Follow Along Let's Shotgun Some Push Conversions

[DEADZ]

I made $50 profit yesterday off offers that pay $0.03-0.07 yesterday. Many members here have a goal of making $50/day

#facts

Thanks for sharing!

 

[Nick]

Yep

lol I wish I could read.

 

[Voluum]

Yeah, I see myself still running most of my Zeropark campaigns through Voluum, but I'm happy they're giving everyone the option

Zeropark + Voluum is the best duo as the integration is simply on another level - it's faster because the data points for the visit (UA, click ID, tokens) are passed in the back end. Meaning that there is no intermediate redirect through the tracker and the user is sent directly to the offer. So if you buy Pop or domain traffic from ZP and track with Voluum, it is a hair faster than for other solutions. Does not happen for push yet, but will in the future.

 

[Luke]

Alright, my $20 is spent and we have some winners and some losers:

I don't want to focus on GEOs with no traffic. I think I can optimize the targets on the ones with traffic to get something worthwhile. So, to start, I'm going to setup campaigns for Spain and Poland.

I have some data on Sources and Targets now too:

There's not much here though besides knowing that my target CPA ($0.05) is going to be tough. I have a nice conversion rate and my eCPA on both the Source and Target level are what you would dream of...assuming the offer didn't pay out $0.05 lol

Not sure I'll be able to get this profitable, but we'll give it a shot with Spain and Poland

 

[Luke]

I had a Spain campaign setup already so I just copied that over for Poland (love the copy campaign feature on Zeropark). I realized I didn't pause browsers besides Chrome in the Spain campaign so I wasted $3 in that. Not a huge deal, but when you're focused on a $0.05 payout, every penny counts.

I'll set a budget of $10 for both of the new GEOs with the same RBOs setup as the original main campaign.

 

[Luke]

One question people have a lot of times is how do you know when you have a good bid on Zeropark. Well, here is an example of bidding too high:

On the target level, a good starting bid should get you about 10-20% of the traffic (based on my previous experiences). If you are getting 50% or higher (prior to optimization) then you are probably bidding too high. I had my bid set at $0.003 and it seems that was too high. Dropped it down to $0.002. Might be able to get away with even dropping it to $0.001.

 

[infotech]

I use Cloudflare for affLIFT, my blog, and a bunch of other sites.

For pop traffic, Aaron recommended not using it because it can slow down the click speed. He gave me a bunch of technical reasons which I am not sure I fully understand yet, but recommended I use the default domain he had already provided instead for this test. I trust his judgement on it so I went with it.

@Kintura maybe you should make a thread explaining why it’s better to not use Cloudflare with more detail?

Do you use default domains on offer links or custom domain names?

 

[Luke]

Do you use default domains on offer links or custom domain names?

I've been just using the default tracking domain from the affiliate network(s) and a custom domain on my tracker.

 

[Luke]

I decided to kill both of these campaigns. I'm not seeing much on Poland or Spain as far as traffic goes and the conversion rate is too low for it to be profitable.

Since I'm killing these, I'll show another campaign for Italy for a different offer (domain redirects from Zeropark just like the original campaigns for this follow along).

Here is a Google Sheet with the top targets. Low payout, but this one actually seems like it might have potential

EPC is calculated by the payout/spend and optimal is 80% of that (20% margin to start). Not sure if I'll get this one green before I pause because it'll be a lot of work optimizing these bids, but overall is could be a small profit producer

 

[Kintura]

Don't get me wrong. CloudFlare is awesome but it's application-specific and not toward the application of high performance traffic distribution for tracking purposes. You should most certainly use it for everything that isn't tracking clicks. CloudFlare is just a reverse-proxy. Kintura is essentially a CloudFlare-style architecture as it is -- we already have a global network of reverse-proxies capable of handling 70,000 requests per second. You're just putting another one in front of it but also with a second SSL "handshake" -- under normal conditions this handshake takes 8-15ms...with CloudFlare it sort of depends on which load balancer your user lands on. That's because CloudFlare is serving everything from video streams to...video streams.

The process of adding a domain to Kintura is pretty much identical to adding it to CloudFlare. The SSL is automated here, too. A lot of cloud trackers rely on CNAME records for custom domains. In certain geos, especially outside the US, you can see your users hitting 1200ms delays on those. Pretty sure only Kintura and Thrive do direct domains. Not sure if Thrive's are self-service like ours, though.

Also: CloudFlare might be very helpful on your landing pages. Separate issue.

According to the custom tracking domains I see in our system, they are best created by making a fist and softly hitting your keyboard a few times...followed by a carefully typed ".info" #protip

 

[Nick]

According to the custom tracking domains I see in our system, they are best created by making a fist and softly hitting your keyboard a few times...followed by a carefully typed ".info" #protip

Are .info domains faster? Can you elaborate?

 

[Kintura]

Are .info domains faster? Can you elaborate?

It was a joke, sorry about that. I just mean that the domains seem completely random. I don't know of any advantage to using one TLD over another in terms of DNS performance.

 

[Nick]

It was a joke, sorry about that. I just mean that the domains seem completely random. I don't know of any advantage to using one TLD over another in terms of DNS performance.

ROFL. I was like WTF how is a different TLD faster?

 

[Luke]

According to the custom tracking domains I see in our system, they are best created by making a fist and softly hitting your keyboard a few times...followed by a carefully typed ".info" #protip

lol, I've seen some insane 342185ueotijalka .info tracking domains. At one point I had ppv.click, which I was pretty excited about, but I let it expire since everyone knew I had it

 

[Morty]

Ah, that top target caught my eye so I added another RBO:

View attachment 3744

Don't want to waste a bunch of my small budget for this initial test on targets that get a conversion, but are going to likely still be ROI negative.

Are these RBO screens from Kintura or ZeroPark?

 

[afiq]

Are these RBO screens from Kintura or ZeroPark?

This is from Zeropark.

 

[Luke]

Are these RBO screens from Kintura or ZeroPark?

Check out these articles:

• https://zeropark.com/docs/zeropark-features/rule-based-optimization/
• https://zeropark.com/blog/case-study-rule-based-optimization/

 

[Morty]

This is from Zeropark.

Check out these articles:

• https://zeropark.com/docs/zeropark-features/rule-based-optimization/
• https://zeropark.com/blog/case-study-rule-based-optimization/

Nice. Thanks to the both of you.

 

[Luke]

I decided to try this same strategy with the push notification smart link from Mobipium. This is working much better

I am still optimizing and just setup the two target campaigns. Nothing really to get too excited about, but it's always fun to see conversions and play with data.

BTW, the more I use the Kintura reports, the more I like them!

Obviously, I am very partial to the rocket

Nice work on that @Kintura

 

[Luke]

I have no idea why I'm so obsessed with trying to get these $0.05 payout campaigns profitable but I think I should probably stop lol. The potential profit is probably too low to keep wasting my time with it.

That being said, I do think this is a good practice for someone wanting to learn how to track and optimize a campaign since you get a lot of data for very cheap. Maybe I'll make a short "course" for affLIFT members.

What do you think?