I have been using ROAS bidding strategy with excellent results. I am talking about campaigns with lots of conversion. Around 80 conversions in last 30 days.
I tested many different TARGET ROAS like 100%, 150%, 200%, etc.
Of course, every time I set a higher TARGET ROAS, the quantity of traffic is less but the ROI is higher.
The question is, how much TARGET ROAS do I should set to get BEST earnings?
As an extreme example, 10% of TARGET ROAS of a BIG amount of campaign invest could be HIGER than 500% of a small campaign investment.
Actually, I have set a ROAS of 200%, but I really want to get the most earning of the campaigns investment. How can I know how much to set the TARGET ROAS in order te get most earnings as I can? Is there any way that Adwords knows the BEST value TARGET ROAS? Is there any balance tool?
Re: ROAS S.O.S[ Edited ]
November 2015 - last edited November 2015
Wow, great question @Isadora R (Mat).
I think the only way you're going to be able to do this is to test and plot your results on a graph to look for the peak actual income. You're in a good position in that you've got lots of data coming in so it shouldn't take too long to gather significance at enough ROAS points to spot a trend. You may, of course, be able to use some mathematical modelling to predict a full curve with only a limited number of data points, though I suspect AdWords won't suit a very rigid model, so the more points you have, the better.
How I'd probably approach this would be to start with the lowest ROAS that gives you a decent profit, run this for N days then increase by 10% and repeat. Plot at least ROAS target, actual spend and net income on a graph and look for the peak in net income. I'd expect to see some form of bell curve so you may be able to stop the experiment once you see clear evidence of the curve "heading downhill". This should work but it has a couple of drawbacks.
#1 - All the conversion optimisers take time to get to their best performance so if you're changing the ROAS target you're really going to need to leave that target set for at least a week, maybe two, before you can be certain it's working as well as it can be. If you're only increasing ROAS by 10% a time it could be 2017 before you plot the graph!! So give some thought to what's a reasonable time period for each level and what sort of increase in steps you should use.
#2 - You run the risk of hurting potential income while testing. In this case I'd probably be inclined to run a Campaign Experiment where you have your Control Campaign set at an ROAS target that you currently believe is the best income generator, then use the Experiment Campaign for the testing. This should give you a steady baseline income to smooth any bumps in the testing.
Re: ROAS S.O.S
If you have any other idea how to get the BEST target Value sooner, please let me know.
Re: ROAS S.O.S
Hi @Isadora R (Mat) thank you for an interesting question, it's made me think about a couple of my own clients...
I don't think there's going to be any quicker way. Even if I could share my own results with you (when I have them!) they could be entirely wrong for you because every business is going to be different. I'd happily share some script snippets with you that I'm using - I'm testing by referring to a spreadsheet that defines the ROAS value for a test Campaign based upon the week of the year. I'm going to increase the ROAS gradually, week on week, then lower it again over the same period. Hopefully I'll see some significance in the data I collect, but it's going to take at least 3 months to do this.