Home > Paid Traffic Sources > Mobile

---

Eliminating Placements For Mobile Pops (9)

---

I have a question on how to approach cutting non-performing placements. My idea was to create a system with two rules:

Low Paying Offers ($2+): Cut when its net loss is $2 or receives 300 clicks and no conversions
High Paying Offers ($8+): Cut when net loss is $4 or receives 300 clicks and no conversions

Is this the right way to approach optimizing placements?

Thanks!

---

So you'd cut an $8 placement after it spent $4 without converting? Sounds like you're requiring a 100% pre-optimization ROI. It happens but this isn't really realistic to expect.

The standardly preached tactic is to spend up to 3x the offer payout before making any cuts. So cut a $2 offer after it spends $6 without conversions, $8 after spending $24, etc. This is more of a guide and it really comes down to personal feel but I think your current strategy definitely has you cutting stuff way too early.

No need to make any cuts based on clicks at all - higher payout offers can take more clicks before converting and be fine.

---

Yes just as dusklife said….

If you cut off a placement before it has even reached the payout, it still has a chance to become profitable and you’re therefore losing a potential placement.
Also regarding the clicks - if you cull depending on clicks you’re assuming the bid is always the same… For one geo 300 clicks could be the same price as 1000 clicks in another geo.

---

you need to base it on "average conversion rate" for you base comparison.

I'd talk with your affiliate manager and find out what the offer is converting at on average. Then put that into split test calculator and find out how much data you'd need for relevancy to determine if the placement is bad or not. Calculation of things should never be based on money but rather based on conversion rate %. This is sample size (clicks) and conversions.

---

I am running on RON, and there a ton more placements so i was wondering if that makes a difference?

Thanks for the comments thus far.

---

Interesting point. And I agree to the statistical significance. However by optimising according to whether you can achieve the same conversion rate as the average - aren’t you missing the fact that perhaps your traffic is a lot cheaper (pops vs banners etc) and you can therefore afford to run at a lower conversion rate and still profit. Thereby cutting potential profitable placements just because they don’t match the network conversion rate?

No this doesn’t make a difference - you still need to test each placement to give each a fair shot. With RON it’s all about pausing the crap and uncovering the golden nuggets buried in there!

---

I cut pretty aggressively with pop and redirect just because there is a ton of fraud/garbage placements out there (especially on plugrush). A lot of the garbage placements can be cut pretty quick based on CTR. For example if the placement has had over 100 visits and has zero or under 2 click throughs I'll just cut it right away. There is just too much crappy zones out there to optimize based solely on the 1-2x offer payout threshold people like to use.

---

clubrock, do you cut Unique ID for the web sites or IP addresses for the multiple visits and 0 clicks or do you cut both Unique ID and IP? thanks,

KK

---

Maybe this will help :

http://stmforum.com/forum/showthread...Banners-Part-2


Amy

---

Home > Paid Traffic Sources > Mobile