Is this a smart testing strategy for Pops on Zeropark? (12)
Hi all, I’m new to the game and although I’ve played around with Adcash, I have decided to try and crack ZP namely due to it’s easy to use and Adcash’s min €50 daily budget is a little out of my budget. I know this means ZP is a more saturated source but I feel like it’s an excellent place to really learn and nail down a strategy for moving forward...
I’ve used a combination of Caurmen’s incredible resources on the mobile cookbook and Rubytunes’ Pops/Redirects Guide to devise a testing strategy for your comment. Please let me know your thoughts below… I then hope to modify it and re-post it below.
Traffic Source: ZeroPark
Traffic type: Pop-up (To date I’ve also really struggled to get many conversions on redirect campaigns, pops seems more affordable and has proven to be more successful for my tests so far)
Offers: Choose 2 offers with a max payout of $5 (as I have a low budget).
Angles + Landers: Brainstorm 21 angles, select three and create a landing page for each.
STAGE 01 - Campaign set-up:
Campaign A_Test:
Set 2x offers to rotate.
Set 3x landing pages and direct linking in even rotation.
Bidding: Check ZP’s volume for Pops, bid just below the average.
Daily Budget: 5x offer payout
Run untouched for 2 days unless…
Pause placements that have spent 2x the payout and is not performing better than -50% ROI OR is taking up more than 50% of total impressions. (Will be temporarily paused until Stage 02 if there is some promise for optimization.)
Wait for campaign to spend 10x offer payout (2 days).
IF no profitable or promising placements after 10x offer payout kill the campaign and try some new offers.
STAGE 02 - Campaign whitelist & Optimization
Pull out the profitable placements: Move profitable targets into a whitelist campaign, continue the test for more placements with poor performing placements paused in original campaign.
So now I have Campaign A_Test and Campaign A_Whitelist.
Optimization for both campaigns
Landing pages: Cut an underperforming lander (or direct linking) using statistical significance provided there are a min of 5 conversions between them.
Cut offer: Cut an underperforming offer if there's enough statistical significance.
Cut any clear non-performing carriers, handsets etc.
Bids: adjust if some placements are converting consistently but are not quite profitable.
Spend another 5x payout per day for 2 days on Campaign A_Test.
Spend another 5x payout per day for 2 days on Campaign A_Whitelist.
(Total spend so far: 20x payout.)
STAGE 03 - Ongoing optimization
Add any new golden nuggets from Campaign A_Test into Campaign A_Whitelist and manually adjust bids if necessary to make placements profitable.
If Campaign A_Test not profitable or at least producing some additional promising placements - kill campaign.
If Campaign A_Whitelist not profitable try for one more 10x spend after optimizing as per Stage 02.
If in profit keep testing, improving and then look to scale!
I have a myriad of questions, but this is what I've tried to put down as a starting point for testing... Please let me know your thoughts below!
Be interested to see how you get on, I'm doing the same this week. What offers are you doing? Vouchers for supermarkets for me.
Cheers Tim, planning Sweepstakes for now. The classic iPhone 6 offers...But will start to branch out now and test some other waters.
I'm really impressed by any effort that's made towards setting up any sort of campaign strategy. Having any strategy is better than having none at all because without a strategy there wouldn't be one to improve on, and from what I've gathered from talking to hundreds of newbies, not having a strategy is one of the main causes of failure...it leaves a person asking "what should I do next?" constantly. And your strategy is pretty darn good for a first try. You will no doubt improve on it as you gain experience. Here are some suggestions on potential improvements.
1) At the very beginning of testing new offers, unless you already have landers that have been proven to convert well in the specific vertical, don't just use your own landers that you've made from scratch, or ad text written based on your own angle. Do some spying or use spy tools such as adplexity and/or adsxposed to rip 5-10 landers you see the most often, and add your landers to them if you like. But by including most often seen landers you can vastly reduce the risk of testing offers with bad landers. These ripped landers will serve to establish a benchmark that you can then try to beat using your own landers and angles.
2) In the initial round of testing, try to include offers that are proven. Those that your AM tells you people are doing volume with. That way you can again establish a benchmark that you can then try to beat by testing additional offers, proven or not.
3) Unless you have a lot of traffic at your disposal, try not to cut much of anything, because doing so will a)slow down your testing by limiting traffic, and b)it wouldn't be fair to cut stuff when it may be because your offer and/or lander aren't good enough. You want to test quickly to find the best offer and lander. Once you have a decent offer+lander combo, you can start cutting anything that doesn’t meet your ROI requirement (for me this is usually 30%), including placements and carriers and OSs and browsers and whatnot. You should definitely still cut placements that are big budget drainers though.
If you have a lot of traffic, however, then you can pause all but the best converting placements and save on testing cost that way. However, once you have a good offer+lander combo, you should retest at least some of the stuff you paused from before, as they may be profitable now with better offer and lander.
Just keep in mind that your initial goal is not to profit, but to test massively to find the best offer and lander so that you can make as many placements and carriers etc. profitable. Let's face it, if you cut enough stuff your campaign will be green. But if you're left with a trickle of traffic after your cutting spree then what's the point. Not to mention you can't scale with the crappy lander and subpar offer.
Again, kudos on bothering to formulate a strategy, and thanks for sharing it with us!
Amy
Vortex has covered all the key points there, including the fact that it's far better to have a strategy than not! Good work.
One more suggestion, however: when you're testing angles, don't just test a single lander per angle. At that point, you're testing the image and headline you used in the lander as much as you're testing the overall angle.
I'd recommend testing at least 3 very different landers per angle. It costs more, but you'll get much more reliable data as to whether your angle or your lander is the problem. You don't want to test 21 angles with the same lander, have all of them fail, and then realise that it was the lander that was the problem, not the angle!
Thank you @caurmen @vortex for taking the time to respond, very useful advice!
I have already gained a bit of a collection of landing pages using Adplexity - great tool! So will use them as a starting point and start to tweak them. And apply 3x landers per angle as @caurmen suggested.
I understand the need to not start to 'trim the fat' with traffic right off the bat, do you think I'm being too optimistic that I can tell if the campaign is worth continuing after spending 10x payout spread over two days without any promising (-50% ROI) placements?
Perhaps applying your guidelines on the go2mobi appetizer to the landing pages would be useful here?
If you noted down "No Outstanding Banners" in Step 1, and the campaign did not make at least -70% ROI (in other words, did not make at least $18), pause it, go back to Chopping And Peeling, and choose another angle, unless you have already tried 2 angles with this offer. In that case, choose another offer. The fact that all the banners performed very poorly means it's much more likely to be the angle or the offer that's not working.
If you noted down "No Outstanding Banners" in Step 1, you did not blacklist at least 2 placements in Step 2, and the campaign did not make at least -60% ROI (in other words, did not make at least $24), pause it, go back to Chopping And Peeling, and choose another angle, unless you have already tried 2 angles with this offer. In that case, choose another offer. If performance is poor evenly across all placements, rather than being [b]terrible{/b] on a couple of placements, this implies that our "leak" is the angle or the offer, not the placements or the banner.
@juiceme - depends. What's your ROI over that time? You can certainly gain SOME insights from that, with 3 landers. I'd spend some quality time with http://www.peakconversion.com/2012/0...al-calculator/ and http://statpages.org/confint.html - that'll give you a bunch of insights into what's going on. (Don't forget to count placements into the equation too, though.)
Originally Posted by juiceme
Amy
Thanks guys, yes I agree some actual stats will help us all dig a little deeper. I'm running a couple of different campaigns so will consolidate some data and post back shortly!
Originally Posted by vortex
I've cut the budget draining targets, turned off wifi as it only accounted for 8-10% of conversions and delivering a -70% ROI. On my follow along, I just posted a screenshot of all my stats over the 20-days it's been running, if you have a moment to provide any feedback that would be great.
My follow-along stats link, click here
Thanks Amy once again for some killer advice! Will implement the above now with a fresh campaign and get back with some more stats.
Ahh answered my own question here: http://stmforum.com/forum/showthread...-Landing-Pages
Otherwise, it's spreadsheet time! Don't worry, this calculation's pretty easy. You're just going to calculate the earning per click from the lander (EPCL) of each offer to figure out what the best one to run is.
For each offer, take the minimum and maximum possible conversion rates (as listed on the calculator) and multiply them by the payout, then multiply by 0.01. That gives you the offer's minimum and maximum EPCL.
So, if an offer has a payout of $5, minimum possible conversion rate of 5.2% and maximum possible conversion rate of 8.6%, its minimum EPCL is 5.2 x 5 x 0.01 = $0.26, and its maximum EPCL is 8.6 x 5 x 0.01 = $0.43.
Now, compare all the offers. If any of them have a maximum EPCL below another offer's minimum EPCL, stop testing that offer. If they've had 10 or so conversions or more each and the maximum EPCL of an offer is very close to another offer's minimum EPCL, stop running that offer.
If they have each run for at least 40 conversions and their maximum and minimum EPCLs are very similar, ask your AM when you can get a quality review on each offer. Pick the offer you're closest to getting a quality review (and hence payout bump) on, and run that offer exclusively until you get to enough conversions for a quality review - then ask for a review and a payout bump.
Home > Paid Traffic Sources > Mobile