Home > Adult Traffic (NSFW) > Follow-along Campaigns

---

Life after Amy's 40 day. Adult Dating - First stop: $XX/day (18)

---

Happy to be starting my 2^nd STM follow-along. The first one was working on Amy’s 40 day guide to kick off 2019.

The follow-along proved invaluable, as it helped me really flesh out my thought processes while writing posts.

Coming off of my first follow-along earlier this month, I had my sights dead-set on FB. I’ll spare the details, but I was procrastinating too much while trying to figure out accounts, so ended up settling on Mobile Adult for the time being while I work on warming up FB accounts.

Let’s get straight to it!
---

Overall Strategy

First and foremost, my goal is to hit $XX/day in consistent profits to start. Gotta start somewhere!

Lead Quality

One of my biggest areas of focus for adult dating is to focus heavily on lead quality, which seems to be one of the most important factors in staying on an offer and having some longevity to campaigns.

I’ll do this by using some of the typical tricks, such as using older models, avoiding words like “free”, and filtering out non-target GEOs, language, proxies, etc.

Ideally, I’ll stay in close contact with my AMs and advertisers to get an idea of what’s working, so I don’t have any surprises in terms of getting booted from offers, though I’ll also keep in mind that getting booted can still happen at a moment’s notice.

Testing Methodology

At a high level, testing prioritization will be uncovering the best of... (in this order):

• Offers/GEOs
• Angles
• LPs
• Banners/Images/LP Text

Starting Details

• TS: ExoClick, since I had $200 in unused funds from my last follow-along. I also have an approved account at TrafficFactory.
• Tracker: Binom
• GEOs: T2-T3 GEOs
• Niche: Adult Dating (Mobile)

I'll post daily stats in the next post, to keep things organized.

---

DAY 0

“If I have porn up on my screen when you walk into the office, there’s an 80% chance that it’s because I’m working.” – me to my roommate when starting this

DAY 1

• Pinged 2 of my AMs for top offers in T2-T3 GEOs. Ended up settling on 2 GEOs with about 5 offers to test. In GEO1, I had 2 offers to test, but one disappeared just as I was about to launch, so I just ran one. All offers are <$5.00 payout.
• Started off with 2-3 banners for each GEO, with a combination of ripped banners from target GEOs, and some from the US, translated.
• I’m starting off by direct linking, just to figure out which offers convert and launching ASAP.

Targeting:

I barely have a clue what I’m doing, so I just threw up some targeting on ExoClick to start

ExoClick – All Adult placements; mobile handsets only; target GEO only; target language only; Premium placements

I started out with [300x100 Mobile] placements, simply because they had mobile in the name. I started expanding out to [300x250] to try to capture some more volume/data for optimizations.

Capping is set at 3 impressions / 24 hours. This was set arbitrarily.

Bid:
Min bid was about $0.03 CPM, so I just picked a random bid, $0.051 CPM, to adjust for the premium placements to start.

Stats:


I think I lucked out pretty hard here conversion-wise.

So far, offer 1 is looking pretty good, especially considering I'm direct linking and haven't done much banner/angle testing.

Had a couple of surprise conversions for Offer 2. Although this is obviously an anomaly, I was feeling very good about this offer at this point.


DAY 2



CTR of GEO2 is less than half of GEO1, leading to high CPCs

CPC of GEO2 300x100 is ridiculously high, at least compared to the other benchmarks, due to very low CTRs (.09%) and high eCPM ($0.046) compared to the other GEOs/sizes

Offer 3 has spent 2x payout without a conversion, so I'm cutting this.

Still direct linking to the offer

Large dropoff of volume of Offer 1 (300x100) compared to day 1, which is strange. I'm assuming it's due to increased bids because it's a weekend day.


DAY 3



Offer 1
Offer 1 is running at -37% ROI in aggregate, but given the higher conversion volume and low CPCs, I think I might be able to optimize this to profitability after going through angles, LPs, banners, and images, followed by placement and bid optimization.

Offer 2
Offer 2 has 0 conversions for the second day in a row, after having a very strong start. It’s a higher payout offer, so I wasn’t alarmed after Day 2, but after Day 3, I’m having to look a bit deeper into it.

This running at -63% ROI in aggregate (again this is direct linked with no other optimizations), but looking at the numbers, I'm not sure if it makes sense to continue. CPC of $0.026 and EPC of $0.0096. I'd have to close that gap, and then I'd feel like I'd be looking at breakeven. Given the volumes in this GEO, I don't know if I could find enough conversion data to make the optimizations needed. I might cut this Day 4 if no conversions and revisit this later.

Offer 3
Already cut

Offer 4
Offer 4 is running at -75% ROI to date. I might cut this Day 4 if no conversions. I don’t really see this one recovering much, and I don’t know if I would even have enough conversion data to make sound decisions.


Going into Day 4, I'll make a decision on final cuts, and begin the optimization process of the funnel for Offer 1. At some point, I'll split test this offer from other networks if it seems like it has legs, but for now, I'm sticking to my current network so I don't split up cashflow too much. If cashflow was not an object to consider at all, I'd start split testing across networks at this point.

---

Oki dokie, so let's take a look at what you did here

I know this very well, had to go through this with my parents when I started, many years ago )

Adult dating will not work directlinked, unless you are using offers that have a prelander build in, such as some smartlinks or certain offers that use affiliate style landers. Even then, it's usually better to use a quality LP.

The CPCs you have : 1cent - 5cents are not high at all, for quality clicks I mean. The thing with exo is that they have lot's of crap in the mix and quite heavy source cutting is needed to make it work. I have many campaigns that are profitable even with 15 cents per click or more, but those are well optimized campaigns. On the other hand, the CVR you have now is very low, 1 conversion from 500 bought clicks is very poor. So there is a lot of room to improve here.

Typically, you can reach CVR such as 1:10 (also higher or lower of course) from LP clicks, so you can do a lot here.

The good thing is, you got some conversions, which means at least some of the offers seem to be working.

The final piece of advice would be quite simple : start using landing pages. Once you set some up, let's share the results and we can move from there.

Cheers,
Matej.

---

Thank you Matej!

When I launched this, I was antsy to get something launched ASAP so I just direct linked -- though Offers 2-4 had pre-landers built in, I believe.

In the future, I'll definitely launch with at least some generic landers :-)
--

Won't go too much into Day 4 & 5, since I didn't have LPs up yet!

Also, quick edit to the previous post -- I have capping set at 5/24 hours. It's fairly arbitrary

DAY 4


Aggregate for all days:


Decided to cut Offers 2 and 4 for now and focus on Offer 1, as it seems to have the most promise.


DAY 5


Nothing really to report Day 5 -- took today to prep for angle testing and LPs.

Performance seemed unexpectedly bad, but not too concerned, as it could be a one-off..


DAY 6

Added angles and LPs to the mix on Day 6!

LP-wise, I'm going with a fairly generic qualification and rules style lander. I will plan on playing around with LP styles once I nail down the angle.

Angle-wise, I'm testing 4:

• Angle 1: custom made (LP: Generic and slightly customized to angle)
• Angle 2: custom made (LP: Generic and slightly customized to angle)
• Angle 3: ripped and customized (LP: Generic and slightly customized to angle)
• Angle 4: completely ripped (LP: completely ripped)

Image/model-wise, I'm testing 3 variations (the same 3, across each angle) + the one included for angle 4.

I will say I screwed up a bit with Angle 4. I only have a 300x250 variation, and no 300x100 variation, so volume is limited. In hindsight, I wouldn't have been so lazy and would have created one adapted for 300x100 :-).

BY ANGLES:


Angle 3 is looking the best so far, with Angle 4 inconclusive due to volume.

Cutting angle 2 due to no conversions with 8x payout spend.

BY ANGLE + IMAGE VARIATION:


Angle 3 + Image B definitely stands out here, but too early to say anything conclusive.


DAY 7 (In Progress...)

It's currently part-way through Day 7, so I'll post those stats after the day is completed, but angles 1 and 3 are definitely performing the best today.


PLACEMENTS:

Regarding placements, there are definitely a couple that stand out that I am blacklisting immediately (at least until I find my winning Angle + LP + Offer combo).

• 300x250: txxx.com
• 300x100: m.pornhub.com

There are two placements that stand out in terms of conversions/profitability already with decent volume (about 10% of total volume). Once I narrow down a bit more, I'll probably split those out into whitelist campaigns, increase bids, and customize banners to capitalize.

This data is aggregated across the past 7 days:




I'm not planning on getting too aggressive with placement cutting until get closer to a winning Angle + LP + Offer Combo.

Moving past day 7, looking forward to narrowing down angles, testing lander style variations, and then start testing headlines + banner variations.

Questions:
1) At what point should I check with my AM to see if these leads are backing out or leading to sales for the advertiser? I'm about 80 leads in so far for this offer.
2) Hows all of this looking so far? My intuition tells me that I should be able to get this to about break-even. I'm counting on my lead quality being good to get some payout bumps to profitability. However, if people's typical experiences are to see some green on each angle before even continuing further, that would be good to know.
3) Matej, you mentioned lots of bot traffic on ExoClick. Do you think I should be checking out TrafficFactory traffic immediately, or wait until I narrow down things a bit more angle-LP-offer wise?

---

Hello again

This is very loose, go to 1-2/24.

From some reason, some of the largest sites do not perform well through exo, I'm pausing them quite frequently too. And pornhub has been fucked for a while, not sure if they sell part of the rotation directly for the past few months or what, but the performance there went to shitter.

I wouldn't ask sooner than maybe 200 leads, the lower the payout, the more leads I'd send before asking for a bump. And the key is to be sending a consistent volume, as the leads need time to mature. So it's better to send 30 leads a day for a week compared to sending 200 in a day, then pausing and asking for a bump. It also depends on what the starting payout was. The networks always have some margin they could give you part of, to motivate you to send more, without contacting the advertiser and actually checking for quality.

Bumps will help for sure, but you should get into small profit without it. It's not like you can get a 50% bump just like that. It does happen, but the quality would have to be super good. So far it looks like you're onto something and yes, breakeven should be possible.

TF is quite different from exo, you will most likely need a tweaked funnel there anyways, so in case you are confident enough about the setup and basic optimization, move to TF now.

---

Long post ahead!

I'm covering about a week's worth of updates and wanted to post a lot of detail to both help anyone following along, and to get as much feedback as possible.

Without further ado...

Day 7 (continued)

• Finishing out Day 7, as mentioned, I cut the 2 ZONE IDs related to the placements referenced in the previous post (txxx.com for 300x250, and 300x100 m.pornhub.com)
• Changed capping to 2/24 hours (thanks Matej!)
• Added desktop traffic to the mix @ 300x250 to try to get some more volume to test faster

Day 8-10

During these days, I was trying to get down to a single offer, then narrow down my angles.

Offer:

On Day 4, I made the decision to run “Offer 1” only. However, Offer 1 had 3 different variations to choose from.

Aggregate data for last 10 days:


On Day 10, I made the decision to cut “Variation 1” based on the “CR” field/metric.

I’ve since added a new field, “Offer CR%”, which is the conversion rate once someone LANDS on the actual offer, which I believe would be a more true measure of an offer’s effectiveness.

Looking back, I think I would have chosen to cut both Variations 1 and 3, based on the stats using Offer CR% (97% Confidence that Variation 2 > Variation 1), and it took me until the writing of this post to figure out this realization

Angles:

Aggregate data for last 10 days:



I’m not quite sure I remember why, but I think I pre-maturely cut Angle 4. I’m not sure if I made this decision based on stats (I don’t think I did) or if I was simply impatient to get closer to a single angle. Here I learn I’m not immune from emotional decisions ��. Noted for the future. I should get better at this as I build a process.

Zones:

Aggregate data for last 10 days:



I decided to cut additional zones that were obviously terrible. I’m not sure how I let one of them get to 7.5X payout without a conversion, but moving forward, I’m thinking 3x payout without a conversion as a conservative cutting approach might make sense.



Day 11-12

We go into this day with 2 angles remaining and 1 offer (running 1 variation). My goal here is to cut down to a single angle, and start testing LP type variations.

Angles:

Aggregate data for last 12 days, filtered down to ONLY include the remaining offer:


I wasn’t really sure which metrics to compare here, since I hear so often that Clicks, CTR, etc., are only vanity metrics, and the only thing that matters is the $$.

I was eager to get down to a single angle to start testing out LP variations, so I simply went with the one that has the best ROI here – Angle 3

At this point, I wanted to ask the advertiser which angles are backing out the best for him/her as I think that’s the best determining factor long-term, but apparently, I was passing my angles in the wrong token to the network, so there is no way to get this information.

Banners:

At this point, being 12 days into this experiment, I was a bit afraid of banner blindness setting in (if it hasn’t already).

Aggregate data for last 12 days:


In terms of banners, I was down to 3:

• Angle 3, Image A
• Angle 3, Image B
• Angle 3, Image C

These are all the same banners with different models.

I made the decision to cut images A and C, so I could use Image B as a baseline while I refresh the models within the banners.

I added a 4 more models, and decided to run a very small test on an additional angle. Probably a bad idea to complicate things with an additional angle, but I was really curious .

This leaves us with:

• Angle 3, Image A On hold
• Angle 3, Image B
• Angle 3, Image C On hold
• Angle 3, Image D
• Angle 3, Image E
• Angle 3, Image F
• Angle 3, Image G
• Angle 5, Image E

I am not trying to figure out the best image at this point, but this was more to add some variety to prevent banner blindness.

LP Variations:

Finally down to a point where I can test LP variations.

My control lander that I’ve been using this whole time has been a generic style quiz & rules lander. Nothing fancy here.

I added 3 more variations of lander styles, giving us the following:

• Original LP
• Cartoon-Style LP
• Media-Heavy LP
• Variation of Original LP

Other:

• I decided to cut desktop traffic (300x250) since it was running at -73% ROI to date. I may revisit this once I have a more optimized funnel.

Day 13-14.5 (Today)

GREAT news!

Out of the blue, my AM messaged me and told me that the advertiser likes the lead quality, and raised my payout by 15%. On top of that, the AM was able to get me an extra 5% from the network, for a total of 20%.

She even mentioned that she might be able to get me another bump at the beginning of next week – fingers crossed.

She also told me to scale – too bad I’m not profitable yet

I hadn’t even asked for a quality check yet, so I take this as a good sign.

LP Style:

So back on topic, my overall goal for these days was to decide on a final (for now) LP style.

Aggregate data for the last 1.5 days:


It looks like we have a pretty clear winner here.

I will plan on cutting everything but the original lander, for now.

Angles:

Aggregate data for the last 1.5 days:


Angle 5 looks pretty promising so far – I’ll let it run for a few more days, or maybe make some banner variations for it so I can send some additional traffic.

It’s a custom angle, so I’ll be careful not to make any emotional decisions due to attachment.

---------------------------------------


Now I want to take a step back and look at everything from a campaign level.

By Day:


I think I dipped heavily into the red at one point while testing, but I’m hoping over the next 3 days or so, I can finally make the big cuts that I need to break into break-even ROI.

PLACEMENTS:


Over the next few days, I’ll cut anything with 3x spend w/o a conversion.

DEVICE BRAND:


BROWSER:


----------------

NEXT STEPS:

• Add the pre-maturely cut offer variation back in to test. A 6% offer CR versus a 5% offer CR should net a 20% ROI improvement, so definitely worth testing.
• Cut all LPs except for the original LP
• Start making ZONE cuts over the next 3 days
• Start cutting out worst-performing targetable segments, such as Google brand phones and Opera Mobile browsers (EDIT: Google brand does not show up in the list of targeting options. Might have to get creative with this)
• Start planning out banner variations

QUESTIONS:

• I still feel pretty heavy in the negative ROI territory, but I feel like with these cuts coming up, I should be able to improve ROI significantly and get break-even. Am I being realistic, or am I just being overly optimistic? (we’ll see in a few days ).
• I originally would have almost expected every day of this experiment to have improving ROIs, but it almost seems like a hump / normal distribution (ok ROI -> bad ROI -> ok ROI -> good ROI). Is that common? It makes sense logically to me, as testing things would mean increased money going to unproven variables
• For some reason, I feel like I threw away a ton of testing budget – I have a hunch that I could be making some large cuts much sooner. Does anything stand out?
• Anything missing from immediate next steps? I’m guessing after those next steps I have to work on bids (general and for the larger placements), capping, etc. Then from there, keep testing more LPs, Banners, Offers.
• Regarding banner blindness, I'm guessing that rotating out models is sufficient short term (within a span of days / weeks?), and changing banner STYLES is something that would prevent blindness in the mid-long term (weeks/months). Does that seem about right?
• Anything else? Anything at a fundamental level I should look into more?

EDIT:
How common is it for the advertiser/network to change the offer within the same link? I feel like they may be switching it out on me, but it might just be me being paranoid.

I'm checking the offer link by:

• going direct (leads me to the right offer)
• using a VPN with a desktop browser on a mobile offer (which leads me to the right offer)
• using a random agent spoofer on desktop to spoof a mobile browser with VPN -- leads me to a very random offer
• using a random agent spoofer on desktop to spoof a mobile browser with an in-browser IP spoofer -- leads me to a very random offer
• using a VPN on my mobile phone -- leads me to a very random offer

It feels a bit difficult to try to give the prospect a somewhat consistent experience when the offer or offer lander/variant seems to change every time I visit the offer link.

---

Latest update:
I've been going through and making some significant cuts.

• There were 2 IP ranges with tons of clicks and 0 LP clickthroughs
• Blocked a few browsers
• Blocked some OS
• Blocked placements with 5% LP CTR w/ 1-2x payout in spend, -100% ROI
• Blocked placements with 3.5x payout, -100% ROI

Here's where I think I need quite a bit of help:

While updating costing in Binom, I've basically been going through and straight up updating costs by day, distributed to clicks in my tracker.

Originally, I went to do it by day by ZoneID, but my tracker was throwing errors when trying to update, since some of the ZoneIDs ExoClick (maybe those with clicks) didn't have corresponding clicks to in Binom.

I thought updating costs by day without ZoneID would be sufficient, but now I'm seeing how incredibly a costly that mistake was, since impression to CTR% would make a HUGE impact in CPC in either direction.

Some of my high volume high ROI zones actually have 1.5x the ROI than what is in the tracker (due to costs being over-stated for that ZoneID in the tracker).

On the flip side, some OK ROI zones actually have 0.5x the ROI, making them hugely unprofitable. (with some CPCs being as high as $2.5CPC)

There are also zones that I cut that have 10x less spend than what shows in the tracker.

To reconcile this, I'm thinking of erasing all cost data and updating by Zone by Day -- although I'm not quite sure what to do about zones that don't exist in the tracker. Maybe throwing all of that cost data into some kind of designated dummy zone would be the right move? Not sure yet.

Am I on the right track here, or is there something that just flew way over my head that I'm missing?

Also, I'm sure there's a much more automated way to do this -- are most people using something custom built?

EDIT: I built a report an excel with revenues from the tracker, and costs from the TS. It's definitely giving me a much clearer view. Decisions TBD.

---

Matuloo is the expert in adult - @angle-lytics you're in great hands! I'll let him answer your questions.

Too bad I didn't see this in time - I've posted something like this in the past - although yours may be better:

https://stmforum.com/forum/showthrea...pampamp-easily



Amy

---

Thank you Amy! I actually did find that an hour or two after my last post.

It was really helpful and actually led to me un-cutting almost all of the segments and placements that I had originally cut out.

Im seeing that it's a bit harder to make decisions on cuts such as browser, device, etc, from the tracker, at least with the data that ExoClick provides and on a CPM basis. Seems like Zones are best to focus on at a high level.

This all leads me to believe that the way to do any cuts outside of Zone (such as browser, device, etc) is either whitelisting larger placements and putting the puzzle pieces together, or using some mathematical methods to estimate costs for each click/impression subject to certain constraints provided by exo's segment data.

On Exo, I'm also starting to think that there's a huge advantage to advertisers who can figure out automation and proper/granular cost attribution.

I actually ended spending about 3-4 hours digging into how to semi-automate updating of costs and merging of tracker data and made some progress as well.

However, after probably about 7 total hours digging into and trying to solve this, I realized that this is probably not the best use of my time at this point hah. I can get to a point that's "good enough" by using that Excel sheet.

I need to keep reminding myself of this throughout my whole journey, so I don't get stuck trying to solve problems that shouldn't be my #1 priority at the time.

It does help to get my thoughts down on paper -- enough rambling. Thank you! :-)


EDIT: to anyone following - I think I should have a big update in about 24 hours or so!

Sent from my Pixel 2 XL using Tapatalk

---

Oh wow, couple large replies that I missed during the weekend

You're doing well man, it's great that you are really analyzing your steps in such depth!

So basically, there are 2 possible problems with your campaign, maybe 3

1. the funnels are not strong enough and you simply have to experiment and come up with better ones. Hard to say whether it's the banners or the LPs and only testing more will help here.
2. the correct cost tracking is quite a pain with exo for sure, sometimes I'm also using a custom xls sheet that my coder made for me. I export data from both the tracker and exo and compare them in the sheet. You can also send a postback to exo directly to check the stats there, I'm not doing that since I like to keep my conversion data private, but it can help with optimization if you wanna take that route.
3. the offer you work with can be poor, or not the right one for this source, it's not like all the offers can work with all traffic ... so even if the AM keeps on telling you some offers are great, they might not work well for you. It's always a good idea to test more offers.

Generally speaking, I think you should be able to improve the funnel by using better creatives. You can obviously start cutting everything that's not performing well for you, but that will also limit your volume a lot.

Looking forward to it

Some more answers for you :

It's probably just the problem with the VPNs and emulators you are using, in case the offer detects it as the targeting they are after, you see the real offer, in case it believes it's not a match, you will get redirected to some fallback offer.

Aggressive cutting can definitely take you to profit, the question is, as mentioned above, how much volume you will be left with. So I'm gonna repeat myself, work on the funnel too, not just cutting sources/zones.

Yup, even a well optimized campaign shows differences in performance from one day to the next, the fluctuations are normal. And the fact that you are testing new stuff just makes this more normal.

I think with the more accurate cost targeting, you will be able to make much more meaningful cuts from now on. I guess you are on it already, so won't write anything else here until you post again.

Gonna repeat myself again a bit, but your focus should be more on the creative side. You need a solid funnel first, then you can focus on the other aspects such as bidding and capping.

Rotating models will help for sure, but I would advice you to work with multiple layouts too ... it doesn't have to be anything fancy or professional, just aim for some variety. I still see years-old banners re-appear all the time, I'm also recycling my old banners often, so yes, they can last a while just make sure you're switching them when the performance starts to drop.

Ok, enough for today, looking forward to your next post

---

Ok – finally got everything together! I have a template now that makes it easy to throw data into to analyze placements.

I have a route to increased automation, but after fiddling for way too long, I think this current method is “good enough” until I get to the point where it's clear that automating will directly lead to an increase of $XXX/day.

So since my last post, I’ve been trying to figure out the best way to analyze placements for Exo data.

The link that Amy posted was a huge help, and I’ll definitely incorporate elements of that once my campaign is more stable.

I finally made some good progress on this to share, and I also have results from a fun weekend experiment to share as well.

BTW – thank you for your responses Matej. INCREDIBLY helpful! I definitely owe you a beer…or 10…or 100 someday ��.


Placements!

Here’s all of my placement data through Friday. I have 2 additional full days of data to share as well, but will do so later in the post.

This paints a VERY different picture of many of the placements that I had analyzed from tracker data in previous posts (at least when looking at specific placements).

There was a placement that I blacklisted previously that I had marked as $7 spend. Turns out, it was only @ $0.70 spend. Looking at this data outside of my tracker was the best thing I’ve ever done!




Cutting While Conserving Volume

So, most of our highest volume placements are deeply in the red. Shit.

I can cut a ton of these, but I’d also lose a lot of volume.

So the things I’m noticing here are:

1. I have a few placements with decent volume that seem to be doing well (Placements 3, 4, and 7).
2. Placement 1 has more then 2x the amount of spend as the next closest, and seems to have at least some traffic that converts.

Observation #1 seems great – this gives me at least a little bit of hope. If I cut everything and just whitelisted these, I could maybe reach the elusive $X/day ��

On Observation #2 , what I would typically do is either:

1. Go into the tracker and see if there are any segments within this placement that are just straight up wasting money, or
2. Call this placement hopeless and blacklist it

I now know that Option A isn’t possible (at least not yet) due to the way costs are attributed.

I think earlier, I was tempted to go with Option B, as it seemed like the easiest path forward.

However, thinking about it some more, I realized that I could literally cut out 75% of the traffic from Placement 1, and it would easily still be in the top 10 in terms of volume, maybe even top 5 after cutting out unprofitable placements. This made it worth it to me to try to salvage Placement 1.

Whitelisting & Digging into an Unprofitable Placement

Since I couldn’t solve this by digging into the tracker, I decided the only way to get any kind of granular segment performance data for Placement 1 was to place it into its own whitelist campaign.

After running this for 2 days, I finally had a little bit of data to work with.

I did a little bit of digging, and bingo! There’s a single device (Mobile – LG) that’s had only 3 clicks, but accounted for 60% of the spend. CPC for this segment is an astonishing $0.95. With EPCs of $0.01-0.05, there’s not a chance in the world this could ever work out. I also verified that this device never converted for this placement.



So starting today, I’ll be blacklisting this device within this whitelisted campaign.

Moving past Placement #1 , I’m tempted to cut Placement #2 altogether, since it clearly wouldn’t be a large player judging by the conversion volume, but I’ll take it as a learning exercise and whitelist it for a day to see what I can find.

This is something I wouldn’t want to do for lower volume placements, since that would create an organizational and operational nightmare, especially when managing additional creatives.

Additional Blacklisting

I cut about 15-20 placements in total – added some to my perma-blacklist ($5 spend with 2 clicks??).

Feels like a high number, but seems reasonable considering I’ve essentially been running traffic without any placement cuts for multiple weeks now. Will probably cut a good amount more in the coming days.

This comes out to cutting about ~12-15% of total volume for now.


My Weekend Experiment

I’m still trying to determine two things:

• The best offer variation (between two remaining variations)
• The best angle (between two remaining angles)

I felt like I was twiddling my thumbs a lot, waiting for conversion data to slowly trickle in, so I decided to run a test.

Partway through Saturday, I decided to whitelist my small handful of placements that were performing well.

For this, I doubled the bids, and increased the 24 hour capping from 2 to 5. I probably should have stuck to the bid increase only, especially so I could isolate factors that lead to changes in performance, but…I don’t have a good reason except I wanted more data, faster.

A few hours into it, it felt like conversions were increasing without too much drop in ROI, so I decided to double down and try it again, going to 4x my original bid (still with a cap of 5).

The difference in volume that I saw was incredible – though the fact that it was Sunday probably played a lot into this as well.

So throughout this campaign I’ve averaged probably about 11 conversions per day or so. Sunday, I hit close to 70 in one day!

Previous baseline for my best placement -- I was averaging about 32% of clicks going from lander to offer, and about 7% of people converting from there

However on Sunday, I had 43% of clicks going from lander to offer, and 16% of people converting from there. I'm assuming that this is due to a better quality of inventory.

Here are the stats for Sunday:

• Cost: $84.92
• Revenue: $81.17
• Profit: -$3.75
• ROI: -4%

+27% ROI and +$18 profit on those placements that I bumped up for the experiment.

I do feel like I kind of lucked into this, but I'll take it :-)

I’m going to keep this going for 1 more day – I fully expect my ROI to tank on Monday/tomorrow, but there’s only one way to find out. I have a feeling it might shoot back down to the -40% range, with high spend.

It seems like every time I start to feel confident, the next day shits on me and reminds me how much I have to learn!

On Monday evening/tomorrow, I’ll change the capping on these placements back to 1 or 2, while keeping the bid the same and see how that impacts volume and ROI.

THE BAD NEWS:

I mainly ran this experiment to figure out the best angle, and the best offer variation.

Here are the stats from this weekend…

Angle:


Offer Variation:


Great…haha.

One thing I am excited about is that Angle 5 is my 100% custom angle, and Angle 3 is one that I ripped that’s being used pretty widely worldwide. I’m pretty stoked that my custom angle is able to keep up!

I’ve been passing the angle as a token to my affiliate network, so it may come down to the advertiser telling me which angle is backing out best.

Plan for this week:

• Follow-up with my AM on the additional payout bump that she said she could likely get me this week
• Continue to optimize placements throughout the week
  
  • whitelisting anything that I want more granular data on, or anything that I need more precise targeting on
  • blacklisting obvious targets
  
  
• Continue to test different levers to improve funnel (this is where I think most of my time will be spent)
  
  • As Matej suggested, I will continue to test more creatives and will probably test a couple more offers
  • If I can maintain volume, I should finally be able to test FASTER, which is critical
  • I haven’t even started playing with banner styles and headline variations yet, so these could be HUGE leverage points to get to profitability. I’m still using static banners, since I wanted to keep in as many zones as possible during the initial data collection, so moving to GIFs alone should help a lot.
  • One concern I have is that to access more adult dating offers for my GEO, I may have to open up more accounts with affiliate networks, which is not ideal for cashflow. I’m still working to hit the payout limits for the ones that I’m already a part of. I think I have about $1K owed to me that I haven’t been able to hit thresholds for. (EDIT: just applied for 4 more networks to get more access to dating offers for this geo -- offers are the most critical part of the funnel, so I shouldn't skimp here.)
  
  
• Start playing with bids for any other placements that are profitable, or borderline profitable. I actually want to do this to prioritize getting more volume over more profits. More volume = faster testing

Questions:

• I’m thinking of sticking with ExoClick for right now and this single GEO to understand the nuances of each for the next couple of weeks. I feel like there’s so much to explore, such as all of the additional banner sizes outside of 300x100 and 300x250, Members and RON inventory, etc. Any thoughts on this? Also on that note, do the other banner sizes have volume? If not, is competition low enough to warrant going after? I’m guessing it depends on GEO.
• ExoClick reached out to me and asked if I’d be interested in a traffic share deal for any placements. To quote: “This would fix the percentage of traffic and price for a 30 day period.”
  
  • What do I need to know about this as a newbie? I have a feeling this would be in the multiple thousands
  • I’m guessing I’d want to have a very stable funnel set up
  • I would probably need to have a VERY solid offer with tested and available backups
  
  
• I’m considering doubling my bid for my general campaign – ah screw it – I should just do it as a test to see the results, right?
• How long does it normally take to build a profitable campaign? I'm guessing it depends a lot on what proven banners/landers you have, and your sources for offers, but I'm on Day 17 or 18 I think and it feels like it's been a while!

----

Phew!

Each one of these takes a couple hours to write, but they’re always so helpful for me. I think this one took over 3.5 hours, since I was combing through data while writing it.

Hopefully someone else is finding this useful as well!

---

Just wanted to update this in case anyone was following this at all.

I eventually did manage to get to $0-$30 consistent profit daily with 0%-100% ROI for a span of about 2-3 weeks, before killing the campaigns to put my focus elsewhere.

After testing 20+ offers, I eventually landed back at my original offer as the most profitable, and got to landers with ~45-60% CTR compared to what I was originally running.

The weird thing is, that all of my profits came from a single placement -- don't know how typical that is, but it outperformed all others by far!
--

Adult was a great experience and I definitely picked up a lot of skills along the way. Ultimately, I think I might step away from adult for a bit -- I think looking at porn all day indirectly impacts other aspects of my life in unexpected ways, so I'll be taking a break.

I may come back to it again eventually, but for now, I'll be focusing on some other stuff! I know jumping around is the worst thing for a newbie, but I'm starting to get a better sense of what I like and don't like, while picking up a ton of skills / context along the way.

Thank you for the help, Matej

---

@angle-lytics Adult has gotten pretty competitive! Congrats on the consistent green! It IS valuable experience!

What are you going to conquer next? Push and native are both pretty close to pop / adult. Push is especially popular these days. Native is harder to get going and will cost quite some money to test stuff and cut widgets, but can be very lucrative once you've spent enough money to cut non-converting widgets (not unlike how you cut placements in pop).

Thanks so much for the update! Been wondering where you had gone.



Amy

---

Hey Amy!

Sure thing -- right after adult, I had an idea for a very specific POD e-comm product that didn't quite exist yet in the pet niche, but after setting up an MVP store and running some FB traffic, it didn't convert nearly as well as I had hoped (pre-optimization), so I scrapped that idea. Fortunately from AM experiences, I was able to tell myself that the 'offer' wasn't worth optimizing due to very low return on ad spend on initial tests.

I have a couple of options that I'm considering now:

• Native - I've been reading up on Native the past couple of weeks and I'm at the point where I think I can start running traffic next week. The more I learn about it, the more drawn to it I am. There is the concern about budget however, which I think I can address one of two ways:
  
  • Start slow and work on generating profits/optimizing on a small subset of whitelist widgets, reinvesting everything back into building up a blacklist as I continue forward.
  • Start off with Native, and if I see promise, get a full-time job or alternative forms of funding to invest into this if I start seeing potential
  
  
• Learn by working for a superaffiliate for free/commission only (if this is something that people are open to). One thing that's painfully obvious, is that learning is hard and capital intensive in this industry. This would cut years off of learning time. Fortunately I'm in a position where I can work for someone for free for at least 3-6 months, and I think I have some valuable operations experience that a lot of people are looking for. I keep hearing that people's biggest hiring headaches is finding their 'right hand man'/project manager/operations guru. I've built my non-affiliate career on being that guy in a few different industries to some incredible people, so I might have a lot to offer here. This would have to be with the right person though. :-)

Thanks for everything as always, Amy! Hopefully I have some better things to report in a few weeks...maybe I'll start another follow-along!

---

Adult isn't for everyone indeed. You're not the first person who finds it too distracting, I was in that phase too but eventually it faded away.

One way or another, it's great that you managed to reach profits and keep them for a while. That was a very valuable lesson, maybe more than you imagine.

Looking forward to your next follow along, whatever vertical/traffic type you choose

Cheers,
Matej.

---

Wow, what an awesome thread. Thanks for sharing. I read it all in one sitting. There was definitely a lot to take in. Is that a thing where you work for a super affiliate for free in return for mentorship or guidance or whatever? If so, where do I sign up lol? I definitely have the time and the work ethic for it and I think it'd be good leverage for both parties.

---

It would be a long shot - but worth a shot!

I don't think there's a central place you could sign up to. You'd probably need to contact each person individually. Another option would be to make a thread here on the forum to see if you get any offers.

If you want to go that route, I would STRONGLY recommend that you somehow prove yourself first, before even contacting anyone.

This means for example:

-Spending a reasonable amount of time and money on running campaigns yourself first to get some experience, and create good track-record (tracker stats, affiliate network stats, AMs that can vouch for you).

-Making some posts here to show what you're learning from your experience. Starting a follow-along for example would be a great start!

Super-affiliates get hit up by a TON of people asking for tips and tricks and all kinds of favors. You need to prove you'll be worth their time. Otherwise, chances are you'll be hearing a lot of "no"s.

Another thing that will help: Go to meetups and/or conferences to meet people face-to-face. It's always more convincing when the person can hear your tone of voice and see your facial expressions - it would give them a way better idea on how eager and determined you are to succeed.

Best of luck!



Amy

---

Thanks for the input. I was just joking a little. I think that being able to turn a campaign green or a few by yourself is such an important step and can only be done by yourself. I also strongly believe that before your able to work in a team you should be able to successfully work independently. Unless you have a mentorship thing going on, that's a whole nother story.

---

Home > Adult Traffic (NSFW) > Follow-along Campaigns