Am I Quitting Too Ealry? (11)
I've been promoting an offer in the beauty niche. It has a pretty good video sales page and the network EPC looks pretty decent.
I figure that since the video sales page does a decent job of selling the product, I could just create a simple pop-up page that links to the offer.
Just to be clear, what I did was create about 20 variations of pop-up pages that had headlines along the lines of:
"Beauty Secrets Revealed" or "How To Get Rid Of Wrinkles".
Those headlines, an image, and a CTA.
I've sent about 150+ clicks to the offer so far with no conversions.
Based on the network EPC, I should have at least 2 conversions by now. I started out paying over $1 per click, but I've optimized a bit and gotten that down to about $80 cents a click.
But since I have no conversions after spending $200+, I paused the campaign and stopped optimizing.
Do you think I should continue pumping money into this campaign?
My traffic source is trafficvance if it makes a difference...
I can see your logic - I used to think the same way as well, e.g. "wow this offer has a high EPC of 0.50! So if I can keep the cost of every click to the offer page below 0.50 I'll make money!" Unfortunately that's often not how it works. Here are several reasons why:
1)Some affiliates may be getting higher payouts than you are, which will effectively bump their EPC as well.
2)They may be running traffic that is of higher quality and/or more targetted than yours - e.g. Facebook / Adwords / Email Lists.
3)They may be preselling visitors a lot better than you in their banners and/or landers. It's one thing to get people to click through to the offer - but whether they'll sign-up is another matter. Adequate preselling can do wonders to conversion rates (not to be confused with clickthrough rates!)
From the sound of it, the offer you were testing is a high payout offer. Those will typically require a much larger budget to optimize, especially if you don't have prior experience with the particular niche. If you DO have the budget though and would be willing to invest money and time into cracking this niche, rewards can be HUGE. However, even in that case, I would suggest testing more than one offer from the start, and stick with offers that you know are converting decent/well (you can find that out by talking to AMs that have given you solid advice in the past, or from talking to your affiliate friends). This way you minimize your chances of wasting huge amounts of money on testing landers (etc.) using a dud offer.
Amy
You made a good choise. If an offer doesnt convert at all after having spend +100$ then it won't work. Move on the next one.
Good to know is that optimization only works if you are in the red but close to being in the green. There have to be conversions for you to be able to optimize (i.e. cut bad placements, push good ones)
Originally Posted by cbrughmans
I would say, see which variation performed better in terms of CTR or cpc maybe, and test some more on those.
Absolutes in this game seems like a sure fire way to filter out a lot of possible good information.
@fabian
Thanks for the feedback. We agree to disagree.
Hangman spend 200$ and had no conversions. My question to you is how can he optimize the campaing on eCPC if there are no conversions? How would you optimize the campaign?
Everyone is on limited funds - even the big shots - so after 200$ and no conversions, I would move on to the next campaign because this one seems a lost cause
I do agree with your point of the 20 variations, if you spread the 200$ over that evenly then it won't give you enough data. I'd never start off with so many variations, max 2 or 3
Well to clarify, I didn't start off with 20 variations. I actually started with about 3. Then 3 led to 7 then 7 to 15 and so on... At the end of my testing, I had been through about 20 variations.
I know my testing was not statistically significant but to give an idea of what I did, when I started with 3 landers, there may have been 1 that really stood out after 200 pops. For this, I would create a few variations.
Throughout my testing, I gave each lander a minimum of 200 pops, then created variations of the best to try to compete.
If any landers had a drastically low CTR after the 200 pops, I would cut it. (Do you think this was waay too early?)
The problem though is that I'm optimizing based on CTR and not CVR. So that might be a reason for my failure there.
Second problem is that i didn't use a flog like traditionally used with these kinds of offers. But based on it being a VSL, I thought I could get away with it.
Originally Posted by cbrughmans
I think you made a decision to kill it too early. I'm not saying I would have let it keep spending, but that's only because I think the approach is a bit flawed, and the angles are a bit too tame and don't provoke enough emotional response. Think more about solution to pain points rather then informational. Let the VSL give the information.
Some thoughts on your testing:
How quick did you spend that money? Was it all from one place? Was it all in one day? Did you give any variation more then even 1x payout?
There's a lot of factors that contribute to converting a VSL... demo (age, gender, affluence)/time of day/day of week/angle of promotion/type of traffic/your ad placement on source/etc...
Start off trying to prove one thing right or wrong... to me it sounds like you tried to test too many things at once here and didn't give them all an honest test.
Originally Posted by Adamw
No single landing page variation got more than 1x payout but there were groups of landers with minor changes that did.
Thank you for your response. This is very helpful...
Lots of great tips from everyone!
When first entering a vertical - it would be a good idea to rip and use landers you see the most often when you're spying, in order to establish some sort of benchmark. If everyone's using a certain type of lander (flogs in your case) and you try a different type, there's no telling whether it's the offer or your lander that's causing the lack of conversions.
Also, using a proven traffic source, i.e. one where you've spied on and have seen your type of offer being promoted, would be a good idea as well. Although the fact that someone is promoting that offer type there doesn't automatically mean they're making money, but if you see the offer type being promoted there a lot, that would be a positive sign indeed. Matching up the right offer type with the right traffic type is essential - get this part wrong and no amounts of optimization can help.
Summary: When first entering a vertical, stick to proven offers, run traffic from sources that look like they're working for that offer type, and use landers you see used most often during spying. These 3 will help to establish a benchmark or "test control" that you can then proceed to beat again and again to optimize your campaign.
Optimizing based on CTR often won't help either, because high CTR doesn't necessarily correlate with high CR (depending on your angle and headline etc.)
Hope that helps! 
Amy
Hello Hangman,
It feels great to see those great ideas in one thread. Let me give you my share of thoughts as well 
When I make a test, I only consider that the campaign has enough data when the costs equal 10-20 times the average payout of the offer. Otherwise, I don't consider the data as solid.
Regarding the variations, I only use 3 - 5 variations at once. I test them, take out the worst performing ones and test new ones. Using only this amount guarantees that I have solid data in each one of them quicker. So I don't need to wait too long to optimize my variations.
Now, speaking about the network EPC being higher than the one you actually have. As @Vortex said, you shouldn't totally rely on this because the traffic that the whole network receives is not the same as yours. You can have an idea of the performance of a specific offer but never expect to have the same values. On my side, when I see a network's eCPM for example, I know that this value comes from Mediabuyers, Webmasters, Remnant traffic, Social Traffic, etc. And as a Mediabuyer, my traffic will normally have a better quality.
With all those ideas from all of us, I think you have what it takes to make more money
Cheers!
So based on all the great tips everyone has shared, I've decided to continue running this campaign a little longer.
I've added a few new landers (some I've found and some I created).
My CTRs with these new landers seem to have gotten worse though. All my landers have a 1% - 2% CTR. I'm thinking they should probably be a minimum og 5% - 10%?
With all the different styles/headlines and images I've been testing, I'm kinda out of ideas as far as landers go.
So I'm thinking the problem must be with my targets.
I've scraped targets from all over. I currently have about 1500. But only a few are getting impressions.
So now I'm thinking that my approach with targets may be my problem ATM. I'm bidding 0.016 on all my targets. So now that I think about it, only the targets that have very little competition are getting impressions.
Now I'm thinking that maybe a better approach would have been to just pick a few targets, (maybe 5-10?), and test those first. That way I could pick targets that are closely related to my niche. I think that would give me much cleaner/more accurate data.
Anyone agree?
Home > Paid Traffic Sources > POP / PPV / Redirect