Gaining Actionable Intelligence - Mobile Campaign Madness to Greatness (6)
Gaining Actionable Intelligence - Mobile Campaign Madness to Greatness
Background Snapshot:
I started my affiliate marketing career close to 9 years ago. My initial focus was ppc marketing through affiliate networks like Commission Junction. I transitioned to SEO and within a few months was doing that full time and have been full time ever since. Prior to that I was in the consulting and venture capital space. Early this year, I decided to focus my time and effort in mobile marketing.
Biggest Mistake:
Getting complacent & going to a 4 hour workweek and not milking Google for every single penny when SEO was booming.
So I decided to spare you the details on a couple loosing campaigns over the last month because I was all over the place testing things like:
- Different types of offers (app install, sweeps, subscription, etc.)
- Traffic Sources: Decisive, Leadbolt, pop, millennial, (ZP is next)
- Affiliate Networks: f5, mundo, click dealer & more.
The goal was to understand how they work, what might convert on what type of traffic, how the affiliate networks differ, getting competent with
WhatsApp Background Campaign 1 & 2
(I ran this a while back but thought i would share the results & hopefully help some folks along the way. I will catch up to current campaigns but had this documented so thought i would share).
I chose this one because I’m having more fun with it after getting comfortable with tracking & seeing some hope on day 1 & day 2 (bot test). ROI -178% & I was actually excited to see it. I ran this one with what I would consider an aggressive lander as my first round of testing.
Day 1 Stats
My Exact Lander for Day 1
So at first I was banging my head against the wall because there were no conversions. I let it sit and a few days later, a few conversions popped in.
Lessons Learned:
- Be patient with conversions. Sometimes there is some lag coming in.
- Do the bot test first. I had told myself I would do this first after my last couple campaigns but still ran with a lander to see if it converted. If it didn’t convert, then move on. I didn’t find many bot placements, but started my blacklist. Also, big thanks to Mr. Green for posting the popads blacklist (that is plugged into Day 3 which is in progress). Hell, with less than $3 spend I received some interesting data. In retrospect I should have spent $30 but I was thinking conversions vs. Blacklisting.
- Pay attention to the categories. Until now, I ignored any potential sign that a category was appropriate for an offer. Don’t ignore your gut. Are people that are on a home improvement/repair site/app use whatsapp? Probably not. But you wont know or get a feel till you test. My guess is that over time, we will all just know after experience what works in what categories and what doesn’t. So I will now take the time to look at each one and segment based on what I feel will work.
- Understanding your prospect: If you are trying to get a visitor to take action, who do you think will convert better? The person looking in:
a) Money and Finance / Investing
b) Money and Finance / Credit Services
My hunch is that the more intelligent profile won’t go for the “you must update your whatsapp” angle. Also less likely to own whatsapp in my opinion. But again, this is a hunch, so test.
- Tracking OMFG: So 99% of my tracking was setup properly but my postbacks were not firing. Thought it was my aff network. More on this for Day 3.
Thank you for allowing me to share, I think I’m learning more by writing this follow along than by just running tests. Please post any comments, questions or ideas.
WhatsApp Background Campaign 3
So what’s new? How is the offer running differently?
- Ran this campaign with blacklisting siteid’s
- Ran this campaign with what I thought was a microtargeted categories but screwed that one up.
- Changed 1pop/24 hrs to 1/8hrs.
- I spent like $4.50 and had a conversion at $4.25. In my mind I was like damn! If only I could have paused it as soon as I saw the conversion I could say I have a profitable campaign. Next morning a second conversion. $15.34 spend w/ $8.50 payout. Paused this and waiting to see if additional conversions pop in like the previous campaign.
- Well I though the conversion tracking was on aff network side so this was still not resolved, but I’m hard headed and wanted more data and figured out the problem which we will apply to day 4. So day 3 cut early.
- I was able to see categories for the conversion in my affiliate panel, but I wasn’t passing the right tokens to see site ID. Without a ton of conversions it really didn’t matter, but fixed this for Day 4.
Lessons Learned:
- Tracking OMFG (part 2): So I decided to hit the support button on
- Setup Templates in
- PopAds Quirk (should have double checked): PopAds has a toggle when you are choosing Categories that says “Automatically include descendant (children) categories.” It is auto checked so all the micro selections I did for Day 3 got undone as a result and included a broader campaign than I would have liked.
WhatsApp Background Campaign 4
So what’s new? How is the offer running differently?
- Fixed all tracking issues (I hope)
- Ran this campaign with blacklisting siteid’s (like before)
- Ran this campaign with microtargeted categories. This may not give me enough volume so we’ll compare to Campaign 3 and adjust accordingly.
- 1 pop / 8hrs
- Run a second offer/lander combo on the same network with the same settings.
- Next phase – Run for a week with multiple landers.
Results: Nothing really stuck on this traffic source that gave me enough data to keep it running. But learned a great deal about pop in general,
APUS Round 1
Last week I decided to go after the good ole APUS Launcher offer and do whatever it takes to make it profitable. First round of testing I chose Turkey as the GEO and Decisive as the traffic source since I'm trying to eat through the balance to get some data.
Campaign: APUS Turkey Decisive
- 2 Angles / 2 Landing Pages: Speed Issues vs. Memory Issues
- 8 Banners for each
- All translated
- Cutting Bad Placements each day
- 3 Networks with a $.24 payout
Results: I wanted to see if cutting placements and increasing bids would lead me to better traffic. After $50 on one Angle and $17 on the other (not sure why i didn't get equal traffic, camps were setup the same), I decided to cut this one as a loss to either traffic source or GEO.
Campaign: APUS Bulgaria Decisive
- So I ran this one with a single angle that seemed to be better, but that is just a hunch since I didn't have any. This ran for several days, but he geo is so small that i couldn't get much data. I did receive 1 conversion but again, just not enough data. Furthermore, APUS just cut the GEO's from 50 to 38 and no more Bulgaria 
.
I decided to change traffic sources and keep testing. Yesterday was a breakthrough (continued on the next string).
APUS Round 2: -52.35% ROI (Optimized -11.73)
Okay so Decisive didn’t work for my initial tests. Lets try a new POP traffic source I have never tried before. The new setup:
- Translated additional LP’s to test Turkey again plus some new geo’s.
- From the previous day I created a campaign just for the profitable placements in turkey from the day before.
- Analyzed the data throughout the day to see where I could optimize
Results: The most excited I have been in a very very long time. Some data finally.
So Turkey seemed to have tanked out a bit. But one of the new GEO’s had a good number of conversions so time to dive into the data. I tested DL, plus 3 aff networks and found the first nugget. While Turkey had a 50/50 split between DL and Lander, the new GEO at the time I checked had only 1 conversion from DL, so I immediately cut that.
New GEO
- LP stats (without DL) have -35.64% ROI. This is starting to get interesting.
- Aff Networks ROI – One clear winner so lets cut those.
Aff 1: -28.63% ROI
Aff 2: -60.92% ROI
Aff 3: -83.08% ROI
- Brands: while some where in the green, it was only from a single conversion so no actionable data here.
- Checked other stats and no actionable data yet.
If I drill down into the Landing Page and then look at the Top Affiliate Network, the ROI would now be -11.73%. Freaking Awesome!
A few more tweaks and we should have a nice test campaign for today. I woke up early and saw some green. Can’t wait to report tomorrow once I have a full day worth of stats. I’m pausing all but 1 campaign so I can start additional split tests.
Other Cool Things: I shot off an email to support asking some questions and showing interest in spending additional dollars. I also deposited an additional $200 which is not much but just showing good faith I guess. It resulted in a dedicated Account Manager.
Todo's:
- Create 2-3 new landers to test with a different look and feel but very similar wording to have a controlled test.
- Setup a CDN – Possible some things were not converting as well due to speed issues. I have a Dedi but when testing camps on the other side of the world, I’m sure I was loosing clicks. I’m going to look into Rackspace but if anyone has a better option please let me know.
- Test similar Utility installs with the same traffic
o Create 2 landers to test for each
o Get approval from several affiliate networks on the new offers
I’m sure I’m missing something but trying to not get too excited until I can see a somewhat sustainable campaign for a few days that has the ability to scale. Still lots of variables left to optimize.
Thanks for reading and I hope there are some nuggets for those in the same boat.
Yesterday was the first day in the green with decent volume. 176 conversions at 33.36% ROI. But also got the email saying the advertiser is pausing the offer. Just my luck!
Home > Mobile > Follow-along Campaigns