[TheFJK] After Dark Dating Follow Along - Period (56)
So it’s time to see things I never wanted to see. Shit, already happened while spying! Here’s my follow along for the Adult dating contest.
Some general infos, I’m running gambling / make money stuff already on those sources, but personally have never ran adult, so consider me being pretty fresh to adult dating. Also I’m not handling the setup on sources for gambling but the LPs / offer optimization, so this is some new for me too. We’ll see how/if at all I’ll be able to adapt to the world of ex-girlfriends looking for a bang and all those super horny, ready to hook up amateur models just waiting for the opportunity… shitload of traffic is there for sure, spying is easy, what more to ask for?
*** BASIC INFO ***
I’ll be seperating this follow along in 4 start posts. The first gives an overview of the general setup and the introduction you're reading right now, the second will show offer information along with screenshots of those offers while the 3rd gives you some samples of landing pages / creative I’m using. 4th post is reserved for traffic sources / sources stats etc. and the 5th is for stats getting updated etc. So a perfect setup to rip with just a couple of click if it proves profitable!
Traffic Source(s): Starting with Exo / TJ
Country: Germany.
Budget: Flexible here, looking to start with 50-100 per source per day for some quick data and depending on success scaling etc.
Offers: Probably a shitload of offers to test. Big thanks to Alex from Adsimilis who provided me a list of strong converting offers, I’ll be split testing those offers and see what offers perform best, show promising results or just tank.
Networks: Splitting Adsimilis & F5
Angles / Ideas: Well what to say here. Bang whatever you want. Going to setup a bunch of creative with different / similar angles and see which of them performs or not. I will not focus purely on bang a milf / bang your neighbor / bang whatever shit.
LPs: Going to start with 6 LPs following a ‘general’ angle on adult. Obviously a lot of rules stuff is including, there’s a reason for those landers running everywhere. In case I ripped any lander of somebody in here: sorry about that, but if it wasn’t me somebody else would have hit it within 24h for sure. I will be showing off 3 landers which are extremely generic / standard and not disclose 3 landers a buddy of mine who’s running adult on other markets is also running. Personally I would post but I don’t feel comfortable to post shit he gave me and he is running.
Goal: Well what to say here, the usual stuff, experience, get profitable, bank hard and never need to work again. Just kidding, like I said, I’m running on those sources with a partner of mine who’s handling the setup while I focus on the LPs / offers for gambling, so getting more familiar with the world of adult dating pretty much described my goal. There’s obviously huge volume on those sources, so getting familiar with adult dating on web first, then taking it further to mobile is something that sounds interesting to me.
*** OFFERS ***
I got a shitload of offers for adult which are supposed to be doing good, here's a brief overview of the offers. In case you wonder, with Adsimilis/F5 you should be able to get all of them:
General/Castual that's supposed to work on adult too:
C-Date
eDates
Casual Club
Secret
Adult Dating:
Erovie
Kostenfrei-Flirten
Seitensprung Area / Fickdates
SexPartnerClub Rotation
Treffpunkt18
Flirt Cafe
Amateur Community
For the start, I will be split testing C-Date with Seitensprung Area / Fickdates. C-Date is extremely popular in Germany due to advertising in print media and television while for Seitensprung Area I definitely like the lander and 'tone' of the website. I know that means shit, but given I like their design and it's recommend as a good converter, I think I should be fine with testing it.
Network Screenshots
C-Date general infos:
As you can see on F5 the offer is 30+. Adsimilis has a much larger pool of landing pages for C-Date, therefore I will be running the majority of the traffic to Adsimilis for the start to get some data on the landing pages. To keep quality up, I included 30+ years of age in the landers and sending those clicks to Adsimilis as they accept 18-29 as well. Would be sweet if ther was a setup on F5 to accept those leads as well, but right now I gotta deal with it.
Adsimilis C-Date
F5Media C-Date
Seitensprung / Fickdate general infos:
Like seen on the screenshots, the offer is supposed to rotate. This can be both good and bad, but I'll definately give it a try. Their LPs are much more 'adult' than C-Date and I like the layout. F5 has 2 LPs available which I will post later on, just for the sake of being easier to overview, I included just 1 offer screenshot. Adsimilis seems to have one offer link for both LPs only so in the campaign setup I got 1 link to Adsimilis and 2 links to each LP for that offer to F5.
Adsimilis Seitensprung / Fickdates
F5Media Seitensprung / Fickdates
Offer LP Screenshots
To make it more structured, I inserted 200x200 resized images of the offer LPs. If you want to see the full screenshot, just click on it 
C-Date Offer Screenshots
Seitensprung / Fickdates Screenshots
For those interested, here are some TV spots running in Germany for C-Date -> hence I've chosen that offer as brand recognition should be strong for German traffic:
Landing Pages Screenshots
So like I explained in the OP, in total, I am running 6 landing pages. I will share 3 of them here as the other 3 landers I've got from a friend of mine and I honestly don't feel comfortable in sharing them. It's ok to post self-ripped stuff, but if people help you, respect that and don't out. Just my opinion. Here are the landing pages I'm showing (click on each image to see in full size, easier to overview than posting full sized screenshots):
For the second LP, I included a geo script and a countdown. Kinda ripped the 'pof style' landers to see how they do on adult. I bet everybody has seen those already, but personally I haven't seen them running on adult so far. Worth a shot, we never know until we test!
Creatives
I decided to run on Pornhub.com first, therefore I followed the following advice found on other posts about creating adult creatives:
- Animated gifs. All gifs are animated to catch attention
- Matching the design of the site I'm advertising on
- Quality: using older females in the ads for the majority of time, adding 25+ / 30+ on the landers as well. No video type ads
- Amateur images. I browsed through a lot of sites and in my opinion the images look very much amateur. In case I'm wrong here, feel free to correct me please
So far, I haven't created creatives that are directly relating to the porntube / chat style as I want to get some data on those first. Definitely I'll try those two angles as well coming days.
So here are the creatives I'm using for a first run:
(Angles are pretty broad for those, everything from bang your neighbour to horny housewive needs a cock or wants to get laid)
Those creatives look pretty different to the creatives running on Pornhub so far (at least in Germany) but that could prove both, good or bad. A first test will show I guess.
One thing on the side, all creatives are 315x300 as Pornhub accepts those for my targetted spots. Obviously I'll have to resize them to take them to other sources & placements, but as for now, I picked them as a starting point due to volume being available according to TJ traffic estimator for Germany. Personally I feel like starting with 1 source / 1 tube should be the right way (at least for the very first test run) much rather than spreading over all tubes available on the net.
Here is the second set used for Exoclick (XHamster DE top right square feature placement):
Sources
So for the start, I'm sticking to placements on Pornhub. Pornhub is Alexa 48 in Germany so pretty much a shitload of traffic should be available here - the total impressions by Germany on Pornhub look like the following:
Overall Germany
Pornhub Video A Overall
Pornhub Video A CURRENT (13.00 CET 16.02.13)
Pornhub Video B Overall
Pornhub Video B CURRENT (13.00 CET 16.02.13)
Pornhub Topright Square Overall
Pornhub Topright Square CURRENT (13.00 CET 16.02.13)
With those 3 placements I am covering around 65% of total traffic that can be delivered by Pornhub per day, so volume wise, I think I should be pretty much fine for a first run.
From what I've read in a different follow along, that TJ 'bug' with a 20k budget is still active so I will settle my bid somewhere between the bids I can see there just to see how traffic comes in / if at all. Here's a quote of Maynzie I saw on a different follow along, so gues I'll have to play a bit to see what works (as always) first:
Originally Posted by maynzie
For the first run, I decided to stick with the Top right square placement for now. It's the 'root' placement that above the fold, therefore more competitive, but at the same time allows animated ads compared to Video A / Video B. To gather some data of Pornhub for DE traffic, I'll run for around $100 on this placement first and check a handful of other sources as well which are supposed to be less competitive for the German market, so expect an update with those new sources to come next days as well.
So far, here's a step by step walkthrough of the campaign setup for the top right square placement on TJ:
Select your placement:
Basic campaign setup (name / timerange / budget)
Advanced setup (Geo / Bids)
Creative setup / tracking links
As you can see, I've included 'tj-rightsquare-NO#' as the KW for that campaign in CPVLab. By doing so you can seperate the campaign performance easily inside CPVLab by sorting for the keyword. I strongly recommend everybody to set up similiar if your source doesn't provide dynamic tokens, as otherwise, you aren't able to track LP CTR + CR seperated by creative => losing the possibility to optimize.
After the campaign setup, I selected to set a frequency capping. I've set it to 4 for a 24h period. This means your ads are shown only 4 times during a 24h period to a user. JT uses cookie/ip detection to set this freq. cap so it hopefully helps avoiding banner blindness. Will be interesting to see how this cap influences volume.
As for now, the campaign is live on TJ and awaiting approval. Stay tuned for the first stats / updates and additions from other sources for my follow along.
Campaign Setup on Exo:
Despite pornhub, I've setup a campaign on Exoclick. Targetting xHamster (one of the most popular porntube sites in Germany according to Alexa):
First setup step. For those never running on Exo: the total volume shown is RON, so be careful about that number if you decide (as I did) to place on premium placements. On the next steps you can select to run RUN or premium.
The next step gives you some basic options like frequency cap, setting your bid etc. Again keep in mind if you run premium not RUN, you will have to adjust your bids in the next stage of the setup. So basically unless RON, the bid here isn't important at all:
Here's the important step if you want to target specific placements. Just hit the 'Premium' selection mark and you get a rolldown where you can select your actual placements. As you can see, you have to specifiy your bids in here again while selecting your placements (side note: right now my account is in Euro currency but I've already talked to our rep so we get another account ready for US $ during next days for easier cost/revenue tracking). In the screenshot my bids are all default yet, obviously you have to adjust them depending on the placements you want to run on:
Pretty much the final step is to add creatives and their tracking link. You can add a max of 10 creatives per campaign, so you gotta split your campaign in 2 if you want to test more.
As you can see, I am using {src_hostname}-BANNERNAME as they keywords in CPVLabs. This automatically inserts the placement of the banner while you have to edit 'BANNERNAME' with a unique identifier so you can see where your clicks coming from:
The campaign is awaiting approval as well and I'll post more stats and details in the other posts in that thread about Exoclick & xHamster placements.
Stats / Updates etc.
*** Campaign Stats ***
Traffic Junky (Pornhub) DE:
Campaign Setup Date: 17.02.2013
Placement: Top Right Square (root)
Initial Bid: $0.46 CPM
Spent: XX
Revenue: XX
P/L: XX
LP CTR avg.: XX
CR avg: XX
Budget for 1st run: $100
Landing Pages (total): 6 LPs
1st run creatives: 14
Offers: C-Date DE & Seitensprungarea / Fickdates
Networks: Adsimilis, F5Media
Creatives Stats:
To come with tracking screenshots
Landing Pages Stats:
to come with tracking screenshots
Some advice:
- Your landing pages could use a LOT of work. If you're going to use your own, I highly suggest split testing them with what your competitors are using.
- Also testing 6 LP's initially is a lot for $100 budget a day, I'd lower it to maybe 2-3.
- I think the images should "pop" from the page rather than blend in. It's all about grabbing attention. Use backgrounds bright colors rather than a black background, on a black website.
- Your idea of using older and amateur'ish women is definitely correct. It's going to take a lot of pictures until you start noticing "patterns" as to what works. Are you doing these images yourself? Maybe you could hire someone to help with the process and improve. I think the ads could pop more as far as the text, headlines, and call to action buttons.
Thanks for your input man, lemme explain why I've done it the way posted above. Still I'm open for all advice, so thanks a lot for jumping in!
Originally Posted by dr_ngo
Nice write up man! I'd have to agree with Charles about having a few too many variations off the bat. Although I do agree with you about making the creatives blend a bit into the site.
Subscribed to this puppy!
Originally Posted by Mr Green
Usually out of those LPs testing at start, from my experience at least 1-2 totally tank and can be removed after a first short test. I usually take around 30-40$ worth of traffic (in case it's low paying offer) for a lander to get some first idea of the CTRs, obviously if CTRs are fine but it's losing finetuning is the next step, but given I get like ~100 clicks with a CTR of 15% it's usually hardly worth to stick with it from my experience.
Ooooo the big Franzy's in town!
Awesome write up, very detailed looking good, also good offer choice I reckon you'll make this work.
Yah maybe too many landers off the bat, I do 2-3, and 10-20 creatives but if this is how you roll go ahead man ! We all have our own processes
DAY 1 UPDATE
So first data is coming in, here are some screenshots that show how what is performing:
NOTE: I've set the CPCs in Cpvlab to $0.01 for an easier overview so those numbers do not reflect P/L properly. I'm tracking P/L in a seperated sheet, the 'real' spent you can see on my summary / on the screenshots:
Trafficjunky - Pornhub
Stats
Spent: $ 23.30
Revenue: $16.20
Profit/Loss: - $7.10
All other stats on screenshot.
Exoclick - XHamster
Stats
Spent: $ 102.60
Revenue: $28.22
Profit/Loss: - $74.38
All other stats on screenshot.
DAY 1 Analysis
Trafficjunky
Sample size is a little small here with a very low spent. Still I see potential here for sure if I manage to get a ~ -30%ROI on my first run with adult dating without any optimization or anything done than setting up the campaign.
Here are my thoughts step for step:
What's good:
- I'm on the jump to breakeven. 1 more conversion only and the campaign was actually about to breakeven / smallish loss. Potential there.
- Although sample size is small, some creatives are performing much better than others. There's no point in keeping creatives that aren't converting and giving a CTR below I'd say %0.1. I love CTRs like 0.3% that are giving me clicks for around $0.06 - 0.07. Huge potential there imo.
- Conversions are coming from several creatives. By this, I could get kind of a 'template' to work with and use different images for optimizing.
- Landers are showing potential. Sample size is way too small, but everything that's profitable so far and above 25% can be tuned in my opinion. Especially after gathering more data, adjusting creatives and images on LP etc.
- Offers seem to convert nice.
What's bad:
- Sample size is way too small
- Spent is too small. I'm bidding close to top bidder but with a budget of $150 set per day. There's a lot of chat about that 20k bug on TJ, I wonder if I should up daily budget for that campaign to spend more. Must admit though that campaign wasn't even running 24h so far, so spending might pick up to around 70-80$+ if running 24h.
- No real 'wow' effects in terms of the LPs CTR. Room for improvement.
My ideas for TJ:
- Trying to get more traffic. I think I'll try the 20k daily budget 'trick'. Anybody experiencing problems with overspent? Like I doubt if your balance is 1k they gonna spend much more than 1k do they? Obviously I don't wanna run in credit for huge amounts. Might sound a silly question, but getting invoices from networks before for traffic I never ran I thought it's better to ask.
- Probably weeding out the poor CTR creatives. I'm curious about how many impressions / spent you usually take in adult to decide to cut or not? My idea is to spend around 5-7$ (offer payout) per creative and if there is no potential at all in terms of CTR / LP CTR for that creative cut it. There has to be clearly no potential, like a CTR of %0.05 and no conversions. Any advice on this is much appreciated.
- After getting a more reasonable sample size, decide on 3 top performing landers and try 2-3 diff. variations of them, means exchanging pictures, CTA etc.
Exoclick
Looks worse than TJ so far, so here are my ideas / thoughts:
What's good:
- Conversions came from 3 creatives only. Those 3 creatives have reasonable CTRs (%0.1 / %0.16 / %0.13 ) so taking those 'templates' and trying to get better CTRs for cheaper clicks could work.
- I've spent around €37 ( ~$50 ) for creatives that had very poor CTR. So how's that good? It showed me that you gotta watch like a hawk. Exo's distribution was kinda spread over 3-4 creatives at the start only which caused a huge loss. Actually with cutting those poor performing creatives earlier, I would have saved quite a % of the budget and that wouldn't look that horrible. Also overall LP CTR can be affected by those obviously, so it's definitely time to seperate.
What's bad:
- Pretty much poor stats after first run with a $100 spent.
- A lot of creatives not performing at all over a reasonable sample size. A CTR of %0.03 over a 17$ spent isn't going anywhere
- Landing Pages aren't performing at all (Btw. those are the same landers I'm using at TJ). Could be related to the non-performing creatives, gotta test that with higher CTR creatives to be confident.
My ideas for Exo:
- First of all, after getting some more sample for TJ, porting those creatives to Exo and see how they perform. It's all about the CPC in the end, with a strong CTR there are dirt cheap clicks to be expected.
- Create 3-4 different versions (new images, change CTA etc.) of the 3 performing creatives and see how it goes.
- Get reasonable sample size for those supposed to be performing creatives and take it from there.
Small update:
After talking to a friend of mine, I feel like I have to change the structure of my campaign setup a bit. The reason behind is following:
I've duplicated my initial campaign on Exo twice, one new campaign for each placement. Setting the budgets to similiar values, I noticed that the majority of clicks for a short-term period came from one placement while the other campaign has not received much impresisons at all.
As there is no dynamic token for placements on Exo / TJ, I'm leaning towards the following setup:
- 1 campaign per site per placement
- Budget of €50 per campaign per day, means effectively testing that campaign per placement, not over several placements like it seems to be happening right now (without being able to identifywhat placemnets the conversions are coming from)
- Stick to 4 landing pages for each test first. Means in my case right now, I'll take the 'top 4' performing LPs and test them placement for placemnet by seperating into campaigns
- 10 creatives for each test, in avg. 5€ per creative which matches the offer payout. Should help to eliminate total failing creatives by that pretty fast.
I'll update later on when those changes have been done on my end. At least I have some creatives that did not perform at all in the intial run, so I will stick to those creatives that have somehow showed results so far.
I think it's better for both, this follow along and better for my personal tracking if I divide placements into seperate campaigns. I'll have to somehow start collecting data from scratch per placement, but it definitely should be worth it while progressing.
All sounds good, for any creatives that look like they have some potential i would test them with more than 1x the offer payout. This is the case with any traffic/niche, need to give it more time to perform.
ex:
- you spend €5, and 0 conversions
- you spend €8 and then 3 conversions come
- you spend €12 and 3 more conversions come.
... and you may have tossed this banner.
Atleast go to 2x the payout for anything with promise and 5x or more the offer payout for each lander.
Originally Posted by fjk87
I agree with stack that you need more data before you can rule out any of these creatives - double the payout is a good baseline to go by. Even that can be pretty early, especially if you're dealing with a DOI where the user might not convert straight away.
Great write-up so far though. Looking forward to seeing how you get on with the optimising!
DAY 2 UPDATE
As promised, divided by placements on XHammer. Here are the stats, 'analysis' follows in the next post:
DAY 2 ANALYSIS
So here we go, my thoughts for day2. Seperated by placement again:
Note: the creatives used for both placements are exactly the same creatives. That's a mix of the former 'performing' creatives on day from TJ & Exo.
XHamster Invideo DE
What's good:
- I'm getting VERY cheap clicks. 0.045€ is around $0.06 which sounds very sweet.
- All creatives seem to at least perform well in terms of CTR. As you can see on the screenshots, some are running over 0.3% which gives me clicks for < $0.05.
- All creatives perform somehow similiar in terms of LP CTR. All too low, but that's good as I think I have to adjust the landers and eliminates one variable.
- The lander performing at around 28% CTR has a different angle to the other landers, seems like I should focus a little bit on this angle while adjusting the lander. It's the mix lander with profiles on it, so I'll try to modficate it a bit to make it look more related to the creatives.
What's bad:
- LP CTR isn't anywhere but bottom. None of the landers performed above 30% which is disappointing, but fixable by adjusting / changing landers. 300+ Clicks should be somehow a reasonable sample size to see the landers aren't good at all as it seems.
- Due to the poor LP CTR, I'm losing a shitload of clicks. I'm pretty much sure that with an adjusted angle on the landers, this can get profitable quite fast due to the cheap CPC I'm getting.
- Offers aren't converting great, although the sample size per offer (offer LP) is still too small in my opinion.
- Far from breakeven
XHamster NTVA DE
What's good:
- I have several creatives that have some potential according to CTRs. My avg. CPC is ~ $0.14 which which is something we can work with.
- All conversions came from creatives that have a similiar angle. Good to know to create new landers and increase CTR / CR related to that angle.
- Again the Mix LP is the best perfoming in terms of CTR although no conversions. The angle is, like mentioned already, related to the creatives more than the other landers that are more or less 'standard' rule landers.
What's bad:
- LPs aren't performing although better than on the Invideo placement.
- The best performing LP has no conversions
- Far from breakeven
What's next / my thoughts:
So as it seems, I'm with a situation we all have faced before: the LPs aren't performing. While the creatives are giving me cheap clicks, the LPs aren't converting them into leads.
The creatives that are performing are not any kind of video creatives which just catch the attention by assuming being a new video. They are all 'bang your neighbour' etc. related so that's at least the positive aspect of it. This means, I have to get new landers to turn the visitors into leads.
So as for now, before running again with TJ (seperated campaigns for placements) I'll get 3 different lander variations for another test on Exo. I'll try to match the tone of the best performing creatives / angle to create a 'flow' between the ad and LP. Hopefully I'll get them ready during today to avoid any delay.
If I missed something important or somebody has ideas on how to improve, feel free to post in here.
****************************
Originally Posted by stackman
Originally Posted by dieter
) This can be good for RON, but since I'm running on premium placements, I prefer to split into a seperated campaign. Originally Posted by Finch
Alrighty, short update, after getting 3 fresh landers ( own creation, not hit and rip ) with a different angle than pure rules landing pages into proper html/php, I restarted the campaign. Expect an update tomorrow morning
I haven't decided for now to post screenshots or not, shit gets ripped fast usually anyway but for sure I'll post screenshots of those landers at the end of the follow along / maybe during tomorrow as well.
For another split test, I will run both placements for €50 budget on a low bid and switching after that to an increased bit which is next to the top bid to see how this effects performance. I'll write down detailled stats about this test tomorrow as well.
DAY 3 UPDATE
Volume dropped of quite a lot, I guess it's due to the low bid. Those are stats after around 18h of running in total, so far from the daily budget especially for the Invideo placement:
Definitely not enough sample size to draw any conclusions, but there's a trend that the 3 new landers in terms of CTR perform better than my previous landers on those placements. Actually for the NTVA placement the CTRs look pretty much strong compared to what I experienced with the previous landers.
The CTR in overall dropped a bit compared to day 2, but I'm not a 100% confident on those as all those CTRs were actually achieved by bidding the min bid / 2-3 cents above the min bid as you can see from the screenshots (Exo has a min bid of €0.10 for premium placements).
*** Setup for day 3 ***
As for now, I'm going to test a different bid size to see how getting priority in the queue effects CTRs and hopefully CRs as well.
I've duplicated those campaigns in both Exo and CPVLab. I also added the 'feature' placement for a test run in another seperated campaign using the same creative for that placement too.
Right now, the bidding looks like the following:
Exohamster Invideo: matching the top bid (very low top bid)
Exohamster NTVA: bidding 3-4 € cent below the top bid
Exohamster Feature: matching the top bid (very low top bid)
I will setup a direct comparison to CTRs / CRs after finishing the test run with the higher bids set. Usually the lower the bid, the more behind your ads get in the queue and you're getting the traffic that's onpage for quite a while already, meaning you're actually not getting 'fresh' users which I think is poor for CTR and banner blindness and CR as well.
Regarding invideo & feature placement: those placements have actually very low min bids which I think is because it's lower quality traffic. The invideo pop ups when a video is paused / has ended and could cause quite a bunch of accidential clicks of people hitting the replay / pause button. The feature placement is only shown after doing a search on Xhamster, means a user pretty much knows what he's looking for and is likely to be more focused on watching a video than clicking our ads. Test will tell I guess but that's my first thought. I'm most interested in the difference of the NTVA volume & CTR / CR with a competitive bid, watch for an update later on.
While waiting for approval and data coming in, I'll browse through some less competitive sources and see what's available for German inventory on there. Exo/TJ is influating quite some time back, I've read several times that many guys advise to try less saturated sources, so that's something I'm going to look for today as well.
DAY 4 UPDATE
So like I said yesterday, I ran pretty much matching / bidding slightly under the top bid to get 'fresh' impressions and visitors to see how this effects CTR / CR. So here are the basic stats.
Surprisingly I didn't get any impressions for the in-video placement within 6 hours although bidding top, so I won't post any screenshots of that placement as there's nothing to show with 4 clicks on LP and 800 impressions over that time.
For the feature placement the CPCs I'm paying are pretty much high, I'd say it has to do with the banner being placed ONLY after a user is doing a search. From my experience with other sources, unless super targetted to keywords (which isn't the case here) these placements work good for niche but not 'broad' as the user pretty much is looking for something on the site.
The landers are pretty much good in terms of CTR but CR sucks so far. While writing this, I'm already running those landers with a different offer to split test and they are showing way more potential than the screenshots above. I'll update tomorrow after running for ~€100+ as I definitely see potential with a new offer running.
COMPARISON LOW TO HIGH BID on NTVA Placement
Although sample size isn't that big, I think it's quite interesting to have a direct comparison of bidding the min bid / bidding max or slightly under top bid. So here we go with a summary of what I was able to see split testing bottom to top bid:
I would have loved to have a more equal sample size but I simply wasn't able to get more traffic with the min bid although camp budget was set to €100 for both campaigns. Take those stats with a pease of salt due to small sample size, but the fact alone that CTR / LP CTR / Adlick CR and LPClick CR CLEARLY outperformed using the high bid should give everybody some food for thoughts. Comparing % wise the avg. CPC by min to high bidding the CRs clearly should make up for the higher CPC.
Expect another (for me, quite BIG update in terms of results) update later today / tomorrow morning as I changed some things and it seems like it's going to work. Breaking even for now without any kind of optimization, more details to follow
DAY 5 UPDATE
Quite a good samplesize for this campaign on the NTVA placement on XHamster in my opinion, here are the overall stats sorted by creatives used:
So as we can see, I'm about breakeven on a reasonable sample size. Right now, the next step is to optimize based on the stats I gathered, here are the overviews for source / tracking:
DAY 5 ANALYSIS
What's good:
- I'm about breakeven with the new sets of LPs.
- CTRs looking pretty good on the landers, especially for those profitable ads.
- I see that 3 different styled banners are profitable. A good base for optimization using those templates with different images and see how they perform.
- All landing pages are performing pretty similiar in terms of CTR & CR.
- Erovie seems to be performing very good.
- Only 14% of my leads are under 30 which should be good for quality.
- Due to this 14% being under 30, I think I'll be able to use F5 +30 offer for erovie without a 'select your age' function on the LPs (right now there is no select your age, just 'we accept 30+ only')
What's bad:
- The majority of the ads failed to be profitable at this point. Each ad had 2 different variations of it, means 5 'templates' for creatives were used of which in the end 3 seem to get profitable. 3/10 sounds bad, 3/5 templates showing potential sounds better though.
- Ad CR could be improved for those profitable creatives.
- I been running on one offer only after testing 2 other offers before. I want to get this first campaign and the follow along profitable, but still obviously there's need to split test offers.
Thoughts & Ideas:
Ok so here's my plan for today:
- Create 3-4 different variations of each creative 'template' and see how that changes overall performance.
- For today leaving the LPs untouched, then analyze results with the new creatives. If it gets profitable, it's time to split test a handful of small edits on the landers to see how that effects performance.
- Setting both networks to 50/50 split as both are converting pretty much strong. This also helps for now as the offer is capped at 100 which I'm not sure to hit today but in future for sure.
- Getting 100 leads to each networks and ask for a quality check.
- Watch for other, more adult style offers to split test (for tomorrow / coming week).
- Going to spy some in other countries to get food for thoughts. Primary goal is to optimize this campaign now but spying is never good as there might be some idea you can rip... eeh sorry, I mean find inspiration in.
- Switch my Euro account to US Dollar account on Exoclick for easier overview. Already done that, waiting for them to approve my VAT ID to avoid extracharges.
No direct link testing at all?
Well it's been proven lp is most likely required in this niche.
Awesome write up man!
I would always run landers in Adult, it keeps your campaigns more stable and seems to improve quality
DAY 6 UPDATE
Today I'll update twice, running some of my budget forenoon. So far, no spectacular results with the new creatives, but I'll give it another shot for a $100 starting at around 6PM today. So here's the first update running during forenoon on my new Exo US $ account:
I will resume at around 6PM today and change some settings as well:
- right now, it's running at freq. cap of 2/day. I will switch to one per day as I'm getting around 25% duplicate clicks according to the networks. From my research on F5/Adsimilis, none but 1 conversions came from a duplicate click so by setting the freq. cap to 1/day I could maybe reduce my costs by 20-25%. Worth a try.
-I will add some 'you must confirm your account within 15 minutes' phrase to the landers for the evening run
Other than that, I will pretty much stick with the settings as for now.
In addition, I'll setup another campaign with yesterday settings (same creatives and freq. caps) that was about breaking even on a different source which I'll update later on about as well. Exo/TJ are probably among the most competitive sources to go with for now so with a less competitive source and similiar performance I could get into the green. Again, watch out for those two updates later on today.
cool fallow along man! Looks like you are going pretty deep with analysis 
. Also nice tip about freq. cap. , the camps I am running right now also have lots of duplicates clicks and I was wondering why hehe. I may change freq. cap to 1/day and see how it goes.
Nice follow along, hope you'll hit consistent profit
Mehdi
DAY 6 UPDATE #2
So here are the total stats for day 2:
Still in the red as we can see, actually I've spent $168.04 while revenue was $98.13. Now the good thing is, I got a total of 4 creatives that are profitable. This being said, I will run this campaign with the following setup today and see if I manage to grab some profits:
- 4 creatives that are profitable so far
- 3 LPs as according to CPVLab, all but 1 conversion for those profitable creatives are coming from 3 LPs.
Due to some hickhack yesterday I wasn't able to setup the campaign on a new source unfortunately. Going to see how the 'optimized' campaign performs today.
The CRs % for those 3 landing pages and the profitable creatives are all > 2.5% (Ad Click CR, not LP CR) so this gives a lot potential for getting in the greens.
Much rather than duplicating the current campaign on a new source, I will take a shot with a slightly modified approach. Let me explain:
How my setup looks right now:
- generic landing pages (means no offer logo, mix of young/old etc.)
- running Erovie right now but I tested 3 other offers as you can see in the previous threads
- I was testing 3-4 LPs and 10 creatives at a $100 budget per day
How my future setup will look like:
- more 'themed' creatives and landers. Means if the says bang a big wale, the LP will be all about about big wales
- adjusted landing pages matching the offer landing page style. Trying to create a better flow between LP and offer
- for testing, I will go with 2 LPs and 5 creatives with a $60 budget per day. This should give around 2 x payout spent per creative and hopefully reduces the amount need to be spent on testing only while at the same time giving me some reasonable sample size per creative / per LP to work with
Next update is sure to come, any suggestions / ideas / words of wisdom are always welcome!
DAY 7 UPDATE
A small update first without screenshots, those are going to follow tomorrow. So far I've been running for ~ $90 spent, here are the stats so far:
Spent: $89,4
Revenue: $117,19
P/L: +27,79
ROI: 31,09%
I've cut one ad that was about -50% ROI after a first test yesterday and has reached a total of 3.5x payout spent now, so right now, only profitable / breakeven ads are running (total of 5 ads running). Excluding the losing ad from stats I'd be at ~43% ROI now. Hoping for those results to hold up and will update later later today / tomorrow morning with some more detailled analysis.
Going to resume at 7PM CET for another $120 so with those optmized ads I will run todays test for a total of $200.
In to the profit... awesome!
This is a great follow along for any newbies getting started. It just goes to show that some persistence and careful optimisation will eventually bring you out of the red. Hopefully you can edge that ROI upwards and then hit the Scale button.
DAY 7 UPDATE (total stats)
In the green! Profit wise, nothing spectacular, but here's the overview. Detailled stats by creative / LP to follow in next post under 'Day 7 Analaysis':
$207.82 spent with a revenue of $243.06. Smallish profit, but there's room for analysis and optimization. Next post with detailled breakdown by creative / LP etc.
DAY 7 ANALYSIS
Alright so here are some stats for day 7 sorted by creative spent/revenue:
One creative is killing it, one is profitable by 26% which isn't bad but great neither, one is at 8% which isn't too much - too small sample size.
For a better overview of what's working or not, here are the creative stats on a larger sample size running for this campaign:
The total overview looks way worse than the overall overview. Now let's divide by LPs as on day 7, I cut out the worst performing LP to see how this effects performance based on previous stats. The stats per LP are calculated by avg CPC for the creative from Exoclick from the total stats, so it's kind of a simulation for running only with the LPs German mix 1, own new lp1 & own new lp2:
Now while this stuff is definitely an overkill doing after that sample size, I just wanted to lay out my process of optimizing here. You can do this MUCH faster by using CPVLab only and taking notes on an oldschool paperblock, but just that everybody can follow.
Basically I took the 5 creatives listed in the screenshots above and calculate stats for them seperated by creative&campaign level (above the black bar in the excel screenshot).
Taking it from there, I see 3 creatives are profitable / breakeven on the 3 top performing landing pages. Below the black bar, there are stats listed for creative & lp combo and we can see that the total profit is coming from LP2/MIX and LP1 in total is in minus. LP1 in this case is not performing for 2 creatives and the lowest performing LP for creative T1, so I decide to cut it here and focus on the winning combinations.
At the very bottom of the sheet you can see a simulation of my P/L if running only those 3 creatives on LP2 & MIX LP. With a 136.89 spent and 200.91 revenue, it's a 47% ROI without on-page LP optimization like switching images / changing CTA etc.
The sample size for this two LP winning combination should be reasonable with an avg. 11-12x payout spent per landing page.
Bottom Line:
So today I'll setup the campaign with the 3 profitable creatives and 2 landing pages that are profitable for those for a total of $200 budget. The way I've setup the landers it doesn't include an option to chose the users age, so to avoid sending non convertable ages to F5, I'll reduce the % to F5 Erovie which only accepts 30+ and try to run as much as possible to Adsimilis for now. I have 2 other networks having the same offer at almost the same payout running hasoffers so some options to spread the traffic left.
Updating later today with some first stats and at evening / tomorrow morning with total stats. If the ROI keeps somehow solid at 40%+ I will be looking to scale as fast as possible to milk as much as I can.
Beside that, I'm going to setup a new campaign and keep creating / spying some landers today. Been a bit busy during the last days, so I wasn't able to hold my word for a new campaign as I planned before, but sure it will come!
Definitely very helpful follow along!
my question to you: how much time you need to spent $200 with your current campaign? i mean, how much impressions / clicks you get per hour?
Originally Posted by drughi
With that being said, bidding very close to the top / top, I was easily spending $50-60 per hour. Keep in mind, that's ONE placement though. Also depends on how many creatives you run / frequency cap etc. Without frequency cap (as personally I've set one) you should be easily able to spend $120-$150 in Germany alone per hour.
DAY 8 UPDATE
Unfortunately not the stats I was looking for, some explanation below the screenshot:
So I'm back to about breakeven / small loss with 194.99 spent and revenue of 183.86. This looks pretty strange given the better performance of the previous day, but has one reason: bids influated on Exoclick yesterday.
Until yesterday, the top CPM was around $0.30 for all the major placements with in-video placements being some cheaper. Since yesterday, whole German exo is covered by a flat all over the premium placements CPM bid of $0.50.
Obviously with the previous stats there's no sense in matching / upping my bid. While it's bad on one side as I'm getting less exposure / less fresh impressions, it proves one thing to me: the higher you bid on Exo, the fresher impressions you get. No rocket science I'd say, but the difference in performance clearly shows bidding higher close to the top bid comes along with a better CTR and especially a better CR.
Nothing comes cheap neither does quality traffic. At this point, I'm not going to continue this campaign on Exo as I was left with only 1 profitable creative at this point but much rather focus on a new campaign on a less saturated source.
As I still got funds left on Exo / TJ, I might try a cheaper placement again with this campaign. Funds are there to stay so I will use them for a new test / later campaigns.
For today, I'll be looking to pick one or max two sources to test with a new campaign based on my experience from this very first adult dating run I did. I'm about to write up some summary of what this first campaign seems to have taught me in the next post, maybe some other guys find it useful and save some time / cash by thinking about those points
Alright, so here we go with a summary of what my follow along / my first adult dating campaign has taught me so far:
1) Stay persistent. If you read my follow along, you can see that from some bigger loss I realtively fast made the campaign profitable (although for a short term) / breakeven. Be sure to test and don't be affraid to lose money, especially not in the first or second run of your test.
2) Structure your testing. There's a shitload of posts about testing offers on STM, it all depends on your experience / budget which route you're going to take. This post by Finch gives a great first view about cost effiecient testing.
Given the response I received in this follow along, a good starting point for your tests in adult could be:
Landing Page: 5x offer payout minimum per lander.
Creatives: at least 2x offer payout with anything that shows promise. Means, a reasonable Ad / LP CTR. Obviously if a creative tanks for 2x spent with a CTR of %0.05 it's unlikely to turn into a goldmine. Better 3-4x payout per creative to see how it performs. Shortlink to a post by stackman and a post by Finch in this follow along about this topic.
Offers: You need to have a reasonable sample size to judge if an offer is working. There's no sense spreading 400 clicks over 8 landing pages with 10 creatives being used. Too many factors to really know an offer converting or not. Finch made that great post about efficient testing linked above.
Many factors come into play here, but the usual chat up with your AM and a mix of spying should do a good job in giving you some idea what's working. Personally, given my results on this follow along, I would stick to 2 offers for a first test that are recommend by your AM and running on adult already. Keep in mind that different offer landers perform different. I made the mistake by spreading over a shitload of landing pages in the start probably wasting time and money on data that wouldn't tell me enough to make any reasonable decision.
3) Get organized. Having a piece of paper for some quick notes is something I'm using fairly regularly, not only in AM. Make sure to use spreadsheets in a way you can easily enter/edit data and overview / compare at the same time. There's nothing worse than an unstructured setup pretty much like I started in this adult campaign by spreading too much in terms of LPs / creatives / sources.
4) Think twice. I made the mistake that I haven't set up landers that allowed the user to select their age. By doing so, I'm facing 2 main problems after talking to my AM and thinking twice:
- I'm unable to send traffic without a potential loss to offers that accept for instance 30+ only and don't have a seperated payout function like hasoffers (cake has not)
- Although so far I'm sending 75%+ 30+ leads to Erovie, I'm kind of unable to control it. Means, I could use a lower paying, not stable offer for those worse quality leads and send 30+ only to Erovie in this example to stay on the offer and hope for a bump.
While not all landing pages are using those options for age confirmation, it's definitely something I will consider when setting up a new campaign.
5) Don't assume. Test. My best performing lander is pretty much different from the usual rules style lander only. Although I wasn't able to scale, with numbers reaching 50%+ LP CTR and CR of 2.5%+ for some creative/lander combinations, it shows that hit and rip isn't the only thing that's going to work. Same goes for creatives. The creatives performing best for me I personally never supposed to do so.
6) The bridge between creatives and landers. Something I definitely messed up in the first run. There's hardly a way (at least in my opinion) to put up a creatives mix between bang your friend's mom, bang your neighbour, bang a BBW and bang a cheerleader. You could include all of them on one lander, but it's pretty hard to accomplish. I'm not talking about niche here, but much rather about some kind of 'template' your following through your creatives to the LP. Means if you're creatives are all about hot MILFs it doesn't make much sense to put BBW on your lander only. I think there's a way to 'keep it general' but at the same follow a certain idea creating a bridge between your creatives and landers.
For example a '100% amateur sex' approach is likely to be better adapted by a lander & offer attracting a mixture of ages than an angle focusing on emo girls only where the visitor at least must have some interest in emo girls to even consider clicking your ad/LP.
7) Focus. I must admit I wasn't a 100% focused in the first 2 days and therefore have done mistakes that made me smile. If you don't have the time to really focus on something, better schedule your tests for a later point. It's hard to stay focused when you got a shitload of other stuff going on, especially as in my case, stuff that was very important to my personal situation both, financially and family wise.
Those are just some short thoughts I wanted to write down after I went over this follow along step by step again.
For today/coming days, my priority will be to test a less competitive source than exo/tj with a better planned testing stage. As described above, you need to have a clearly structured plan for testing, which in my case, will look like this:
- Budget per placement / day: $60 for a first start, scalable if I want more data / see potential.
- Creatives: 5-6 creatives to get around 2x offer spent (given an offer pays ~ $5.00-$6.00) per creative on that placement
- Landing Pages: 2 LPs for this initial budget, otherwise I won't be able to draw any conclusions due to small sample sizes of creative to LP clicks
- Ad-to-LP bridge: trying to match the theme of the ads with the landing page in a better way than I did so far
- Offers: 2 offers to test with no more than 2 offer page variations.
It seems like these sources where people instantly know they're being outbid are really really dangerous and affiliates enter a race to the bottom so quickly.
This is the one area where Facebook was way better.
Yes, I wrote a post about these types of low payout offers and running them on Exo or TJ. It would very hard to profit daily on these $2-$4 offers. The bids are hovering around .40-$1.00 on premium channels and as you stated to get fresh impressions you constantly have to bid higher which is a losing game. If you were promoting an offer with a payout of $8 and above than that may be doable.
Originally Posted by mi6430
The main problem I faced yesterday was the influation of the bids. Going from $0.3 to $0.5 overnight (ALL premium placement) means 66% higher costs if you want a similiar quality traffic. Simply not doable with the ROI I had before, even with the best performing creative.
As it seems a lot of networks do not have bulk German inventory available. I've browsed through 4-5 networks so far, and a lot of them have low-medium volume available, but the problem is that those impressions are spread over 100s of targets with each placement giving you 2-5k impressions per day. Definitely something I'm not looking for and I've picked 2-3 networks that seem to have reasonable volume by specific placements.
I'm getting one new lander done today and modifying another one to follow my ideas described at the latest update. Meanwhile, as I've found a placement on TJ that isn't overbid and looks ok (not in-video, no feature/search placement) I've setup the best performing 6 creatives of my first run on Exo with TJ to see how the results will look after around $60 spent. The max bid is quite a bid lower than the max bid on Exo - should be a good comparison how similiar placements on different tube sites perform with the same setup.
If everything goes smooth, I'll submit the new lander and the modified one to a smaller network for a new test with a modified angle and LP design. Updates to come!
For the placement described above, I'll be split testing Erovie and Seitensprungarea for a budget of $60 today.
Yesterday TJ had some issues with their tracking, so I decided not to run as I don't like running a test without being able to see stats at all. Landing pages are ready so some update coming later on with a new traffic source.
Had a post about ero-advertising here as it seems like their interface had a bug. After getting on phone with my rep, he pretty much explained the situation and said that the live-chat support was making false statements. Ignore this post.
Originally Posted by fjk87
Alright, some lack of updates from my side, mainly due to 3 reasons:
1) Ero-Advertising being unable to verify my VAT ID so far, therefore I cannot deposit. Not possible to run.
2) The only payment processor used by Adbucks hasn't credit my CC deposit for 24h now. Gave them a call and supplied a ticket, but it seems like their support hours listed on their page aren't accurate, at least not for the weekend.
3) TrafficForce not having a lot of inventory for Germany, at least not according to what I can see on the interface. Plus, most of their placement are pretty expensive in terms of min CPM, so it's not really a less competitive source.
Anyway, for those interested in adbucks, I've done some short overview of how your campaign setup is looking there with a small goodie I realized while checking bid sizes:
Step 1: Select your placement
I took the route via the 'estimate traffic' button at top right which brings you to
Step 2: Estimated traffic
On the next step, you can select some additional options and can see the traffic in detail sorted by sources:
Step 3: Select campaign options
One small note here, when selecting geo targetting, be careful about language. Most tubes aren't translated and there will count as English language but the traffic can be from Germany.
Step 4: Select your placements' bids
Here you can finally select your bids for each placement. There's one important thing to note: the green and orange buttons are supposed to set your bid to 'winning bid' and 'recommend bid'. Now while this looks like a cool feature, be careful: at some placements, I've experienced the winning bid being lower than the recommend bid, so always double check about this selection which bid looks better
Step 5: Upload creatives to your campaign
Pretty much self explaining. A great feature by adbucks is the ability to create your own makros to track. They offer a handful of pre-defined dynamic tokens as well:
In terms of my new campaign setup, I've taken care of a lot of stuff I mentioned in a previous posts. The new angle I'm trying out now is targetted for an older demographic (mainly 30+) with the creatives matching the landing page design much better than before.
I'll get on phone tomorrow morning with both adbucks and ero-advertising and try to get my stuff fixed to be able to run some firsts tests on their platform. In case it will delay even more, I'll give this new campaign a shot on TJ/Exo and see how it performs there trying out new placement than XHamster.
Summary
So as the contest is over, I'm writing up a little summary of the experience I gained during that contest. I wasn't at home much last days therefore haven't launched any new campaigns on the less crowded source which is going to follow obviously, but still, here are some thoughts:
- There's a shitload of volume available for cheap on adult tubes. Find a set of winning creatives and combine them with a good performing LP and you're probably likely to get profitable pretty fast. Anything but rocket science, imo a pretty ok niche to start with if you don't have any other background in AM.
- You get what you pay for. I've done some small tests on Exo just for $30 each with the same creatives matching the top bid and bidding 50% below the top bid. In that test (no screenshots posted on this follow along) exactly the same creatives more than doubled their CTR (every creative!) matching the top bid => cheaper clicks than bidding low with crappy CTR. In addition, LP CTR also improved by around 8% over this small sample size for almost every creative. Bidding high might seem expensive at first but can be cheaper overall. In the end, it's about the LP CPC much more than the network CPC as those are the clicks we are looking to bank from.
- Set your plan for testing. Described that in a previous post in here. Set your budget, adjust the number of creatives / landers you're planning to test to get significant data. Testing 20 creatives with a budget of $100 with 5 LPs won't give you anything.
- Don't use free / free right now etc. in your creatives. I've had good feedback about the quality of my leads so far and I'm supposed to get bumped over a bigger sample size => higher ROI / profit withtout any additional work involved.
- Match your landers to your offer: if your offer is about cheating wives, create an angle on both your creatives and landers that somehow matches.
- Outsource stuff you lose time on: personally I'm not strong on coding so I got a VA doing this for me. If you find a good, reasonable priced and (BINGO) intelligent VA, pay him on time, make small extra payments of a couple of bucks. While it seems shitty to you, believe it or not, those people actually care if you pay them $20 or $25. And most important, they'll instantly start on projects you assign them as they know they are paid on time and higher than average (while still pretty cheap for 'us' standards).
- Neither lander or creative will decide if your campaign is a winner. I've had landers performing on 50% ROI with some creatives while tanking for other creatives. Always test more than 1 'template' for your creative per landing page, you never know until you test.
- Adult traffic works for non-adult. I launched a make money online campaign (using roulette method) on a pretty big source 3 days ago. Since yesterday, it's spending around $350-$400 on a 50-65% ROI. Proven campaign from a different source but shows that adult traffic isn't only suitable for adult offers.
- Semi rip. It's not about ripping a person's lander. If you do it, make sure not to copy it step by step. I've had best results with mixed stuff (mixing rules with profile pictures / stuff that's running for ages on adult tubes but modified a bit, / matching the offer etc.). While it can work for sure you're likely to be at the end of the line by just copying and running it.
- Read other follow alongs. Obviously this helps while analyzing your own stats and finding mistakes you've done before. While a sudden rush of follow alongs popped up due to the competition, I hope for some guys to continue their follow along. At least I will do so and update several times a week with results I'm gathering.
- Start your own follow along. What I noticed is that you're analyzing much more in detail and keeping an eye on details when posting a follow along since you somehow don't want to post obvious / lazy mistakes in your follow along. It doesn't take that much time to update once or twice a week but I bet you'll receive some feedback from more experienced guys which will nothing but help.
That's basically it about my 'summary' for this follow along. Like I said, I'll def give adult another shot. Potential is there, lots of traffic, lots of offers, tons of supportive people on the forum, what more to ask?
By continuing this follow along hopefully I'll be able to show some totally fresh faces that it is possible to get profitable without any proper experience in that niche by using a structured setup and analyze in detail. If it helps others / motivates, the job is done!
Great follow along, I like your analyzing methods and definitely the adult tubes will be my next big traffic source to master.
By the way, what do you mean by saying "roulette method"? Some kind of rotation?
Originally Posted by peanut
Hi fjk87,
did it happen to you that ExoClick shows only a few creatives from 10 uploaded?
As I understand there is no tool to set a rotation weight as TJ does.
This happens usually if you got a lot of creatives they spread it out uneven. Let it run a bit more, but honestly spoken on 1 mio impressions it should have been even a little more, never experienced it as you did in the screen, just in the beginning of a new campaign starting.
Now it 2 mil impressions, paused first six banners in order to get some traffic to the last four.
EXO allows rotating up to 5 banners if I'm not mistaken.
Originally Posted by bbrock32
10 banners for sure, my second campaign got 1.2 mio impressions yesteday and rotated 10 banners evenly. It seems there was a bug with the first one.
Thanks for the detailed follow along fjk87. About to take a crack at TJ and this thread gave me some great tips.
Home > Adult Traffic (NSFW) > Follow-along Campaigns