Bro, do you even mobile? First steps into the mobile world. (46)
Bro, do you even mobile? First steps into the mobile world.
Holla Stackers and Stackerinas,
Been quite a bit since my last follow along in the adult contest. The follow along helped me a lot the last time so I'm up again with a fresh mobile version - though not turning to tits and ass this time,it's all about carriers, handsets,OS and Apps now!
So basically I never ran any kinda mobile except a handful of campaigns on a source I practically owned which was organic so I didn't even bother to track as it was just additional, free traffic from mobile devices.
Speaking so, everything I know (or at least think to know) about mobile comes down to having browsed the whole mobile section while taking notes, reading some blogposts / basic guides including TheAngryRussian's intro (kudos for that, great read!) and talking to a couple of guys - thanks to Mr. Green and Nusolutionz by the way who helped me with some basic questions + server tips. Beside being part of a beginners mobile mastermind, I think this follow along should be a nice addition to help stay focused and get some feedback / tips from you Stackers out there.
Mobile seems to need a pretty structured approach when it comes to testing and especially tracking, so I'll pay a lot attention to stats and numbers, even more than I did in my previous follow along. If it gonna work or not I can't say, I'm a virgin in this vertical. I don't have any fancy target such as $100/day in 2 minutes or XXX,XXX/month after 1 week - my first target is just to get my feet wet in mobile, learn the basics of the variables involved and how to track properly. There's probably no way not to lose in your first campaigns so I'm fine with it, as long as I get something out of it.
As I try to format this follow along as a little mini guide for the basics of mobile as well (kind of a collection from a beginner view), I much appreciate all comments in case a statement in the opening thread is wrong. That's what I've noted down so far and what I'll stick to for the beginning, if you feel I'm missing something, just going plain wrong with an assumption or being totally mad, feel free to reply or PM me 
I'll update this OP later on with some first stats about my setup and my first creatives, so consider this to be kind of a note collection I did before even starting my first campaign on mobile. Maybe some others looking to start with mobile will find it useful as a short summary or collection of bullet points
- Where to start
So before funding and setting up my first campaign, I read the mobile section and talked to a couple of guys I know are into mobile. I got 3 pages full of notes from the mobile section alone plus another page of notes I took from personal convos, so here's what I came up with for my start:
- The lower the CPA for an offer, the more data you get for a reasonable budget and can draw conclusions faster (cheaper). Lower paying offers such as 1 click flows (outside mainstream countries), app or game installs seem to be suitable for beginning in mobile.
- Those offers mentioned above most guys I talked to recommended to direct link. By going direct link first, I eliminate the possibility of loading time issues of my mobile LP and just creating LPs that suck. In my theory, it's narrowed down to getting cheap clicks and finding profitable carrier/handset/OS combinations first which is done by the creatives/angles - at later stage, dipping into offers requiring a lander such as sweepstakes and/or split testing with landers is the next step.
- The above put in a nutshell: start with less technical offers where I understand the conversion flow better and can direct link.
Conclusion: I hit up my AM at Mundo and asked for some tips on which offers to start with that can be direct linked. I'll focus on an app install which I describe in detail in the next post.
- Some opinions / tips I wrote down for the beginning:
- Stick to 1 ad network first rather than jumping around between sources and try to make it work. Each source has its pros and cons just as in any vertical.
- Prepare for losses in the first/second/third run. In the beginning of each campaign in mobile, it's crucial just to get some data and analyze based on carrier / handset / OS / placement much more than hoping for a positive ROI. Generally spoken, my target for the first campaigns/runs is to learn about mobile, about all the variables involved, much more than getting profitable and/or reduce losses at first.
- Most offers only work on specific carriers / handsets / OS combinations. That's where tracking comes into play, so again, the previous paragraph applies.
- On a first test, it's just about setting up some targeting in a way you can quickly see what seems to have potential or not. Next step is breaking out the targeting parts that seem to perform. Important here is to avoid 1 variable (such as 1 specific carrier or placement or handset) eating up all your budget in a test.
- Site / App traffic seem to convert different and have to be tested separated. Same goes for Carrier/Wifi traffic.
- Mobile Penetration in LATAM / Asia is huge, even bigger than in the most mainstream countries, hence I decided to start in those countries. Traffic is supposed to be cheaper and going non-mainstream in a new vertical is more likely to pay dividends than jumping into US/UK.
- Tracking is key to getting profitable, so is a solid server. More about my setup in the next post.
- Talking to your AM / Source Rep is key. Getting stats/infos about what offers are doing well and how the traffic distribution works (there are more things than just eCPM in mobile) is crucial.
- How I'm going to start
So that's how I gonna start, more details about creatives, bids, targeting to follow in the next post:
Server: LiquidWeb Storm VPS SSD 3GB (only used for mobile)
Tracker: IMobitrax
Affiliate Networks: Mundo & F5
Source: IncrowdAds
Offer: DU Battery Saver & Widgets Brazil ( $0.25 CPA @ Mundo )
I'll update later today with my first shot, including more details about creatives, targeting and other factors involved.
Alright so here's some info on my initial setup:
Tracking
I'm going to use Imobitrax for tracking on a Storm SSD. There are other options which are free like the STM tracker or Mobaff, just Imobitrax is pretty much the CPVLab version for mobile. I'm used to CPVLab, so decided to invest in the tracker. The interface is awesome in my opinion, same are the functions.
The tracking has the most important mobile variables included to track, such as carrier / handset / os / etc.. Reading about tokens compared to the Imobitrax outpout, I've found a lot of posts where the source reports different to imobitrax, so I'll use the c variables for tracking via dynamic tokens provided by IncrowdAds. In the end, it's all about how the source defines a value much more than being correct or wrong - I mean I have zero advantage knowing their definition of handsets is wrong or inaccurate, I still need to target those handset in their interface.
Source
I'll stick to IncrowdAds for the start. They got a damn sweet platform with a very nice design and interface, are very open to suggestions on improving their platform and offer a lot of tokens for tracking. Kudos to them, especially Andrea, who's available like 24/7 on Skype and very helpful about questions you have.
They offer you a huge selection of tokens, including conversion tracking. I'll personally use their tokens in my tracker for the following values:
- Media (App/Site)
- Placement (App ID/Site ID)
- Carrier
- OS
- Device
- Category (App/Site Category)
- Subcategory (App/Site Subcategory)
Additionally, I got 1 variable set to track the banner ID a click is coming from.
First campaign setup:
The offer I’m going to start with is described above, a Battery Saver App for Brazil. The offer pays $0.25 on Mundo, accepts Android only and goes straight to the Google store. According to my rep, this is a good offer to start with and can be direct linked. F5 has the same offer at $0.26, so I’ll split between both.
Before setting my targeting for a first run, I’ve looked up the planner tool on IncrowdAds and took a couple of notes about their traffic distribution, carrier breakdown etc. and came up with the following:
- about 1/3rd of their Brazilian traffic is carrier based
- by far, the biggest ad unit is 320x50
- the majority of their carrier based daily traffic is shared across 3 carriers
- the site / app traffic is pretty balanced (almost exactly 50%)
- prices are pretty much different based on ad units, specific apps or sites, wifi/carrier
So after comparing the suggested bid prices (according to Andrea from IncrowdAds, those are neither min / mid / high bids, just recommended bids that should get you some reasonable wins in their bidding system), my first campaign is going to have the following targeting:
- app traffic only, 2 separated campaigns for Wifi / Carrier traffic
- for carriers, I’ll be targeting 3 carriers only
- no apps/sites blacklisted. At the beginning I’ll keep an eye on the sources to see if there’s not 1 or 2 placements eating up my budget and adjust accordingly
- in total 6 different creatives, each with a $10 daily budget to start. So in overall, I’ll have $120 worth of data - $60 for wifi/carrier - if everything goes as planned after the first day and take it from there. The creatives are in different colors and use different CTAs / angles, all animated, translated into Portuguese (Brazilian Portuguese) by OHT. I’ll worry about optimizing angles (creatives) later. Got to have some data first!
So in a nutshell, this is how I’ll start:
1) Carriers, In-App only, Android Only, CPM $0.20
2) Wifi, In-App only, Android Only, CPM $0.16
6 creatives in total, $10 per day per creative. Direct linked, splitting between Mundo & F5.
The bids are based on some average numbers I calculated from the planner, since you can’t combine specific factors to suggest a bid price. So for example, the average bid for a carrier might be 0.22 while the avg. bid for Wifi traffic is shown as 0.18 with in-app being priced at 0.14. The CPM looks a bit high to me especially for Carrier traffic, but I’ll see how it goes. Based on the first data, I can play around with the bids later on.
In case I'm taking a wrong step on something or completely mess up my first setup, feel free to give any feedback, no matter it's a bless or telling me how dump I am.
Awesome. Subd!
First update: after having the campaign let run for about an hour, I noticed that I wasn't getting any kind of traffic. I upped my initial bids to $0.20 for Wifi and $0.25 for Carriers. After doing so, traffic started coming in, but I was stuck in a situation where the majority of my traffic came from just 2/3 placements.
Checking after around 5h again, I've spent an average of $0.80 per creative in that 5 hours having excluded those sources that sent me the majority of my traffic. Definitely nothing to work with. After checking the planner, it's obvious why, there are a handful of Apps that seem to deliver a lot of traffic, but their bid is suggested in the 0.45ish range by IncrowdAds.
As I am seeing somehow reasonable stats for those 3 targets so far, I'll include them again and just let it run to collect some data. So far, the stats look kinda ok without any optimizing, with some placements hovering around 4.5 - 5% CR without any optimization.
Regarding the networks, Mundo converts at around 4.5% for both, Carrier and Wifi, while F5, after a total of 300ish clicks, stays with 0 conversions so far. 300 clicks are not a reasonable sample size at all from my understanding in mobile, but I think it's reasonable to assume that Mundo does a better job on this offer, hence I'm sending all the traffic to Mundo over night.
So to put up a summary, this is how it looks right now:
- I needed to up my bid after talking to IncrowdAds
- I saw traffic coming in, but majority of it was from 2 placements. I excluded those placements, traffic got stuck again.
- Mundo seems to convert better than F5 for this offer.
- With that being said, I will reactivate the 2 placements for both campaigns that gave me the most traffic. Before activating them again, these placements are at 4.8% / 3.8% average (combined Carrier and Wifi), so there should be room for optimizing.
First lesson learned, always check (if possible) the avg. bids / recommended bids for placement if the networks provide them to you.
I'll update tomorrow with some first stats and trends I see and hopefully see some first trends I can optimize based on.
A couple of questions that came to my mind in this scenario, it would be awesome if some of the mobile guys could jump in:
1) If a situation like this occurs, would you recommend to focus on those placements giving the most traffic and optimize based on them? With a CPM of 0.45+ I think especially for a low paying offer like this, it's close to impossible to get profitable.
2) Is it too early to switch the traffic to Mundo only after 300 clicks and no conversion for F5? I know it's split over 3 carriers, but still, the difference is quite huge.
3) For optimization tomorrow, with the current status of just 2-3 placements delivering reasonable traffic, would you optimize only based on those placements or still watch the overall performance when it comes to carriers / handsets?
4) Is it common in mobile for some campaigns working only on a handful of placements? Logically thinking, the demos for apps are completely different. Just think of a chat-app for teens compared to an App about historic events (example only). From what I've noted, it's definitely about converting carriers & handsets, but does it apply to specific placements too? Following the general advice to stick to one traffic source much rather than jumping sources, I'd assume the logical step would be: see where the traffic is coming from, get stats from those sources, optimize based on those.
Gonna update tomorrow with some P/L and CTR and all the sexy stats we affiliates love, so stay tuned
I'd definitely recommend testing out the Bayesian split-test calculator linked in my signature for this one: could seriously reduce your maximum spend needed.
My gut feeling is that you're on the boundaries of statistical significance as far as the offers go, but probably on the right side of it. However, the Math Will Know
GREAT follow-along: super-detailed and really interesting. Thanks! Looking forward to seeing what happens next.
Alright so here's the first update. Lots of data, as expected a wealthy loss, but I don't care. I got first data to take a look at which was my intention. I'll divide the stats by Carrier & Wifi so it's easier to follow:
Overall Stats:
Spent 95.74
Revenue 16.5
ROI -82.77%
Carrier Stats:
Spent 47.46
Revenue 7.75
ROI -83.67%
Avg. CPC $0.05 (definitely too high)
Right off the bat, I noticed 2 things.
1) IncrowdAds shows a higher CPM than I selected for some placements
2) Majority of the traffic comes from 5 placements with 1 eating a lot of budget. Described that before already, I'll probably focus in this placement first. Surprisingly, it's not the top volume App shown in the planner.
STATS Carrier
I will exclude the P/L breakdown from imobitrax since it's not accurate, the important thing is the CVR first. I don't care about positive or negative at this stage, I just want to get something to work with today.
2 carriers seem to do better than #3, so here are the screenshots including some data for all placements and the carrier #1 & #2:
So given the stats in the screenshot above, carriers #1 & #2 seem to be better than #3 for this offer. I noticed a stronger performance CTR wise for the creatives on #1 + #2 as well, along with the higher CR rate I will exclude carrier #3.
When it comes to placements, clearly 1 placement provides the most traffic with 1 placement not doing much and 3 placements looking kinda ok CR wise. The problem for those is that with 40-60 clicks for each, I don't have any significant sample size to make a decision based on.
I have spent less than I planned per creative. After a chat to IncrowdAds they told me my bids look pretty high and should get me some reasonable volume, but god knows why I didn't get it (bidding CPM). Waiting for them to come back to me as they wanted to check something about their ad delivery.
So given the data I have for the carrier campaign so far, my plan for today looks like this:
1) I will duplicate the banners b1 - b5 on IncrowdAds and target 4 placements only (see at the Placements overall screenshot). B6 did poor compared to the other banners, so instead I will focus on the first 5 and get more data for those. I'll run them to the existing campaign on my tracker to get the data combined.
2) Exclude carrier 3 in the targeting.
3) Test if I get any traffic bidding $0.20 CPM for those 4 placements. I'll watch for about 2 hours and depending on winning bids in IncrowdAds, stick to $0.20 or increase back to my initial $0.25.
So in a nutshell for today for the carrier targeting: Cut banner6, cut carrier #3, try a lower bid and target 4 placements.
The Wifi write up is about to follow later today in the next post.
Reserved for Wifi.
Originally Posted by caurmen
My gut feeling is that you're on the boundaries of statistical significance as far as the offers go, but probably on the right side of it. However, the Math Will Know
GREAT follow-along: super-detailed and really interesting. Thanks! Looking forward to seeing what happens next.
What CTRs are you getting on your banners?
Your CPCs are about 5 times too high. I wouldn't cut any of your targeting until you see those click prices drop a decent amount. You will only have a small amount of traffic to work with if you start cutting at this stage.
Also.. I think you are focusing quite a bit on carriers right now. I'd focus on getting your CTR higher and costs down, then start looking at the handsets that are converting as data builds up. It's all about the handsets/device type and screen resolution IMO.. (aswell as carriers lol! too many variables)...
Originally Posted by Mr Green
Right now I'm hovering between 0.03 and 0.05 CPCs. I'll follow your tip and ignore the first data I collected and set up new creatives first with the intention of finding high CTR creatives and taking it from there.
So plan for today + tomorrow is to get new creatives up, get them translated and see how they perform in terms of CTR. Shall be a long evening with WRW, I definitely need to get a better feeling for how mobile banners should look first. Lesson learned, cutting too early makes you end up with traffic shortage right away and kills your campaign before it even started.
Originally Posted by dynamicsoul
From your guys both replies (correct me if I'm mistaken), the first step before cutting anything shall be to get high CTR creatives, therefore cheap clicks. All further optimization should be based on those creatives bringing in a high CTR, not on the first results including lower CTR creatives. So it all comes down to cheap clicks, then cheaper clicks, and as the last step, even cheaper clicks right?
Yeah, pretty much.. ^
Just about to upgrade my server and jump into the mobile game myself but I'll be starting with adult traffic. Your last follow along was great and really helped me learn TJ. Best of luck fjk87, will be following this thread closely.
great follow along! tons of great information already!
Damn big follow along man, lets connect on skype this week.
Very informative, hopefully lots of lessons being taught out this thread love these mobile ones!
Thanks for the kind words guys, for sure that's motivating.
So right now, I'm waiting for Incrowd to approve new creatives. I changed my approach a bit in more 'general', hopefully higher CTR creatives from pretty much battery style targeted ads first, so I'm eager to see how this will affect CTR and especially CR. The first on the plan is to get cheap clicks now and take it from there. On my plan is to get cheap clicks first before anything else.
So after finaly receiving my 'one hour translation' for the creatives after around 7 hours or so, everything's uploaded and ready to fire up. Setup for today looks like the following:
- 5 different creatives, all themed very general and not targeted to Battery, to get cheap clicks. Think you have a msg / 1 app ready similar styles. In total I have 12 translations ready but going to test 5 creatives first.
- I've set up 1 campaign for the 3 biggest carriers according to Incrowd's planner for each, App and Sites. Each creative is set to run for $10, so in total I hopefully will have $50 worth of data for App and Site each. I decided to split app vs sites to see if I notice different CTRs, different stats in overall. For future campaigns, I'll stick to either Carrier or Wifi and split test Sites to Apps, I think that's a better approach than run Carrier along with Wifi on one media type. There's enough variables involved already, so probably diving by the 2 'main' groups of Carrier vs. Wifi gives a bit more structure in my tests.
- Different to my first run, to avoid cutting too fast, I will let the $10 per creative run and look at the stats after it's all spent. I want to avoid cutting sources too early which seem to take a lot of the budget, but looking at the planner, it's to be expected if they have like 5% of the overall traffic for Apps according to the planning tool.
- After the first $10 spent per creative, I'll take a look at the stats and sort by placements to see if my data is based on 1-2 placements only or if it's a bit more spread. I'll post a screenshot of the placement stats before continuing to run and post my thoughts along.
- I've upped my bids a bit since I noticed a pretty big CTR change on a quick shot for 10 bucks on a creative yesterday. After upping my bid I ended up with cheaper CPCs, they were still too high, but comparing to my initial bid, CTR almost doubled.
Eager to see how this run goes. I'll spend a bit time browsing through WRW and getting inspiration on banners and update once the first run is finished.
Alright so here's some update on yesterday:
Couple of goods / bass seen yesterday:
The goods:
- new creatives perform strong CTR wise, getting between <0.01 to 0.025ish CPCs
- seems like there was some issue with the F5 offer, I'll include in my next runs again as there was some issue on the pixel firing. Seems like it converted better than Mundo, but I'll run both 50/50 in the next run to really have some data that I can analyze
The bass:
- After 24h of running, my campaigns for both, Carrier Sites and Apps, have spent about $5 per creative only. My bid is pretty competitive according to the last convo with Incrowd, I think there's some issues with their delivery system right now. While setting up the campaign yesterday, I saw Brazil disappearing from the planner tool and coming back, I'll chat back and wait for a response about this to see if it's my bid or an issue on their end.
- For Carrier traffic on sites, I received 1930 clicks in total, of which 1291 came from just 1 source which converted at 0.7%. In overall, for the sites I'm at a 1.24% CR which is poor, but looking at the 60% of clicks coming from just 1 placement, this data isn't really convincing.
- Similar issue with the App placements, with 966 clicks and an overall CR of 2.69%, but again, majority of the clicks coming from one placement (418 clicks) . The only good aspect is this source so far converts at around 3.83%. The data I got for app placements is for around 75-80% just from 3 placements, so pretty hard to draw any conclusions so far.
I'm waiting for IncrowdAds to respond to my message and see what's up with that delivery problem. Generally, traffic is very low and stuck on a handful of placements only although I selected the 3 major carriers for both, in-app and display. Combining both together, I see 1 carrier outperforming the other 2 by a bit, but I need a more equal distribution on several placements before starting to cut down targeting first.
I'll update later today about what Incrowd recommends to do for a better distribution and more traffic and probably get some data posted later today or tomorrow morning. If somebody has similar issues with Incrowd for today & yesterday and found a solution, feel free to hit me up please.
Incrowds have a really good platform.. but a few bugs to iron out.. and there, s not a great difference in the sites / apps from places like sitescout.. they are ok.. but I found it hard to get a wide variety in volume. . Most from a handful of pubs that weren't converting... Ive stuck with my other sources for now. No point changing something if its working for me.
Try split test between CPM and CPC. I literally ran a test 5 minutes ago, and found that I got more volume for cheaper with CPC.
May do.. I seen the CPC was beta. So haven't tried it.
Dude anything in beta is an opportunity to take advantage of. I interpret "beta" as beta try this shit before other affs catch on.
Yeah I get ya.. but also can mean.. "doesn't work properly yet, but we are offering it anyway".. lol.. 
much like "bars" and "phablets".
Alright sorry for the missing update yesterday, had a pretty busy day at the office.
After having an in-depth chat with Incrowd, I follow their suggestions to remove carrier specific targeting at all and just selected 'mobile' including all carriers. I got all creatives set up for a new run again having excluded that one placement giving me 70% of the traffic, hopefully I see better traffic distribution today. Depending on how it goes, I'll obviously stick to the offer or probably take a different offer as I want to stick with Incrowd first to avoid source jumping right off the bat.
Update coming later today, hopefully with some better spend
Short update: not looking good so far, after about 2h running, $0.50 in average spent per creative. One placement is taking 55/105 clicks so far and the CTR on this placement is horrible compared to the other placements.
I'll let it go in the hope to change, but I'm not too positive. Basically I got 3 choices now:
1) Search a different offer in a different country and stick with Incrowd making sure they got more distributed traffic for that new geo.
2) Give Wifi a shot and see if that helps, though my initial plan was to stick to carrier site+app traffic. Plus, I faced the same delivery problem with Wifi traffic in my previous run.
3) Look for a new source that seems to have more traffic / more distributed traffic in Brazil for carriers.
From what I've read and my notes, I think I should search for a different country and stick to Incrowd. The only thing that puts me in a bit of itchy situation now is that after F5 credited the leads I had from them, I had almost double the conversions on all my previous runs, hence CR looking pretty solid.
Another option would be to stick to Incrowd for a new offer and look for a different source for this Brazilian offer. Kinda not 100% confident how to proceed, though I think 2 campaigns would certainly be doable at the same time.
As always, thoughts and tips on the choices described are much appreciated guys!
Definitely recommend you trying wifi traffic. I just looked at incrowd's planner and the bulk of Brazil traffic was categorized as wifi.
Alright so here's some stats for Wifi, don't know what's wrong, but not much traffic at all. After about 3-4h of running, each creative spent around $3.80 only. The only good thing is I'm getting very strong CTRs in the range of 1.5% - 2.9% giving me < 1 cent CPCs.
Round about 1.2k clicks, with 900+ clicks coming from effectively one source which is double listed in Incrowd. My bid is about double the recommendation for Wifi, feels like it'll be hard to get any volume, even including the biggest targets.
I'll have a chat to Incrowd today again but given that even including all sources on Wifi which is supposed to have more traffic, I will have to look for a different solution, at least for Brazil.
Anybody else experiencing similar issues running Brazil in Incrowd? I love the platform and they're awesome when it comes to communication, but it's like nothing I can do right now with that kind of traffic given.
Right now, it looks like I have to watch for a different ad network if I want to continue with Brazil, F5 added another couple of conversions so I see potential in this offer. Going against the advice of sticking with 1 traffic source is actually what I'm leaning towards, if there are strong arguments against doing so, feel free to hit me with them!
Going to update later and talk to my AMs about alternative offers as well to be able to stick with Incrowd, update to come!
Hmm I'm not sure what is wrong with your set up. One of my Brazil campaigns got around 15,000 clicks yesterday, and I'm not exactly bidding aggressively.
Originally Posted by Mr Green
^^ haha fuck sorry I was meant to say "I'm not exactly bidding aggressively".
Subscribed!!! Good luck for your follow along!!
Shame on me for the lack of updates. Simply reason: I'm busy as fuck with some SEO clients, budgets are increased to spend as much as possible to avoid taxes, Christmas is coming... it's like the busiest month of the year!
So without fancy screenshots or anything, here's a bit of summary from my recent run on Incrowd (think it was Saturday or Sunday):
I focussed on Wifi and 1 placement only and was able to achieve clicks in the $0.004-$0.008 range for all 5 creatives I ran. Without any optimization, I was at about -60% which I found quite amazing despite the loss, first time I saw something not completely tanking in mobile since I started this follow along.
Now the issue was even bidding pretty aggressive, I wasn't able to spend more than 15$ over roughly 1.5 days I let the creatives run. As explained in an earlier post, Incrowd was really helpful to me so far and explained the issue which sounds logical why I was a bit stuck on traffic.
So after finishing a marathon of meetings tomorrow, being out of town for the last 2 days I'll give my best to start updating at least daily from the new week on, most likely Tuesday.
My idea is to take the following route now: as Incrowd doesn't look perfect for Brazil for me, but the offer seems to convert reasonably well, I will add a 2nd source just for this Brazilian offer. I'll definitely stick with Incrowd and dip into a different offer in a different country as well, so probably will have 2 campaigns+ from the next week on which I'll look to get profitable over the coming days.
I know these non-continued follow alongs suck like mad, apologizes for everybody who expected more, it's just I can't leave my clients alone now with the business hammering me harder than expected. Sounds a bit odd for a problem, but it definitely is for me. Will update coming days, thanks for the feedback so far guys
PS: If you feel like having trouble with SEO / questions, catch me on Skype for a quick chat. Some SEO guides are coming soon to STM as well ^^
High-quality problems like having too much work on are still problems Looking forward to hearing your updates when you get time to breathe!
@fjk87 I'm just working on some tests now to help you out.
Alright so finally back after a busy week with some hunger for moarrr
So as described from the above posts, Incrowd doesn't seem to be too good for Brazil so I took a shot with Aduiqity and set up my campaign there dividing Wifi into Sites & App placements. The campaign is running right now, according to my convo with my rep they got quite a bit of Brazilian traffic, so let's hope I'll see some traffic coming in today!
To stick with Incrowd, I started asking my AMs for an offer that's doing good and which has shown growth over the last couple of days, both right up recommended me to give Mobigenie (INTL) a try. The offer is a rather generic app download excluding just a handful of countries, so the perfect offer to stick with Incrowd and give it a shot. Offer pays $0.20ish on Mundo & F5. I've setup 2 campaigns in 2 countries after getting a top performing country list from my AM. Each campaign is 'researched' on IncrowdAds to catch the majority of the volume available in the specific country and have a somehow equally balanced traffic distribution upon more than just 1 or 2 major sites/apps according to their planner.
So in a nutshell, Adiquity for Brazil with the same setup as before on Incrowd. Mobogenie in 2 countries on Incrowd with 5 pretty different creatives. One thing to note is the countries are not English speaking but I kept my ads in English first to see how they perform CTR wise since after a quick shot with 1 creative being translated over a $10 spent, I saw English actually outperforming the native language. Bidding is set up in the way David from IncrowdAds recommends in this post, so now I just have to wait to get everything approved and take it from there. Update to come later today or tomorrow, depends how much traffic I'll get, probably going to spend the day browsing through WRW and contacting sources about their traffic distribution as I need to send out some funds before the year ends to reduce profits and therefore taxes.
Ok here's the first updates, I'll divide 1 post per campaign to make it easier to read.
Campaign #1 (Carrier & Sites):
The data is a little bit corrupted as it includes a quick shot on regular CPM for 1 creative (b2), the creative with smartCPM bidding actually performed at 1.4%CR and a 2%CR in average. This quick run on regular CPM sent a lot of traffic from one source which I blocked afterwards as it was eating a major part of the traffic. In average, I spent around $9 per creative. The CPCs range between $0.004 and $0.16, pretty dependent on carrier and placement combinations.
Thoughts:
- 3 creatives seemed to perform best, I plan to just let these 3 run today to collect some more data and tomorrow create new ones based on their angle.
- Mundo performed a bit better than F5, while at the other campaign (diff country, Wifi + Apps) F5 was doing better. I'll keep both offers running today for a better sample size and then decide to switch to Mundo/F5 only or not.
- I was running 5 different creatives stopping 1 create that didn't perform at all, CR was kinda good but the clicks been extremely expensive, CTR just sucked. Hence the stats for 4 creatives only.
- My traffic is coming from a lot of carriers which are performing different, both CR and CTR wise. Some carriers average at 1.2% CTR while some are at 2.2%. I'm not sure if I should focus on the carriers that have 200+ clicks only so far, according to the planner they're good for about 80% of the traffic Incrowd has on carrier. The carriers with 85 clicks / less are a bit short on sample size I think to cut already. The only carrier I think about excluding is the one with a 1.09% CR.
- There are some placements that don't seem to convert compared to others at all, but I think to have them included today running with just 3 creatives as I feel 1xx clicks is a bit short to cut.
Plan for today:
1) Let the 3 best performing creatives run, collect more data to analyze. Meanwhile, create some new creatives based on these 3 angles.
2) Cut the carrier performing at 1.09% after 274 clicks. Otherwise, I'll keep all carriers included.
3) I'm not sure about the placements, tempted to cut them but then again the data is a bit short I think. Any opinion if the clicks are enough to cut (talking about the 3 worst performers with 100+ clicks) is highly appreciated, my gut feeling combined with lack of experience in mobile says compared to the other placements it's ok to kill them, then again, it's not too many clicks on them.
Campaign #2 (Wifi & Apps)
Thoughts:
- similar to the other campaign, 1 creative pretty much tanked CTR wise so I decided to run the remaining 4 only. 2 creatives clearly perform best, both are around 1.5% CTR in overall. I'll run these 2 today to collect more data on them.
- I got 2 placements which my traffic is coming from in the majority, both doing kind of ok. In fact, these 2 placements are about breaking even already so definitely big potential there. The next 2 placements in terms of volume doing pretty much shitty so I plan to exclude them.
- Different to the carrier campaign I lack carrier(s) to exclude, so I have to focus on other factors like certain handsets or OS versions. At this stage I have not enough data to exclude based on handsets or OS versions.
Plan for today:
1) let the 2 best performing creatives run more to collect data.
2) exclude the 2 non performing sources
While setting the campaigns back to run I'll go browsing on WRW again and brainstorm a bit for angles related to the ones working along with designs. I'll spend about $20 per each creative today to collect some more data, so expect the next update by tomorrow.
Interesting news on the English ads! Are the translated copies you're using pro translations or (as per Stackman's case study) Google Translate based?
The translations are done by OHT but both countries are known for having a lot of web content in English.
Yeah you have enough data to cut those shitty placements based on your payouts.
Btw, have you tried a splash lander yet?
Campaign #1 (Carrier & Sites) Update:
So the campaign pretty much tanked compared to the previous day:
So it's time to take a look why and I think I found the reason. This is one point I would love to get some feedback from the mobile stackers if my thoughts are moving right or wrong:
Lack of sample size. In overall, my data was based on mainly 4 creatives with a total of about 2000 clicks. This data was spread across carriers & placements. Since the country has a lot of volume, I crawled through the Incrowdads planner again and took a closer look on my own tracker as well to figure out if in fact I got something I can use for decisions or not. Turns out not at all.
Even the banner with the majority of clicks (b2) doesn't have any sample size in the end. Sorting the stats according to creative and the biggest carrier and biggest placements, even the 2-3 biggest placements received only 35 clicks for this carrier with 1 or 2 conversions. That's around 1x CPA spent for them and the stats are totally irrelevant on this small sample size.
The mistake I think I did is basically on such a huge geo with a dozens of HUGE placements and several HUGE carriers, I did not separate at all. I just set up the campaign and let it run for not enough money to be able to draw any conclusions. In fact, my only targeting was the OS and creative sizes along with excluding Wifi. If taking this kind of broad targeting right off the bat, I need a much higher spend to collect data and therefore analyze.
Update for the second campaign to come later, it's doing much better number wise though not being profitable yet.
Notes taken for future campaigns / giving this geo a fresh try:
- With loads of traffic coming from different carriers and placements, there's no way to draw any conclusion after 2000ish clicks. Even with this small payout, it's hardly spent 1x CPA per carrier & placement. Not even starting about devices. Probably started to cut way too early and using completely irrelevant stats for my analysis.
- The only thing I can take from my previous data is which carriers got a bit of traction or not, limit my tests on a few carriers that drove some clicks and take it from there again with some better structured testing.
- Get my shit structured. With the shitload of variables involved, I need to have a plan about what I'm actually doing which wasn't the case here.
From what you've said above. . Id be inclined to take it one carrier at a time.
Separate them out into separate campaigns one by one..
So camp 1 carrier 1 all placements x 10 banners..
Camp 2 carrier 2 all placements x 10 banners
Then cut from the back up.. cut the lowest converting banners first.. then placements and see if each carrier can be made profitable or not.
Just my two cents.
Campaign #2 (Wifi & Apps) Update:
With the elimination of the worse creatives I took a look at placements + banner combinations. One banner clearly did best on the one placement I was left with traffic, so I decided to focus on this and ran it for about $60 spent with a return of $50, so definitely moving towards the right direction. Here's the screenshot for this creative on the placement from my tracker:
Right now I'm at around -20% ROI with this campaign which I love, but the problem is excluding manually and combined with smartCPM, there are just a few combos left since looking on some handsets, some clearly outperform others. The problem that right now it's narrowed down to 1 (big) placement but I doubt this is the right step micro optimizing. Honestly spoken, with a -20% ROI after 3-4 days as a first result I'm absolutely happy.
So after this 2 short campaigns I have a bit of 'click' feeling when it comes to setting up new campaigns / going more into mobile. This stuff is basically advised all over the mobile forum, I even had it in my notes before, but seeing it on your own campaign / numbers probably teaches much better than reading. Thing I'll try to change / take care of for my next attempts:
- avoiding excluding / cutting the targets too early. More than any source it's a numbers / combinations game and I need some sample size to make decisions. I definitely cut my targeting to early which leaves me with lack of options and therefore lack of potential for optimization and volume.
- taking a closer look to the geo I'm going to run. I made the mistake with the first campaign. I was amazed about different carriers available that I ignored the lack of sample size for each carrier + placement combination. Again, cut too early, had the wrong plan right off the bat. If there's much inventory available, I need to set up a more structured approach to separating carriers to get reasonable data.
- I'll force myself to try more totally different angles when starting and cutting the non performers after some reasonable sample size.
- I got to get more used to the amount of clicks compared to web, it's a totally different game with so many variables involved. At least I'm feeling like understanding the basics a bit more now.
So with the notes taken and a better plan on what to do, I'll continue with the same offer for now. Next updates to come, let's see where I'm at in a couple of days.
Great progress mate. Remember the offer is open to like 95% of countries. I'm not sure how you picked the first two countries, whether it was random, or whether your AM gave you a tip. If it was the later then maybe the bids have been driven up because of that.
Monster mobile campaigns come from tweaking your lander to suit the traffic, rather then tweaking the traffic to suit your lander. If a carrier/handset gets more then 0 conversions, leave it on. If you've run decent traffic to say that those handsets/carriers convert at 0 then thats okay, they most likely don't support the offer.
Cutting your targeting SEVERELY lowers your volume. I was always one for running many many mobile campaigns, just a maze of carrier/handset combos that worked and yes this SURE still works for the technical person, but I find myself a bit more creative so using my landing page as leverage to the traffic gave my campaigns an extra digit
Super old now but still wondering:
how did it turned out at the end?? Things you have learned?
I really liked this follow along i can't believe I only saw it now.
Thanks for the input!
Haha fill us in fjk87 !
Home > Mobile > Follow-along Campaigns