 193
CTR: 4.1%
Conversions*: 1
CRV**: 0.52%
Revenue: $40
Bayes chance of being best: 20%
Lander Pl (URL variation):
Total clicks:204
CTR: 3.9%
Conversions*: 0
CRV**: 0
Revenue: 0
Bayes chance of being best: 0%
My Analysis
-The URL variations aren't working so hot, they're getting cut. They worked a month ago, but that was a month ago. I'd like to start split testing variations of the landers that are working, but I think they have clicksteals in them.. so no point yet as I'm getting cleaned up versions probably tomorrow.
-Small pockets of profit in the adsets. Age group 1 did all of the conversions today, but age group 2 has significantly cheaper CPCs. I've seen Age group 2 work in the past and it's only one day's worth of data. Too early to say. It'll be interesting to see how facebook optimizes spend allocation.
-Step 2 take rate sucked balls (50% and no offense to people that like sucking balls). I'm going to attribute the low take rate to not having much data and daily variance. Unlucky, but if my spend had been $1000 I think it would've been higher.
-Conversion rate sucked balls too (7.27%). My pixel isn't optimized and my targeting is still broad (no lookalikes). I'm going to say that an optimized pixel and a lookalike audience will bring this up to the 10-12% range.
Upcoming
-Have a new angle I'll be testing. Waiting on the landing pages to be delivered before I launch it. I've got a good feeling about it.
-Going to keep budget at ~$300/day for now until I can be more confident that these adsets are moving in a good direction. But here's a problem - one is at $100 daily and the other is at $200 daily. If they do well I'll want to increase their budgets, but can't because FB will hit me with more expensive clicks for doing so. Instead I'll have to duplicate and start the new ones with a higher budget. This will be an issue as adset optimization will start all over.. I think it was a mistake to have adsets with budgets lower than $250-$300. But I also didn't want to jump my spend so much as I feared it'd raise a red flag. Anybody have advice on this?
PS: What's the shortcode needed to insert full size images into a thread?
11/22
Cut Adset 2 as it's a different image/same angle but with less CTR.
Adset 1:
Another day of abysmal CVR.
Clicks on offer from adset 1: 40
CVR: 4.4%
Rev: $160
Spent: $210.
ROI: -24%
I'm not sure why CVR is so low. Perhaps users are just sick of this angle, perhaps they click to the product and realize "hey, every skin product seems to have the same offer page, I'm not falling for this", or something along those lines. That means my job as a marketer is to come up with a more compelling ad/angle/lander to force them to bite anyways.
Going to switch out the product for tomorrow, also focus entirely on the best performing lander, as well as a copy variation of it. Budget @ $200.
Also tomorrow, pending receipt of my new angle's lander, I'll be launching a new angle/lander combo that I'm hopeful about.
Viva la stacks!
Did you approach FBQueen about helping you get over these issues? That's where I'd start!
Originally Posted by iAmAttila
11/23
Today I have happier news to report.
The pixel on this account has sufficient data and has started to optimize - furthermore I had my cleaned up landers - furthermore I tested a new URL variation - split an offer between networks - and tried a new offer. The combination of these things yielded positive ROI.
I saw the trend this afternoon and decided to dupe my $200/day adset(1) and set budget to $300 in the duplicate adset(2).
The duplicate adset has much way less CTR and more expensive clicks, and no conversions, but it's only spent about $100, so I'm going to give it a day or two to optimize before killing it and trying another dupe.
Stats (as of 11pm EST)
Spend: $300
Revenue: $478
ROI adset 1: +139%
ROI adset 2: -100%
Offer 1 CVR network 1: 20% (take rate 100%) - Bayes chance of being best: 74%
Offer 1 CVR network 2: 9.5% (take rate 100%) - Bayes chance of being best: 24%
Offer 2 CVR network 2: 7.7% (take rate 100%) - Bayes chance of being best: 5%
So according to Bayes, Offer 2 @ network 2 is getting cut. Not entirely sure one day's worth of data is enough to make this call, but Bayes says so pretty definitively and worst case scenario Offer 1 tanks and I look for alternatives.
--------------
Happy Thanksgiving everyone!
Hey all - if you'd like to continue along, check my new thread here:
https://stmforum.com/forum/showthrea...318#post330318
Home > Facebook & Instagram > Follow-along Campaigns