First Mobile Campaign and mistakes made (11)
Hey Zoe! Props for taking action!
As far as I know Portuguese is the most widely spoken language in Brazil. Translate your banners 
I would say for a low payout offer like DU speed booster you don't have to spend so much - Especially if it's your first campaign.
A click difference happens all the time, usually it's a loss and not an increase in clicks throughout the funnel though 
Could have many reasons, I get 110% LP CTR from time to time haha. Would not worry about it as long as it's an increase.
For optimization I better shut my mouth, I've got almost no experience with Tapit! and others are way more qualified to give you advice with that.
Good luck and keep us updated!
Hi Kevin thanks for the advice. Will get my banner translated.
TapIt allows a minimum daily budget of $20 for each campaign. I have no choice but set a $20 daily budget.
TapIt! let's you optimise by publisher IDs so make sure you are using their token to send those through.
You must pass the
Your banners are hideous, which is good haha, but make sure the text is easily readible. It's not very sharp and the red text moving over the image makes it hard to read.
Hi Zeno thanks for the advice. Anyway there are few questions here.
1. May I know what is publisher ID? I'm afraid I'm not using them now. May I know what will happen if I'm not using them?
2. I did append the append the click ID to the offer URL as per
3. Yes I am bad in designing. A chemistry student though. Alright gonna improve the text color.
Appreciated for all the advice 
1. A publisher ID is a code that identifies where your ad was placed - i.e. the site or app. On blind networks these are just numbers/codes rather than descriptive names. These are 100% the main things you want to optimise first when running mobile campaigns as different placements give wildly different performances.
It will pass through the site using the [site] token. Information from TapIt! - http://tapitmobile.zendesk.com/entri...k-optimization
2. Yes. Go to Neverblue > reporting and go to the click report. You should be able to see clickIDs under one of the subid columns provided you passed it correctly. I would not tick "Append click ID" in the affiliate network setup section of
3. Chemistry? I approve (if you didn't know, my PhD is in physical chemistry).
Good luck!
Hi Zeno thanks for the note. Yes I posted the clickID from Neverblue to
1. I used the site token and I found they are just number as you said. 6 out of 7 conversions are from this site ID. This means it is from Site traffic? When I checked my
2. 7 conversions might not be good for $70 spent. I planned to upload the ads (translated to Portuguese and adding a landing page). Should I target the same wifi-app, wifi-site, mobile-app, mobile-site? Or I should just optimise base on these 7 conversions? Or should I try new angle?
3. Any recommendation on testing budget for low payout offer such as this $0.22 offer? As Kevin said above I seems like spending too much.
Thanks for all the advice given.
The site IDs are specific placements - there will be hundred or thousands of them for both app and site categories.
If you drill down into your data at TapIt! you can see which site IDs provided traffic - so can see impressions, spend, clicks, etc. Often one placement will take a lions share of the volume.
You don't yet have enough data to kill off carriers... you'd need to see a global trend showing that you run at a loss on WiFi or Carrier traffic (you must look at TapIt! data to see how much you have spent on these various metrics unless you are paying a set CPC). You can also remove specific carriers from your targeting but again, you need solid data. If 90% of the traffic comes from carrier X and it gave 4 conversions, whereas the others so far have given 0, you do not have enough data to justify killing off everything other than carrier X - far too premature.
2. $70 spent and like $1 made? How many clicks got to the offer? This is very much a situation where you should kill it and try new angles or completely ditch the offer. Ideally you would never go to this much spend - 300x the offer payout - before making such a decision.
You should always split by those 4 listed to keep data analysis convenient. You could optimise based on 7 conversions but this is sensible only if the data justified this in a statistically significant way. Without seeing the data I can't help much.
3. Translate first, not afterward. Follow the mobile cookbook as it lays out some optimisation thresholds and guidelines. Test lots of angles and kill poor performers without wasting too much money on them. Bid CPM and don't use sources where you have to bid CPC (e.g. LeadBolt).
Thanks for the reply again Zeno. I agree with you as the data is just too little to conclude anything.
1. There is 1588 clicks shown in Neverblue and 7 conversions. Spend almost $80 to get $1.5. Not sure if my angle or the offer suck.
2. If my offer is low in payout, is it better to bid CPM rather than CPC? When the payout is high CPC will be better?
I'm very grateful for all the advice given and really glad to sign up STM!
It's probably a combination of offer and angle. The offer is certainly proven but has been smashed in Brazil already.
I would advise bidding CPM on most mobile DSPs... CPC guarantees you clicks for your $$ but I think it's rare for CPC bidding to be more cost effective, especially in countries with low CPMs where a $0.01 minimum CPC bid might be twice what you'd get per click on CPM, even with average ads.
Hi. Tried to translate the offer banner to Portuguese and didn't turn up well too. May I know when to bid CPC? When the offer payout is higher?
Now I am researching other offer to run. Gonna talk to my AM and keep things running. Will keep you guys updated. It's gonna be a long learning process!
So most of the placements have spent 3x the payout without conversions? How many placements is the data spread across?
Right now, 400 clicks and 1 conversion to me says your angle is not working well but I would also advise testing things other than a browser/AV/utility app in Brazil.
Whatever you test, consider testing a lander straight away as well.
Re: CPI - it means cost-per-install which is where you pay per install event. You can only do this if you have the necessary conversion tracking integrated into the apps back end, which you won't have. Incent traffic is where users are rewarded for clicking and converting/installing. This traffic is very low quality for advertisers and you should send this to offers only when they have explicitly said that they allow incent.
Home > Mobile > Follow-along Campaigns