Converts but Red ROI

[TrafficYoda]

Im a little stuck with what to do when I have converting pubs but it's RED ROI.
I did block -90% ROI pubs but what to do with those converting but not enough.

What's the best approach for this type of problem?

 

[Graybeard]

Better ad quality -- maybe. This assumes the offer is viable <<<a good value that will scale
Better landing page -- maybe. This assumes the offer is viable <<<a good value that will scale

Is it a repeat sale or a one shot deal?

 

[KwarkAds]

While everyone has their own specific method of cutting placements, the goal is to remove those that:

a) do not generate any revenue
b) persistently cost more than the generated revenue

Assuming most bot traffic has been blacklisted and with sufficient data stats, I generally follow the following for cutting placements further:

1. if cost is same as 1x payout without conversion, then its removed. Why pay same amount if there is NO revenue.

2. if cost is same as 2x payout with conversion BUT negative ROI, then its removed. Why pay twice but revenue is a persistent loss instead of profit. Hence, i cut placement before cost is x3 x4 x5 etc....with negative ROI.

3. if ROI is positive, I separate into white listed campaign. Those in turn are also monitored to remove fluke placements that may have converted luckily or traffic dried out. Similar to #2 above.

I hope that gives you an idea to craft your own approach for similar situations.

My 2 Cents ;-D

 

[Patrick3333]

Im a little stuck with what to do when I have converting pubs but it's RED ROI.
I did block -90% ROI pubs but what to do with those converting but not enough.

What's the best approach for this type of problem?

I guess the only thing that could really help will be to run some more tests, and see what goes on, also you could try to check your traffic, it will be very helpful TrafficYoda!!

 

[Mobidea]

Negative ROI is a normal thing in the beginning. If the offer converts, there is a chance to make it green.
Try to go to deeper optimization and cut the cost. Maybe cut the placements or sources. Or vice versa, try to find the best and increase the bid there. There are many different options, so try to play with it around!

 

[TrafficYoda]

While everyone has their own specific method of cutting placements, the goal is to remove those that:

a) do not generate any revenue
b) persistently cost more than the generated revenue

Assuming most bot traffic has been blacklisted and with sufficient data stats, I generally follow the following for cutting placements further:

1. if cost is same as 1x payout without conversion, then its removed. Why pay same amount if there is NO revenue.

2. if cost is same as 2x payout with conversion BUT negative ROI, then its removed. Why pay twice but revenue is a persistent loss instead of profit. Hence, i cut placement before cost is x3 x4 x5 etc....with negative ROI.

3. if ROI is positive, I separate into white listed campaign. Those in turn are also monitored to remove fluke placements that may have converted luckily or traffic dried out. Similar to #2 above.

I hope that gives you an idea to craft your own approach for similar situations.

My 2 Cents ;-D

I understand that i can do this for publishers
How do you determine to blacklist Os version, browser version and cities

 

[KwarkAds]

Using your tracker and campaign stats, just group them by those category (OS, OS version, Browser, etc...)
Then look for patterns of conversion. Analyze and conclude which KPI (device, os, etc) are converting.
You can then exclude or include targeting of such on traffic source.

How long has your offer been -90%? Did you chose that from a group of tested offers?
If you have already bot filtered placements, and done 2-3 separate optimization on placements and its still
-90% ROI, I would drop the offer and skip further optimization (LP, angle, etc). It may just not be converting enough to be profitable. For me, i would work on offers ONLY if they are at least -50% ROI or better. Higher chance (but not guaranteed) of profitability and postive ROI.