Home > Mobile > Follow-along Campaigns

---

6 Months to Profit or Quit AM (10)

---

Very nice to see you start this follow-along! I was just getting ready to sleep when I saw this thread. Instead of typing out a rushed reply I'd much rather look at everything with more attention and focus tomorrow when I'm refreshed. Just want to make a brief post to be subscribed to this thread. Nice progress!


Amy

---

Looks quite promising! Don't worry about the low traffic - once you get profitable you can scale this to many other sources.


You can always use this tool to cut placements:

http://stmforum.com/forum/showthread...Banners-Part-2

However, this tool can be pretty slow in identifying placements to blacklist (although it's accurate). Since you're getting a lot of traffic, feel free to cutting placements aggressively using rules of thumb or even by gut feeling, so that you're only using the best placements to test your offers and landers. That will save you money when testing.

If you want to use this approach though, remember that once you're green, you'll need to retest at least some of those placements then.

Once you have a winning lander, I would highly recommend using that to test more offers!



Amy

---

So i took Amy's advice and reached out to 3 networks and got 3 offers that are similar, took 7 landers from Adplexity and fixed the code and came up with some real good data. The offer that paid the least is actually profitable . What i did was i took the campaign that had the best ROI from my previous test, created a new camp and added two more offers from different networks, kept my 2 converting LPs and added 5 more. Basically this test was an offer/LP test to see if could find that winning combo and try not to cut any placements in order to not decreased my traffic.

The traffic estimator estimated at least 10 impression but i reached 7.6K bidding 2$ more from the average price

Im day parting running the campaign from 6pm - 12am at that GEOs time:

Here are the overall stats for this offer/LP test:

6/29/2016
Cost: $12.97
Revenue: $8.80
ROI: -32%



LP4 and 8 are already profitable, i just need to analyze and drill down to see what offer converted with these 2 LPs. The spent is still low so i might just keep gathering data and make a decision later what to keep and eliminate.

Offer 1
Cost: $4.25
Revenue: $4.50
ROI: 5.95%

Offer 2
Cost: $4.18
Revenue: $3.60
ROI: -13.81%

Offer 3
Cost: $4.54
Revenue: $0.70
ROI: -84%

I think it's safe to eliminate Offer 3 and keep running 1 and 2 and collect more data to make better decision on what to keep and exclude.

Next step is to create 3 campaign to test the bids - low, average, and high to see if i can increase the ROI. I probably should of done this from the beginning and please advice if i should take a different approach.

---

My staggered bids campaign are done and i am now analyzing my stats. This morning i was tempted to make changes but did not want to screw up the data so it let it run and the outcome is very surprising. At first, the only campaign that had conversion was my "high" bid campaign and now that im looking at my stats, all campaigns have conversions.

Im shocked that my "Average" bid campaign outperformed all the others based on ROI. I find it weird that this morning, there was no traffic or ad spend and now i have a very close to profit campaign:

High Bid Test
Cost: $35.13
Revenue: $6.20
ROI: -82%

Average Bid Test
Cost: $7.22
Revenue: $7.10
ROI: -1%

Low Bid Test
Cost: $0.71
Revenue: $0.50
ROI: -29%

Im very grateful that i split test all 3 offers and now i can clearly see which one to focus on. The offer with the lower payout has 5x the conversions than the higher offer so it's very important to split test. None of the LPs are profitable, i dont know if it could be from the offers not converting and money being spent and i need to focus on my funnel more. I have to do the math and make calculated risk.

Here are the overall stats for my LPs from the past week :



Right now, im going to focus on the average bid campaign, check the stats, see what modifications can be made and keep updating here.

Im sorry if im not providing enough details, i only have 4 hours to work on AM on the weekdays and im on 4 hours of sleep and high of caffeine.

I will try my best to provide as much details as possible but i need to get back to work!

---

On my last campaign I took all my winning LPs that were profitable and created a profit campaign with the hopes that this campaign would finally hit the green after doing so many test and my stats are not what i expected:

2016/07/02
Cost: $9.61
Revenue: $7.50
ROI: -21%
Profit: $2.11

After testing 3 offers, i took the one that had the most conversions and seem more profitable. Im only testing Mobile on this offer and that was due to my previous test when I was running another offer and mobile had higher and better ROI. I don't know how this offer would perform with WIFI and Desktop so the only way to find out is to test.

Here are my lander stats:



One thing i have not done is to check my lander speed and try to fully optimize my funnel which i will work on today. Lander 8 has the best ROI but the least amount of clicks. Im thinking of re-creating that page with other alerts and test in hopes of getting something profitable.

Stats are all the same, i have 5 ISP/Carrier that are in the green but i think the traffic will be too low. I could create a campaign based on these 5 carriers and collect some chump change and focus on other campaigns to fully optimize my funnel .

What you guys think?


I also have 11 profitable placements but again, very little traffic and it's hard to make a decision with a $7.00 spend.

Any suggestion is very appreciated.

In the meantime I will

- Check lander speeds
- Run this offer testing WIFI (OR should i run it on RON?)
- Optimize the funnel on LP8
- Look for more offers to test

---

Beautiful! Kudos for being willing to test more offers and landers, ESPECIALLY more offers! Even nicer to see that your effort has paid off!

Are you only testing carrier (aka cellular/3g/4g) traffic? If so, and if you're considering testing wifi, I would strongly suggest that you start separate camps to do that - you know the drill: 3 camps at 3 bids, set low budgets just to test the waters. Depending on your particular offer, wifi can potentially convert a lot worse. The bids you set on wifi should be lower than for carrier traffic as well, because carrier traffic tends to be a lot more expensive (from people promoting carrier-billing offers, driving up the prices).

You can certainly set up a separate camp for tthose 5 ISP/Carrier, but chump change on single-digit spending/day...I think your effort will potentially be better-rewarded elsewhere. Why not use your time to do some testing on 3 other major traffic sources instead? If you can find a source that has a lot more traffic for your geo and targeting, testing there would get you data so much faster, and you'll be more motivated to get green due to the amount of traffic available.

What are you using to decide which offers and landers to cut? If you're not using stats calculator you should be! Allow me to introduce you to each other.


For landers, use the split-test calculator here: https://www.peakconversion.com/2012/...al-calculator/

How to use it: http://stmforum.com/forum/showthread...Banners-Part-1

Basically, you'd need to identify the "current-best" or "most-promising" lander first. If you've been rotating them evenly (like you should!) so that they have approximately the same amount of impressions, your "current-best" would be the lander with the most conversions. Next, use the calculator to compare EACH of the other landers to this current-best lander. If a lander's "probability of being best" falls to 10% or below, cut it.

(As a final check before actually cutting the lander, drill down into OS/device/browser stats to make sure that the lander isn't the best performer for any major OS/device/etc. Some landers intrinsically perform very well, but may have some js/css screw-up that's making them display/function incorrectly in certain OSs/devices/browsers. If you find this to be case you can potentially fix the issues and make the lander perform well over more traffic segments.)


To compare and cut offers - this will be your new best friend:

https://win-vector.shinyapps.io/CampaignPlanner_v3/

I'm forever indebted to the STM member that introduced us to this calculator - they have provided instructions on how to use it here:

http://stmforum.com/forum/showthread...ou-pick-offers

This calculator will allow you compare offers that have different payouts. This is why we shouldn't use the split-test calculator at peakconversion.com - that one assumes all conversions to be equal.


And yes - testing more offers can't hurt! Coincidentally I've made the same suggestion to participants of the 6-week AMC course minutes ago:

You're making very impressive progress! Keep on my friend - double-digit profit days are just around the corner!




Amy

---

I believe im doing both on this and all my previous campaigns. Here is a screenshot of my connection targeting:




Very ignorant of me to not split test these 2 settings, will do on the next round, i feel like Im over the place right now and i need to create a strategy.


I want to master one traffic source (popads), but that sounds like a good idea.

Thanks for the split test calculator. That's what i was using to help me based my decisions on lander cutting but the cut offer tester is new to me and thank you for the link. That's exactly what I needed.


--------------------------------------------------
Here is the profit campaign that i created and ran over the weekend:

Profit Campaign:
Cost: $26.68
Revenue: $20.50
ROI -23%

Im so close and with all the tools i just got, im more confident in my ability to get this campaign profitable.

Here are lander stats for this campaign.



I ran the LP test and LP4 has a probability of 58% on being the best.

This campaign was a good learning experience and im going to run one more test were im going to split test the wifi and carrier and only LP8 and test 3 different bids.

In the mean time i need to figure out how to speed up my landers speed and find more offers to test. I will be completely honest, Ive done very little modifications to the LPs and i want to come up with a different approach instead of ripping landers and competing with every one else in the field:

These are the results for my speed test:



I tried to upload the .htaccess file to my amazon cdn to no luck.

That's all for now!

---

I created 6 campaigns seperated mobile and WIFI, bidding low, average and high. Results are below:



These are the stats for today. WIFI has the best ROI but at this point, im not too sure what to do as far as trying to get this offer profitable.

Im down to one lander, one offer, targetting one OS and one browser. I don't want to cut placements and limit my traffic more than what it is. Im almost close to $200 in ad spend for this offer and with it's little payout, im not sure how i shall proceed.

Just thinking outloud, i can modify the LP and work on the funnel or I could just hang it up, find a new offer and implement all that i learned into a new campaign.

What should i do?

---

Actually there's nothing wrong with targeting both wifi+carrier at the same time. And if you do that, in most cases 90%+ of your traffic will be wifi anyways.

The more important question is: Do the offers you're testing accept just wifi, or carrier as well? Or just carrier? Based on your stats I'd say it's converting well for wifi.

If your offer isn't designed to convert well for carrier, I'd say don't even both setting up a separate carrier camp to test it, because carrier traffic is expensive, so unless you're running stuff like 1-click carrier-billing offers that convert really well on carrier traffic, it's not worth the price.

If your landers keep going head-to-head after a few more conversions, feel free to just pick one and put the other two on pause. (Yes, this is acceptable practice!)



Regarding lander speed: Nice find! 2 seconds is the threshold caurmen talked about in his posts on lander speed - this is the point when visitors will start clicking away in droves.

To find tips on how to speed up your landers, check out:

http://stmforum.com/forum/showthread...-AM-FAQ-Thread!

Do a search on "decreasing your page load speed" and you'll find relevant posts.



There's nothing wrong with sticking to one traffic source! If that's your approach then by all means keep working on popads. It's an interesting network with unique targeting options. The only beef I have with popads is they make it real confusing when you need to target by carrier. You may not need to know how to do this now, but if you ever do, here's how you do it:

http://stmforum.com/forum/showthread...esting!-Part-1




Amy

---

You made a post as I was replying lol!

Your "thinking out loud" is the perfect answer - i.e. either improve your funnel or hang it up.

However, instead of just testing variations of your lander, I would suggest testing more offers first!

That would be the best way to potentially increase your ROI by the biggest amount.

If you can find a better offer, maybe it would be worth it for you to come up with lander variations to test. Focus on testing the bigger things first, then do the smaller tweaks.

Alternatively - if you're tired of playing with this camp, you could always find another geo or even another vertical to play with.

Yet another option: Set up on another traffic source that has more traffic in this geo, and see if you could still break even there. If so you may be able to cut your way to green.



Amy

---

Home > Mobile > Follow-along Campaigns