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

---

The Italian Job - 2018 Adult Dating Follow-along (23)

---

UPDATE:

I just noticed the postback URL from my Affiliate Network sends my payout back in EUR, and this is screwing my stats up.
I set it to manual for now, but now I have to re-work the old data and do some calculations in order to figure out exact ROAS.

If you guys know how to automatically update the payout on old stats in Voluum, please let me know

P.S. I will be adding more details to the optimization process & how I cut out the offers & calculated the thresholds shortly, I just wanted to put this up to get inertia working in my favour LOL

Hope some of you will find this follow-along helpful and/or entertaining :-)

---

Great job at documenting what you've done so far Let me post some comments ...

List your campaigns in Voluum, select the date you want to alter, select the campaing and click the "update costs" button.



Enter the cost for the entire period that matches the date you selected and save.

.
.
.

50% CTR of an LP is very high, 30% is usually enough to reach profits. LPs with higher CTR usually contain some misleading element, which will lower the CVR in the end. So if you feel like you can do better, aim for 35% or so, 50 is to much IMO.

I see you are targeting quite a few placements on TJ and still mention rather low amount of traffic. Let me give you one tip, TJ uses the account balance and daily budgets a lot, if it's low, they limit the volume. In case your account balance is lower, it's better to target low amount of spots and give them a reasonable budget ... instead of spreading it out too much.

This is what I would do next ... focus on the best LP, prepare more banners, pick just a couple spots (3 max) and test on them.

---

Hey Matuloo! Thanks for the wisdom as always, your help is much appreciated.

I will be analyzing the data on the banners, and pick the spots that gave me the lowest CPC

I am having a lot of trouble setting up the conversion postback between Voluum and TJ.

Have been following their guides, but I can't seem to make it work.
I am planning on contacting support tomorrow morning.

Will keep you guys updated on that one too.

Wrapping the day up, will be posting campaign stats tomorrow

---

You don't need to pass conversion data back to TJ, I never do that either ... I like to keep as much of the data private as possible

---

Oh ok...and so how do you track which banners generated a conversion when split testing?

---

Ciao Xechel0nx! e benvenuto su STM!

---

TJ has a token for this, use it.

Here is a screen for you, these parameters are basically all you need in order to be able to optimize TJ. Configure TJ in the "traffic sources" section like this :



Then you will be able to analyze your campaigns based on these tokens.

---

Ciao brokerbabe! Grazie mille

---

Awesome Matej thank you very much for taking the time

---

This approach will not get you anywhere in adult, I'm sorry.

Adult is an IMPULSIVE vertical, which means there is bigger stats error than in other verticals. I'm not sure how many conversions you based this analytics on, but I guess not that many, which means your predictions are most likely wrong.

You tested across many placements, with small budgets for each of them, during isolated hours due to the budget ...

As I told you earlier, focus on a limited amount of spots and spend more on them. Then you can scale to more of them. Spending $5 per spot is not gonna give you any insights. In adult, conversions happen randomly to some extent, you need MANY of them to spot clear trends.

What you need now is the best possible offer you can get, the best possible LP (2 or more preferred) and a set of well performing banners. Focus on one larger spot and find it there. Scaling to other spots comes later. Do not predict, assume or expect ... let the data show you the way

Matej.

---

Alright so first things first, thanks for the reply.
Now for the sake of clarity let's define a couple things:

Well Performing Banner:
According to what I understood reading the forum, a 0.2% to 0.4% CTR is considered to be a good performance

Well Performing LP:
30%-35% CTR seems to be the sweet spot

"Best" Offer:
I don't have a clear picture of what a great offer looks like,
But I am thinking conversion rate & payout are the most important parts.
I have been looking for a low payout SOI offer.

My reasoning is the following:
Low payout because it's easier to understand if the offer is working without spending large amounts of money.
Also because if an offer converts really well, the street payouts for new affiliates is lower

In my case it's a SOI 18+ $1.11 payout 24% CVR (on offer clicks)

Have been split testing it with 3 other offers

Cutting methodology: 10 times offer payout spent with no registered conversions.
I gave myself a penalty here because I was also testing LPs. I've found you generally wanna do 4-7 times.

Now I've found a seemingly 30% CTR LP, so no need for the penalty anymore.
Now the focus is on finding the first placement + spot that generates a positive ROI.

Same cutting principle applies to a banner placement.

Q: HOW MANY SPOTS ARE YOU RUNNING YOUR BANNERS ON?
A: 4 Right Now

Q: HOW DID I PICK THEM?
A: The best performing ones in terms of CTR / CPC


Matej:

I see what you're saying here:

Pick just one Ad + Placement and Banner and stick to it

The part that resonated with me the most is the fact that I could test the banners only on a set
number of hours throughout the day due to the limited budget.

But if this is true then I don't understand how statistics come into play on validating a spot.

On some placements + banner combinations I got 48943 impressions and no clicks.
This is a clear indicator I should cut the placement out from what it seems to me.

On other placements I got 1011 impressions and 4 clicks
I flagged the placement as green for further testing.

I have also considered the CPM and CPC, if a placement has a low CTR but vert very low CPM
it could also be considered good as long as it generates conversions.

But for now I stick to the ones that looked the most promising ones.

Am I doing something wrong here?
i.e. how do I pick a good spot without first measuring it's performance?

---

Oki dokie, let's dive more into it.

You're on TrafficJunky ... this is a clean adult network with pretty much only their own sites. All proven brands, that rely on organic traffic. As such, you do not need to worry about the placements/spots, they are all legit and can work. Except for the tier3 RON, stay away from that now. This fact allows you to worry about one less thing, you don't need to look for spots that can work and label them as green, just stay away from the RON and you're good to go. Of course the performance will be a bit different from one spot to the next, but Im sure you understand what Im talking about here.

What I'm suggesting is to focus on a limited amount of spots, or even just one, because you need to properly test it. This means 24/7 of traffic, with no big limits ... On top of that, TJ limits the delivery based on budget and account balance. The more spots you select the more budget and balance you need, otherwise your delivery will get throttled. And you don't want that, you need to see the full potential of the traffic, not somehow limited delivery.

Use this spot, or limited amount of spots, to test your funnel ... so banners/LPs/Offers and try to identify the best ones. The best ones are those with the BEST ROI ... not the CTR. CTR is an indicator to follow, but profitability is more about CVR and the final ROI. Banners with lower CTR often outperform those with high ones and the same goes for LPs.

Based on what I already said, this might or might not be true. CTR is just an indicator that is not the most important one, ROI is. 0.4% banner CTR is already extremely high on many spots, which indicates misleading elements and conversions will suck. 30-35% LP ctr is also VERY solid. I'm looking at my campaigns now, I have some where banners are around 0.1% and LPs with less than 20% ctr and these are still profitable. Do not get obsessed with CTRs, it can be very misleading.

As you mentioned yourself, even low CTR can work, when the CPM is low too ... keep in mind this is all based on BIDDING.

Obviously, banners with lousy CTR cannot be profitable and the same goes for LPs ... when the CPC goes to something like 30 cents per click, it will be hard to profit from it, it's still possible in some GEOs though. But as soon as you cross 50 cents per click, it's game over pretty much. Or when a LP only get's like 5% CTR ... impossible to make it work again ... UNLESS, the clicks are extremely cheap. Its good to watch the price of a LP click (so click on the cta on your LPs), if these cross certain limit, you won't be able to profit as there are certain possible limits to the CVR of any offer.

This is solid thinking yes. Low payout SOI offers are the best to start with.

Cheers,
Matej.

---

Alright so since I want this thread to be useful for guys like me starting out in affiliate marketing I will be posting my thought process as well as my actions, because I think my doubts are pretty common among n00bs.

I have been reading some premium posts at finchsells.com

It's really great stuff, Finch really knows what he's talking about when it comes to adult.

And I have found a lot of similarities with what Matuloo was saying to me.

When gathering data for a placement you have to consider the day & the time of the day because there's big fluctuations

That means the data is useless if you test a lander say on Friday from 9am to 6pm and compare the stats with a lander you tested on Monday from 1pm to 7pm

There has to be consistency among the tests.

Finch recommends at least 1 full day(24h) from 9am to 9am to perform an accurate test that is somewhat reliable.

__________________________________________________ __________________________________________________ _______

MISTAKES I HAVE BEEN MAKING:

• Testing multiple offers at the same time with a limited budget

Actually this is something I took away from John Crestani's videos on youtube, he said you can perform "rules of thumb" tests
But now that I look at the stats, it was just too few

• Making decisions on data that is not statistically significant


THE QUESTIONS I HAVE THAT ARE MAKING ME PROCRASTINATE

1) Since everybody is using the same rules lander, advertising on the same placements, how do you beat the competition if you cannot differentiate yourself in a significant way? The adult world seems to be pretty limited on the choices you can make

From what I figure you can either:

A) Get better offers or better payouts then the competition

This is something you can do when you have leverage generally

B) Out-hustle

Bidding for lowest traffic and testing different angles that appeal to different niches or pockets of traffic that the big spenders are not bothering to target. So monetising what these guys are leaving on the table basically.

C) Innovate

But this is a riskier approach and also not suitable for a newbie that doesn't have benchmarks.


I've also understood the differences between the whitelisting and blacklisting approach

I have been using a whitelisting approach, this approach doesn't allow you to have contractual power with your AM since you are driving very little traffic to the offers, but it's less expensive then the blacklisting approach.

RULES OF THUMBS ON DATA (averages)

Banners CTR: 0.1 - 0.4
Lander CTR: 15% - 25%
Offer Conversion Rate: 10% to 15%

Now...the reality check taught me I have less knowledge about running traffic via Display Ads then I thought I had.

A) Burned $400 not understanding the importance of frequency capping...don't do that!
B) Spent another $100 on multiple placements and making decisions based on data I gathered in a non-consistent way.

I will be regrouping before I start running traffic again

---

This is mostly because we are talking about ADULT! Weekends, late night, couple beers in the head ... now compare this to a regular office day, 11 am on wednesday... people certainly think and behave differently The interesting thing is, you can be profitable on wednesday, just as on the weekend day, but the performance will be different.

Where Finch recommends 1 full day, I'd opt in for a full week, so you can spot the weekday-weekend difference. Certainly dont try to apply a 2-hour test results on the whole campaign.

More offers, more budget required ... this is an universal rule. John was into Nutra, thats different from adult dating again.

The logic of the LP is the same, but there are many ways you can skin a cat, right? Change the headline, the females on the LP ... this can affect the performance a LOT!

People in adult tend to be lazy and copy each other, quite often it's just about making SOME changes, so it looks a bit fresh.

And then the OFFERS ... not all of them are created equal, you need to find the good ones. SO test test test ...

Cheers,
Matej.

---

Redtube is a good site, good CTRs though usually higher CPMs required. Its one of the smaller sites on TJ, but there is still enough volume in certain GEOs.

As to your projections ... these can be accurate or totally off, I'm not a big fan of going this route, you will have to test it properly to see the real numbers.

I know that there are offers with higher payout than 1.1 in IT, try to find them. Dont expect to hit the jackpot with the first offer you try.

The CPM of 0.10 is actually pretty low, it can be way higher in some GEOs and on certain spots. Looking for cheaper traffic is not the way to go here, Im afraid.

That's an option too, but I think it would be giving up to early. You need to test many ads, lps and a solid bunch of offers too.

---

REBOOTING.
I want to commit $1000 to making this campaign work.
I picked the current offer in a wrong way since I didn't send enough clicks to it.

Here's what I plan to do:

A) Split test the current SOI offer (payout $1.11) with a DOI offer on the same network(payout $2.35)
Same GEO
B) Creatives testing strategy:Send at least 200 clicks per creative before selecting a winner

With a 0.05 per click this is $10 per creative per placement

C) Offer testing strategy: At least 200 clicks per offer before picking a winner

With a 0.25 per click it's $50 per offer

D) Test period: 1 full week

Of course the costs will depend on the CPM and CPC of the placements.

Started today will keep you posted on this

---

These payouts still seem a bit to low to me, what network are you with again?

This is a bit to much, its close to 10x offer payout, you don't need to run more than 3x in the initial tests. In most cases, the bad banners start to fall behind after 2-3x spend.

This will depend on how the various banners and LPs perform. Once you block some LP or banners, you need to disregard the data pretty much, and only focus on the data from the creatives that you keep on running.

This should be enough to get a good idea about the potential of this campaign.

Looking forward to your results.

---

Advidii

This is a great info thanks a lot for sharing it

Why do you say you have to disregard the clicks sent from the creatives you stop?
If this is true, then basically 90% of the initial traffic flowing trough the funnel is useless.

This would also force you to keep all the clicks strictly separated like:

Dedicated Campaign for Placement -> Dedicated LP -> Dedicated Offer/offers

I am using this approach here:

Dedicated campaign per traffic source -> LPs -> Offers

So basically the conversion can be traced back to the creatives and landers only digging deeper in the campaign stats.

What approach do you prefer to use?
__________________________________________________ __________________________________________________ _

SO NEW PLAN ACCORDING TO THE SUGGESTIONS IS THE FOLLOWING:

3x offer payout spend per creative per placement a day.

The test of the creative + placement is gonna be 3 to 7 days depending on the amount of conversions
I get from that particular combination. If after 3 days running 24/7 there's no conversions, I'll kill it.

Generally speaking, if the spend goes way beyond the payout, and a couple of
quick random conversions can't possibly get the placement back in the game,
I'll kill it.

I will be testing 5 different placements & 3 creatives each.

I will have different payouts, so I will average them out
to calculate the max spend per placement.
(is this the way to go?)

- Still confused on how many clicks to send to an offer before picking a winner -

I will also ask my am for offers with better payouts to test in this GEO, Both SOI and DOI

---

The traffic is not useless, but clicks coming from a banner that you paused and through LPs that you are no longer using ... it's not fully relevant anymore.

This traffic helped you to find the winners and pause the losers, but in order to optimize further, you need to focus on traffic delivered through creatives that you are running NOW.

You can still use the old data to spot trends and find the most profitable targets, but you shouldn't cut placements based on traffic from banners and LPs that you are no longer using, because the new banners might work way better on the same placements.

As you will find out later, even though you can track conversions back to ad and LP by digging deeper, when you isolate these detailed funnel and run just those, the result will suddenly become worse. That's how it works, especially in adult. The statistical error is simply bigger here.

I still do this from time to time, when the isolated funnels perform WAY better than the rest.

The approach I use the most, is to focus on what is working the worst and getting rid of it. I start with lots of banners and LPs and offers, and quickly cut the worst parts of the chain. It doesn't take to long in most cases and once I have cut enough to find myself with just the promising ones, I start focusing on the other parts of the optimization process.

This looks ok to me, once you start running traffic, you will see that in some cases its possible to cut sooner .. some ads are simply so bad that they fall behind fast.

Don't focus on clicks, focus on the spend and ROI.

Also keep one thing in mind, you need to decide this in 2 phases so to speak.

Phase 1 is getting rid of the worst part of the funnel, you can be pretty deep in the red here and it doesn't have to pose a problem.
Phase 2 is trying to optimize the campaign with just the better performing ads/LPs etc ... when deciding whether to cut an offer, only look at data collected in phase 2.

Advidi is a solid network with good dating offers most of the time, so talk to them and they might have something better for you. Quite often, they can also do a small bump straight away, when they see you are serious about working with them.

---

Hi xechel0nx,

It's been almost two weeks since your last confession! I'm interested to learn how you're getting along based on the above advice. I've enjoyed reading your thought processes and your approach, so thanks for thinking out aloud! I'm not currently that organised in my thoughts personally. I am looking to post my own follow along either today or tomorrow. I think I'll leave out my thought process in my post, as based on matuloo's response, at this stage in the process as my thoughts will likely be counter-intuitive, whereas I will likely need to have the data inform and align my thinking for me.

Thanks for your follow along and I look forward to any updates you may be posting soon.

Alan

---

I'd love to see an update too

---

Hey guys! So basically I've understood what I was doing wrong after banging my head against the wall for a couple weeks and spending a bunch of money.

Basically I didn't start from a proven offer but instead a bunch of offers that "looked good" and that's about it...the main lesson I've learned from this camp is: START FROM ROVEN OFFERS

Something you already see working and being actively promoted on the traffic source you want to target.

One more thing: I understood I didn't enjoy the research process on adult websites, and finding angles to persuade "the horny lonely guy", not there's something bad in it per se...it's just not my thing.

So I moved on, I decided to abandon this traffic source and went back to native

Currently promoting Health VSL offers.

One more thing I was doing wrong is making decisions about the campaign too quickly...
I was in a frenzy, touching ads, bids & everything all the time, another huge mistake.

I really learned a lot from this cap though & am very grateful to Matuloo and everybody who decided to follow and respond to this thread ;-)

Have a good one my brothers

---

This is a very frequent mistake that new affiliates are doing ... glad you could identify it and learn from it.

I understand that working with adult traffic isn't for everyone ... so good luck with native

---

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