A First Foray into the Mobile Kitchen (19)
Hey everyone,
So I joined STM about a week and a half ago and spent my time so far trying to absorb as much as possible. Having read caurmen’s guide to getting started in 2014 - I’ve decided to commit to learning mobile.
Given that it’s my first campaign, it seems sensible to create a follow along so you guys can (hopefully) point out if I completely screw up along the way.
I’ll be following the appetiser recipe:
- signed up to Decisive and F5 Media
- Starting budget of $2000, to which I can add around another $1000 each month – having read some other posts this sounds a bit on the small side but enough to start learning and I'm prepared to lose it all doing so
- My goal is to learn as much as possible with the budget and ultimately, in the long term, go from being a cook following a recipe to a chef creating a recipe.
Offer: Psafe Antivirus Brazil
CPA: $0.37
Initially I was going to promote the Dolphin Browser offer since it had a higher volume but, after looking at the reviews on Google Play, I noticed all of the recent ones were terrible - like 1 star bad. Given the awful reviews I didn’t think it would convert as much now as it might have done in the past - was this a good call, or do reviews tend not to make much of a difference?
Spent tonight creating my ads which are listed below - The angle is on cleaning your phone to increase memory/speed:
Set up the campaigns following the recipes instructions. Just waiting for decisive to approve my ads then they’re ready to simmer.
A couple of questions:
- What is the reasoning behind waiting until each campaign has spent $15 to cut and $7 dollars in pay outs? Should this be tailored to CPA? i.e since the CPA for this offer is way lower than $1 can they be lowered? And vice versa if the CPA was higher than a $1?
Thanks in advance for any feedback - it’s really appreciated.
Looks good - you've done a very good job of trying a number of different ad styles there. One suggestion, though - try some more photographic images too, just to cover an even larger spectrum of possibilities.
PSafe is pretty oversaturated these days, but it's still worth testing - you'll probably see some conversions and it's a good way to learn the ropes.
$15: yes, that is indeed a general rule that you can tailor to CPA a bit. Be careful not to overdo it, though - mobile has a lot more confounding variables than Web and it's easy to test too little, which is why I suggest $15. You could try a smaller budget - maybe $10 - per campaign, but in that case make extremely sure to keep a very close eye on placements and pause ones that are eating your budget.
Thanks for the reply, caurmen
Do you have any recommendations of sites with royalty free images? I had a look myself but couldn't find many decent ones, which is why I reverted back to icons in a lot of the creatives.
Just completed a first check over the live reports, only one site to blacklist so far
I wouldn't worry about royalty free images. Google will be fine.
I would focus your banners to look like alerts, system alerts, variations of what a typical alert looks like etc.. These send the message that something is urgent. Just be ready for a lot of ad denials from traffic sources. That's the catch 22
A quick update on the campaign after around 1.5 days of it running. Stats look like this:
Still in the simmering phase so just cutting any site or app that needs to be. Just to double check this though, does the site or app need to tick all of the three factors listed in the mobile cookbook to be cut. i.e does it need to have spent more than 3x total spend of campaign AND not had a conversion AND spent more than 3x payout. I've only cut a few apps and sites, just checking I'm not burning money unnecessarily.
The spend on a couple of the campaigns is also very low compared to others. Is this to be expected or is there a way to boost the spend?
I've also noticed the timezones for reporting appear to be different on Decisive and F5 media. I can't see where there is a setting on Decisive to change timezone to reflect F5 media though...any ideas?
I'll start creating banners around Stackman's angle (thanks for the advice) this weekend to run alongside these campaigns.
By 3x total spend, I'm guessing that's a typo and you mean "1/3 total spend of the campaign"?
IN theory, yes, all three of those boxes need to be ticked. In practice, you can be a bit more selective - if the placement's eaten $3 and the campaign's spent $12, and the other two apply, kill it. The 1/3 rule is mostly for when it's only spent $5 or so, just to avoid killing things too early.
Timezones - do it the other way around. Decisive reports in your timezone, so change the F5 timezone to be the same as yours.
Spend - put your CPA bid up a bit.
Hope that helps!
Thanks for the advice, caurmen. Yep that was a typo, my bad.
One campaign has reached $15 just wAiting for the other ones to catch up while I create more ads around the alerts angle
Good work!
Start working on a new angle too - it's always a good idea to have more angles in the pipeline.
Two campaigns have now hit the $15 mark - but both have made considerably less than the $7 needed to continue ($3.33 and a big fat $0). Upping the bids on the other two seems to have worked with both getting more traffic and one tip-toeing into profit for today (still -ROI overall however).
Created another set of campaigns around the alert angle - see below - and now just waiting to see if they get approved (or rejected) by Decisive.
Hopefully these campaigns will produce a better ROI.
Okay, just got an email from decisive kindly letting me know that all of the ads have been rejected for being misleading - which is fair enough since I was pushing the line.
Going to redesign them now and not mention a virus at all - hopefully that will work
All of my campaigns from the first angle have finally reached $15 spend. Reports on each are below, but none of got near $7.
App campaign:
Site campaign:
Mobile app campaign:
Mobile site:
Looking over the results there seems to be a lot more conversions coming from app placements rather than websites. I guess that's because it was an app download and therefore people already on apps are more likely to download another. Is this usually the case?
I also don't like how quickly the first two campaigns hit the $15 mark. I'm sure 2 days doesn't give a fair reflection so I think i'll lower the daily limit in my next campaign.
Lessons learnt going forward:
Given that there is going to be a lot of failure as I start out I think it will be important to reflect on what I have learnt after each campaign. Given that this was my first I have learnt an awful lot, including:
- How to sign up with affiliate networks and traffic sources
- how to set up a campaign and the initial stages of testing and all of the small steps that come with them i.e inserting a tracking code, targeting etc.
- Photoshop skills - I was a complete novice to start but I already feel a lot more comfortable and am a lot quicker. Still need to learn shortcuts though.
- The need to keep a careful eye on campaigns. I don't think I was quick enough to cull some of the placements on these campaigns and it's something I will need to address going forward.
A new angle:
My new set of ads got accepted a couple of days ago and have been running since then. Not all reached $15 yet but those that have haven't made $7 either. Although there is a marked improvement in conversions across the board. I will post up the results when they have all finished but ad creatives are below:
New offer:
I've also signed up for a new offer - APUS Launcher Android
payout - $0.16
Location - international
I'm yet to sit down and think of angles/banners for this offer yet, but I'm definitely going to promote to an english-speaking audience. I guess UK/US/Canada etc are all fairly competitive, but what about New Zealand/Aus or even English-speaking african countries like Nigeria, Ghana, South Africa etc.
Does anyone have any recommendations about english-speaking countries to target?
Originally Posted by caurmen
@afernald - I'd be hesitant to push an AV offer in Brazil unless I knew it converted really well (and wasn't PSafe) at this point.
@currers - African countries are definitely worth considering, although the difficulty you'll have at the moment is finding good offers. Nigeria's starting to have a few, though.
Targeting AUS / NZ is definitely a solid plan - not a lot of volume, but they convert. Consider other English-speaking countries like Ireland, too.
Campaigns created and launched on Apus Launcher Android App.
I've created 8 banners around the angle of getting a faster phone (apus is apparently the fastest launcher available) and am testing them in both Australia and India. Halfway through today I realised that Decisive wasn't tracking conversions on the India campaigns because I forgot to enter the trackback URL...D'oh! I've corrected it now but have a couple of conversions I can't place. I don't suppose there is anyway to find this out?
Another thing that I wasn't sure on concerned affiliate links for different campaigns based on the same offer. Should the affiliate link change to reflect the different name and location of the campaign? I.e should I change the link be different for the Aus campaign and the India campaign or does it not matter?
Originally Posted by currers
If you didn't have Voluum you should have bidhashes at the network end in the click report, and you can just plug these into your postback URL to send the info to Decisive manually (i.e. literally put the postback URL in your browser, add the bidhash value and load the address).
If you did have Voluum, you may be able to correct the postback URL then manually upload the Voluum click IDs to see if it triggers the Decisive postback again.
As for affiliate links, the network will provide different links for campaigns/offers in separate geo's quite often, especially if they have different payouts.
If the offer says blah blah (IN/ID/TH) or something then they all use the same network link... so no you wouldn't need to change this for different campaigns.
However always check the links section to make sure, for a particular offer, they don't have some abnormal setup with separate links/landers for different countries.
Thanks for your detailed reply @zeno.
I'm not using volume tracking - realised what was wrong though. I hadn't entered the postback URL into f5. Rectified this and it's tracking correctly now.
Could you explain bidhash values though? From my understanding, the s1 value is for non-unique IDs and s2-s5 is for tracking unique IDs - but what is the difference between the two? Also does it matter if you use any 's' parameter i.e s3, s4 etc. or should you always use s2 as the Appetiser guide uses?
Thanks
You can use any s parameter other than s1 for unique identifiers. We just use s2 for simplicity.
A unique identifier is one that changes with every click. Bid hashes are unique per click so are unique. Campaign names, for example, don't change on every click so aren't unique.
Ah APUS looks like a good offer. Interested to see how this goes.
I would focus on international, its the beauty of these apps.. since you can advertise in countries where there aren't many mobile offers!
@caurmen Thanks for the explanation. So do we use the unique identifier so that we know from exactly which placement each conversion has come from?
@stackman Glad to see I've picked something potentially decent. I've started campaigns in India and Aus (Some conversions in India but best campaign is still only -80% ROI, no conversions in Aus but really struggling to get any spend on it. Have upped the CPA to nearly 3x the payout). I've also asked my AM which countries have high volume and high conversions - so on his advice I'm about to launch campaigns for South Korea and Puerto Rico.
Are there any countries in particular that you recommend?
Home > Mobile > Follow-along Campaigns