How much do you spend to test a campaign?
- Thread starter illmill
- Start date
When you guys are talking about testing and spending $500-1000 before you change a lot, is it usually just on one ad platform before you try getting traffic elsewhere?
Say hypothetically you've got a particular combination of keyword, LP, offer, and traffic source, that overall is profitable.
Click cost is .50, payout is $20, conversion rate from ad click to purchase is 5%. Over the course of 1000 clicks from this keyword you've spent $500 and made $1000, 100% ROI.
If you look at your traffic, you've had 50 buyers out of 1000 visitors, or 1 buyer for every 20 visitors.
My question is, how clumped together or evenly distributed can these things be? Obviously you wouldn't expect a perfect "19 non-buyers then 1 buyer, 19 non-buyers then one buyer" pattern to happen. Also you wouldn't expect 950 non-buyers followed by 50 buyers. How distributed the buyers are seems to be important for how much needs to be spent to see if something is profitable.
Click cost is .50, payout is $20, conversion rate from ad click to purchase is 5%. Over the course of 1000 clicks from this keyword you've spent $500 and made $1000, 100% ROI.
If you look at your traffic, you've had 50 buyers out of 1000 visitors, or 1 buyer for every 20 visitors.
My question is, how clumped together or evenly distributed can these things be? Obviously you wouldn't expect a perfect "19 non-buyers then 1 buyer, 19 non-buyers then one buyer" pattern to happen. Also you wouldn't expect 950 non-buyers followed by 50 buyers. How distributed the buyers are seems to be important for how much needs to be spent to see if something is profitable.
^ This. Last 4 months I have been testing a niche with seo content only.
I wrote 30 articles and optimized them for high volume keywords.
Now I just look at the data to see which keywords/phrases converted and use this for my adwords campaign.
The campaign is profitable from the start. It takes more time to get the data but I think it's a good alternative instead of spending $xxx to test keywords on adwords.
Well suppose the main thing missing here is what is a test?
Is a test everything you try before you give up and move on to another offer/vertical?
Or is a test everything you spend to get a campaign profitable before you give up?
If number 1 is the case I think there's a case to be made for saying 2-3k if you try different presells, traffic sources, angles at selling it, etc.
With regards to truely testing a campaign that seemed like it might have some legs we've spent in the 100's of thousands testing. And that was just testing to see how we could get it to perform. Now that doesn't mean we lost 100's of thousands that means it was still at a loss while we were optimizing out and building it.
As I've said many times before ask a better question and get a better answer.
Is a test everything you try before you give up and move on to another offer/vertical?
Or is a test everything you spend to get a campaign profitable before you give up?
If number 1 is the case I think there's a case to be made for saying 2-3k if you try different presells, traffic sources, angles at selling it, etc.
With regards to truely testing a campaign that seemed like it might have some legs we've spent in the 100's of thousands testing. And that was just testing to see how we could get it to perform. Now that doesn't mean we lost 100's of thousands that means it was still at a loss while we were optimizing out and building it.
As I've said many times before ask a better question and get a better answer.
We usually do it the other way around. Spend a bunch of money finding the keywords then you spend the time to seo.
Suppose it goes back to the proverbial money or time thing.
Isn't success somewhat harder to judge with this method? It seems like a lot more variables would come into play because of the diversity of your traffic sources...instead of a targeted demographic from fb ads you're getting incoming clicks from SE results, social bookmarking links, who knows where else...I guess with this method you ignore your traffic source since you can't really split test?
IMO there are several pre-requisites for a successful campaign:
- traffic source that is scalable
- proper pre-qualification ads
- converting LP
- offer that performs with your LP
So, once I find a traffic source that can be scaled I'm pretty sure there's a way to monetize it - especially if I see other guys running stuff on it.
After that I make ads that I think will attract the right segment of people that will end up buying.
Now, finding the right LP is the big problem, since you can't really know right off the bat what your traffic really wants and what motivates them. This is where the most testing is done initially.
With each type of LP I test several offers. After I've found one that converts I'll start optimization of the given campaign.
But in case it doesn't convert that easily you need to ask yourself:
1. Is there a way to get the same amount of traffic at lower cost?
2. What kinds of different ads could I build to drive different segments of people to my LP?
3. How could I change my LP to motivate that segment of people?
4. Which offers could work better with this type of visitors?
With each step you can (hopefully) get several different answers... and different variables mean SEVERELY different performance on your bottom line, which means that ideally you'd want to test as many combinations as possible.
What I'm trying to say is that there is no way to give an estimate of how much one should spend on a campaign. It just comes down to how far are you prepared to go to make it work.
- traffic source that is scalable
- proper pre-qualification ads
- converting LP
- offer that performs with your LP
So, once I find a traffic source that can be scaled I'm pretty sure there's a way to monetize it - especially if I see other guys running stuff on it.
After that I make ads that I think will attract the right segment of people that will end up buying.
Now, finding the right LP is the big problem, since you can't really know right off the bat what your traffic really wants and what motivates them. This is where the most testing is done initially.
With each type of LP I test several offers. After I've found one that converts I'll start optimization of the given campaign.
But in case it doesn't convert that easily you need to ask yourself:
1. Is there a way to get the same amount of traffic at lower cost?
2. What kinds of different ads could I build to drive different segments of people to my LP?
3. How could I change my LP to motivate that segment of people?
4. Which offers could work better with this type of visitors?
With each step you can (hopefully) get several different answers... and different variables mean SEVERELY different performance on your bottom line, which means that ideally you'd want to test as many combinations as possible.
What I'm trying to say is that there is no way to give an estimate of how much one should spend on a campaign. It just comes down to how far are you prepared to go to make it work.
what about the decision to use an LP or not? i'm currently messing with an offer that pays great for a 1 page submit.. got a friggin great CTR on my ads but the shit just would not convert.. even some of the email traffic i got which was extremely targeted did not convert..
i wanted to take the route of not bouncing from offer to offer and trying to sit down and make something convert but if the advertiser's page just doesn't convert what can i do besides jump ship on the offer? (on this particular one there isn't a million choices of offers its not Acai or Biz opp etc.. )
i feel like a pre-sell LP could help but if I'm seeing a great CTR on my ad and this is really just a simple per lead offer.. is it even worth it? I know the answer is to just test it but just looking for a little input..
i wanted to take the route of not bouncing from offer to offer and trying to sit down and make something convert but if the advertiser's page just doesn't convert what can i do besides jump ship on the offer? (on this particular one there isn't a million choices of offers its not Acai or Biz opp etc.. )
i feel like a pre-sell LP could help but if I'm seeing a great CTR on my ad and this is really just a simple per lead offer.. is it even worth it? I know the answer is to just test it but just looking for a little input..
I'm a total newb and I haven't even tried any ppc yet so my answer is probably shit, but i have been reading a ton and I like to think i'm a somewhat intelligent guy soo..honestly it sounds like if you're getting a good ctr and everything else is great up to the offer, you've done everything you can and the offer must suck, right? Unless you can find a second similar offer to split test with it sounds like you should just ditch it and get on to the next one.
imp:
there are many different mentalities of internet surfers. just because your CTR is high, doesnt mean that they are the converting type..
- it could be the type of people you target with your adcopy,
- your LP doesnt inspire conversions
- the offer is shit
just troubleshoot each one of these at different times to use process of elimination
yeah im back on running the offer (of course now im having trouble with approval but just gotta wait for the interns to quit being a-holes)
im going to try some other ad copy and give it some more clicks after looking at my data i may have not given the offer enough time yet. if it doesnt work im gonna do a quick LP and if that doesn't work ill drop this mofo..
im going to try some other ad copy and give it some more clicks after looking at my data i may have not given the offer enough time yet. if it doesnt work im gonna do a quick LP and if that doesn't work ill drop this mofo..
for testing!
I want to spend as small as possible.. this is a test that is why..
I voted 50 to 100 only. :zzwhip:
I want to spend as small as possible.. this is a test that is why..
I voted 50 to 100 only. :zzwhip: