Home > The Newbie Zone >

---

My results after few rounds (+5) (8)

---

Hi @vortex and all the other amazing newbies. ��

So as you know I was running this tutorial for a while but without success.
Than something happened I need a help with to understand.

I set the report from PopAds as you suggested but today, I added a variable: instead of leaving date range: this month. I choose "today". Because i was running this offer for two days so I wanted just the data from today.


My bet was the Geo Iran. So the data from the whole month were mixing everything and showing all negative ROI.

Filtering for Today...the situation changed:


I have some mixed feelings because I'm very tired and not sure how to read these data but am I wrong or I see a lot of positive ROI? ✌️
I mean like a lot!


I'm not sure also if I found a positive ROI or not becase:

1) considering all two days and the overall budget spent, I'm in negative


2) considering only the last round of this campaign, it's like super awesome!

Can you Amy help to read this properly?
Which should be the next step?

---

Two major factors are at play here:

1)Pop traffic can be very volatile - you can be losing money one day and be profitable the next.

2)The "today" stats have such low spend - we're trying to make observations based on a few cents here - which isn't ideal.


So in another thread I've already suggested to take the best-performing geos and set up separate campaigns for each. Another thing you can do is just to take your initial campaign where you're targeting multiple geos, and cut geos that are performing badly. And then look at placement stats and cut placements that seem to be performing badly across multiple geos.

You can do BOTH - keep this initial campaign alive, and set up separate ones at the same time. Often, when you're targeting more broadly - for example multiple geos in the same campaign - you can get more traffic volume at lower cost. That's how the algo works. Therefore, sometime, you'll find that when you take individual best-performing geos and start their own campaign, you'd get worse performance than when those geos were in the original multi-geo campaign. So testing both arrangements would be best.

This type of testing is fun, because you can do all kinds of testing and collect data very quickly on low spend - because of the extremely low payouts.



Amy

---

A quick update, I think I did it! Kudos to Amy @vortex
Small, but still profit ✌️

I was not looking for the profit, I was willing to learn and your tutorial helped a lot!

---

Thanks as usual Amy. What you mean by "placements" ?
Ok, found what you mean:
https://stmforum.com/forum/showthrea...ing-Placements

---

Hell yeah! I understood the placements thing so I started to exclude zones which are not profitable.

If I understood correctly the optimizing process:

1) we run a lot of geo
2) we cut of the geos with very low ROI
3) we choose a winner GEO or GEOs (the ones with the best ROI)
4) we run only these Geos
5) we cut placements by cutting all what is below 1% ROI

In theory now we have only profitable Placements so we should have higher ROI.


Question: the next time I will run a campaign in this Geo, how can I use these data to perform better on that campaign? Can I or should I start always from zero or can I leverage the data from previous Campaigns?

---

That's pretty much it! Except the 1% ROI - you got to look at the actual spend of the placement as well. For example if the payout is $1 and the placements has only spent $0.30 without making a conversion (yet!), the ROI would be -100%, but you probably won't want to cut it yet.

It's not really possible to only have profitable placement - I wished this was achievable also. Truth is that there will always be ultra small placements that only give you a few impressions here and there, that you can't really optimize - I talk about that in the lesson on how to cut placements (which you linked to above).

And yes, the idea is to leverage existing data, so that the next time you run in the same geo on the same traffic source, you can exclude at least the worst placements from the start to save some money.



Amy

---

Oh ok so the ID's of the placements are always the same? Should I save the worst on an Excel file?

Can U tell how to see the PayOut?

---

Yes the ID of each placement should always be the same. A "placement" for pop traffic is simply the website that shows your ad.

You can save placements any way you like.

Payout is just the revenue per conversion - you can find this at the affiliate network. Or just divide the "Conv. value" by "Conversions".



Amy

---

Home > The Newbie Zone >