Transformation from software developer to marketer (10)
Hello,
I'm a 24 y.o. software dev, but realized writing code is just not where the money is! (well, at least the big money) You can either get paid by the hour and your only way to increase your income is by begging your boss to give you a raise (basically). Or you can start a start-up and work on a super-cool application for 6 months and then it turns out the market doesn't really need it! AM is the exact opposite in that sense - everything starts from the market and by testing we determine what to sell. The feedback loop is very short. I want to feel the "pulse" of the end-consumer.
I'm starting out with mobile pops as it seems like the easiest traffic source where I can learn the basic workflow/routine of AM.
My current traffic sources: Popads, Popcash
Current networks: Adsimilis, F5Media, Clickdealer
Tracker:
Spy tool: AdPlexity Mobile
Hosting: AWS S3 + Cloudfront + Route53
I ran my first test today, here is the info:
Offer: AV pin submit
Traffic source: Popads
Payout: 0.22 USD
Max bid: 0.0008
GEO: Indonesia
Targeting: Indonesia, Android smartphones only, WIFI and 3G
Result:
Questions:
1. I set my total budget for the campaign on Popads to 10 USD. It took about 15 minutes for it to spend it all.
a) The average bid in "Inventory" of Popads for Indonesia was 0.00075
b) I added +20% to it as I read some reccomendations on the forum that said to do that.
c) When I saw how fast the budget was going down I lowered the bid to 0.0008, but it kept draining as fast.
Is this normal or was I bidding too high? How long does it normally take for such a test to run after being approved by the traffic source?
2. Considering it took me quite a lot of time to set up the ripped lander on new domain + Cloudfront CDN + S3 - would it make sense to use the same domain for mass testing offers by just creating one folder per campaign on the same S3 bucket?
3. Why are my costs & ROI not showing up in
4. Testing budget
For test budgets I've seen reccommendations such as:
avg payout X offers X landers X 3
Which in my case would translate to:
0.22 x 1 x 1 x 3 = 0.66 USD
So can it be said that I spent way too much cash on this test?
Thanks for reading!
Indonesia is a big GEO with lots of traffic, without a blacklist you will spend your money quickly.
Most of it will be traffic that nobody else wants.
Either hit up a traffic source rep for a basic blacklist or try to get one from other affiliates your friendly with.
Set the cost tracking to CPC and use the same CPC you have set on the traffic source.
However there will always be clickloss, so you need transfer the real costs from the traffic source to
Originally Posted by sebastian_r
I have emailed my TS regarding blacklist, but I'm not counting on their help too much. Instead I launched bot tests in both Popcash and Popads using caurmen's guide (https://stmforum.com/forum/showthrea...highlight=test). Will see how that goes.
Getting a blacklist with help a bit but you will still need to spend a fair amount to make this GEO work well.
Here is what I suggest you do...
You're running an AV Pin, so for now, completely remove the Wifi traffic because it doesn't usually convert as well and there is a ton of wifi traffic.
Split your campaigns up, break in down to one campaign per mobile carrier. By doing this PopAds will give you an estimated bid price that is more accurate to each carrier plus you are able to allocate a more equal budget to each carrier.
In the beginning try setting your bid to 25% less than the average bid that PopAds estimates for you, if that gives you very little traffic, then I would increase the bid to 25% above the average bid. Run a test, you will need to spend $10 or so per carrier just to see if there is a carrier that will convert. I highly recommend trying to find a few offers to split test and 2-3 landers to begin with should be enough.
This is just to see which carrier/offer/lander combination has potential, if any at all. If one or more segments stand out, then you need to continue testing as many offers as you can find. If you are getting some conversions but it's just so few that it doesn't make sense to continue, I would test Zeropark.
If you find something profitable, then later go back and test the wifi but bid at a lower CPM because it will convert less.
Another big tip - your CTR is extremely low. Before relaunching any campaigns again, make sure there are some high volume placements that have a decent CTR as in above 5%. You can blacklist from the beginning any placements that are show less than 2% CTR and over 100-200 visits. But if overall your placements are lower than 5% CTR, then you should test more landers.
Andrew
Thanks Mr. Payne!
Yesterday I:
1) Launched 3 separate campaigns with the initial lander, one per carrier
2) Tested 2 more (ripped) landers
3) In all cases excluded WIFI traffic and targeted only the carriers that the offer supports
Results:
Analysis:
1) Lander 2 converts, and I'm not surprised - it's very agressive compared to the other 2.
2) All the Lander 2 conversions are from one carrier. But that carrier's campaign brought 2x more visits than the other 2.
3) For Lander 2, I also accidentally had prefetch on in Popads, so the visits/clicks stats are messed up.
4) I did not block any placements yet
5) I also got a conversion from Caurmen's bot test, which was essentially direct-linking to the offer
6) From my bot test I didn't really find many placements to cut (no significant placements with 70-80% or more bot traffic)
Now I plan to:
1) Cut placements according to Mr. Payne-s suggestions: less than 2% CTR and 100+ visits
2) Test Lander 2 with each carrier (using the new blacklist)
3) Test more offers in this GEO
I found a sweeps offer that many people have been running for a long time in ID. Will test 3 landers + direct linking too.
Originally Posted by monkeybusiness
Keep in mind most of your landers have low CTR but the one that converted has a high CTR.
I would run more traffic to the one with high CTR only.. then after a bit of spend I would look at the placements for traffic just through that lander and then cut by the CTR method I mentioned.
Andrew
Originally Posted by Mr Payne
Good luck and keep this follow along going!
my suggestion is that if you chose this traffic source, do what you can to master the f**k out of it
in regards to offer, landers and prices etc...
that's what I see many successful people do, and saw it worked for me too
Home > Mobile > Follow-along Campaigns