The keywords are “money button”

I think everyone who works with Search Arbitrage sooner or later comes to the point where they try to find a formula for selecting keywords with the highest RPCs 

The first thing that comes to mind is to use such services as Semrush, Google Keywords Planner, Ahrefs, etc., but agree, it would be too easy))))

In idea everything is very simple, we look for big Google publishers , we check what keywords they buy traffic for, we look where the highest RPC, High Competition, and that Volume would be as low as possible. According to this logic Google should pay us more for such clicks, because it does not have enough traffic on these keywords. But, in our practice it rarely works, although sometimes we can really find high RPC for such keywords, although you can just as well choose randomly - when you get lucky :).

By the way, RPV (Revenue per View) is more important for us than RPC (Revenue per Click), because the first shows the real picture of how much we earn. It is not always more profitable that the key, which has a higher payout, because click on it may be less often )

So still, how to search for keywords, and in general, is it necessary? After all, you can just take them from the spay, and do not worry (no).

In principle, the keyword test stage is the last one for us. We work according to the following algorithm:

1.We choose the offerer, sometimes take from recommendations , sometimes on the spay look where mass run advertising for a long period of time, or just personal ideas 

2.We look Creo by spay, analyze what approaches shed the longest, unique, adjust to our Compliance and run on the test. There are different methods to test creo, someone waits 2-3 days until the CPM falls, and someone makes a decision after a few $ spand, depending on what money-management :)

3.We see what the RPC on the basic key found on the spay, what CPA on the best creo and analyze in what direction to move us 

Of course initially we are trying to get to + ROI method of testing dozens of creatives to get a minimum CPA, and only when we realize that cheaper lead is not to make we start to go through keywords. But we do it only when we realize that we can raise ROI, and we see that the payout can really be higher).

Usually we take 3-4 packs of keywords and split 5 per domain. We do a split test, we shed at least 50 leads per domain and determine which ones have shown themselves better.

We keep the best 1-2 domains and run more volume, see how the keys behave when scaling and how much volume they can withstand. If the RPC is ok, we keep running, if it drops, we create a new domain with new keys. It often happens that on a couple of hundred conversions you are in the plus, but at scaling RPC drops and you do not know what to do ).

I know that many people ask themselves the question “Why does the RPC drop?”).

On this question you even managers from the affiliate will not give you a clear answer, but according to my observations most often such reasons:

1.Not enough advertisers buys traffic on these keywords, and you with your proposal to kill demand, the result is the price falls. Then just create a new domain with new keys and hope that here you are lucky )

2.Quality of traffic . Yes, yes, Google analyzes traffic, and if your leads do not make the necessary actions, after a day your payment will regularly decrease. We have tested this more than once , and indeed better traffic = higher RPC . This creo with massage from depression may convert, but your RPC will be low))Make more targeted creo - and the RPC will be higher).

But sometimes the payout for the same keys can be different on different domains, you will say ! )) 

Yes, and why this happens is a mystery to me, but we solve it by running the same keys on many domains and see which one has a better payout). 

By the way, the important point is that we do not change the keys on an existing domain - usually the RPC will not change much. 

New keys = new domain !) so the picture will be clearer)

As a result, we have not found the button "money" (shit) and do not have a clear strategy where to take the best keywords, yes we have a database of thousands of keywords on our team and we understand which best show themselves - but this is only theory, always need to test that to understand )

That's why we have a staff member who regularly tests dozens of keywords every day, and builds analytics for our media baers).

Well, if someone from you still has a cousin/brother-in-law who works in Google, and he will tell you the insights - write to me without fail, we'll give him a cup of coffee - and he will tell us where is this magic button))

If you have any questions, or wishes, write me in a lis, let's talk http://t.me/searchmaks.