7 天掌握 Mobile Marketing (Day 3)
本帖最后由 清心 于 2014-9-1 10:49 编辑本系列文章来自国外收费论坛AFF PLAYBOOK的精华教程，不敢私藏，特意分享给大家，但是因为没有时间翻译，所以英文不太好的同学可以复制到谷歌自动翻译看看
7 Days to Master Mobile Marketing (Day 3)
Day 3: Tracking Mobile Campaigns
Tracking is one of the most important elements to your campaign. Not only will it allow you to get more out of every dollar spent, it will help you find the hidden gems in your data that you might have overlook otherwise.
I will go over how to analyze your data in the next few days, for now I want to make sure your tracking is set up correctly.
Setting Up Tracking
First you need to get yourself a server. I would prefer that you use a dedicated server that can handle the quick redirects necessary to succeed on mobile. Because of the nature of mobile and the number of clicks coming through I HIGHLY advice you DO NOT run on a shared or even a VPS.
Some of you may even want to link directly using your affiliate network links. The only draw back is in most instances you will not be able to split test offers as easily, but this is a small sacrifice compared to loosing 10 to 30% of your clicks. Also, for those of you who think its important to hide your data from networks, most important parts of mobile can all be derived from User Agent and IP data of the visitor which in most cases can’t be cloaked anyway.
Whatever you decide make sure that your tracking allows for multiple subid fields. Subid fields are basically placeholders at the end of a link that allow you to track multiple parameters.
Here are your options for self hosted trackers:
CPVLabs – Great tracker with lightning quick redirects. Originally designed for PPV traffic but can be used for mobile as well.
Prosper202 – Newly updated and ten times faster than its dinosaur cousin. Is currently being optimized for mobile. Best part it’s FREE!
Bevo – No comment as I’ve never used Bevo.
Here at MobAff we are building out own custom tracker so if you have the resources you may want to do the same.
Again both trackers above, or your affiliate network will have resources on how to set up tracking effectively so I will not cover that here. Go Google it.
Most networks will have some sort of subid fields allowing you to generate a link similar to this:
http://www.myaffiliatenetwork/?offer...321321&subid1=
If you’re using your own tracking you would import the link with 1 subid field to which your tracker will append a unique transaction id each time a user clicks. When you generate your link you will use the c1, c2, c3, fields to populate notes about your campaign targeting.
For example: …&c1=JumpTap&c2=Android&c3=ATT&c4=ugly.jpg
What is great working with networks that have MACROS is you can insert a variable like JT_ADBUNDLE that will automatically replace the name of the campaign from which the click came. More on that later…
Adding Pixels
Here is another crucial step that will save you time. Make sure to add pixels to your affiliate networks. A pixel is usually a javascript, img, iframe, or postpack url that the network places on the confirmation page of the offer that lets your tracking system know there’s been a sale/lead/download. Each network is different, but ask your AM and see if they have a universal pixel option as this is the easiest and quickest Method. CPV Labs and Tracking 202 all use pixels.
We found that Server 2 Server Postback URL system is the most effective at tracking mobile conversions, as it doesn’t rely on cookies or javascript.
See that wasn’t so bad. If you’re still getting confused talk to your reps, that’s what they are there for.
Check your ad networks wiki some even have a postback system to let them know of your conversions making it easier to track conversions right in their dashboard.
Now I will show you how to use dynamic variables from the ad networks to ensure that you’re tracking things effectively.
Let me explain to you the strategy and then I will show you a few examples.
Say I was to test MeetMoi offer. Upon my first test I may want to see if there is a difference between Feature phones, iOS, and Android. To do this I might decide to create three separate campaigns with a budget of $15 each for a total of $45. To decide on budget I simply multiplied the offer payout ($3.00) by a factor of 5.
The above seems like a great test simple enough, but what if I told you with that $45 you will also get data as to which Carriers and publishers perform best?
You’d be pretty excited right… well here is how you can do it.
If I were using Jumptap I would use their dynamic macros. Here is the full list of the macros.
JT_CAMPAIGN: the name of the campaign responsible for the user’s click
JT_ADBUNDLE: the name of the ad bundle responsible for the user’s click
JT_KEYWORD: the category or keyword matched from the advertiser’s campaign
JT_REQID: the unique ID for the ad request
JT_HANDSET : the handset/device associated with the ad request
JT_PUBLISHER: the publisher ID associated with the ad request
JT_SITE: the site ID (for a given publisher) associated with the ad request
JT_QUERY: the search query passed by the publisher; if applicable
JT_OPERATOR: the carrier associated with the ad request
JT_TIMESTAMP: the time associated with the ad request
So lets say I assigned my subid variables as follows:
C1 = JT_CAMPAIGN
C2 = JT_HANDSET
C3= JT_OPERATOR
When I run my C1 report I may see data like this:
NOTE: THIS IS NOT REAL DATA JUST FOR ILLUSTRATIVE PURPOSES
Clicks CR CR
C1 1872 177 9.46%
Android 420 52 12.38%
iOS 651 41 6.30%
Feature 801 84 10.49%
Clicks CR CR
Now looking at above it looks like Android did better so you may want to focus there moving forward.
However, you were at the same time tracking carrier so you might want to pull a report for your C3 variable and you find something like this.
Clicks CR CR
C3 1332 118 8.86%
ATT 141 21 14.89%
Verizon 418 85 20.33%
Sprint 773 12 1.55%
T-Mobile 540 59 10.93%
So now with just 3 campaigns you in fact did multiple tests. You were able to see which carrier, handset, and operators perform and can now make changes to your campaigns to make it back out. You can even run the above campaign with the data nested to see the exact combinations that yield results.
Here is an example:
Clicks CR CR
C1 2943 270 9.17%
Android 420 52 12.38%
ATT 100 25 25.00%
Verizon 24 5 20.83%
Sprint 197 8 4.06%
T-Mobile 99 14 14.14%
iOS 651 41 6.30%
ATT 250 12 4.80%
Verizon 199 5 2.51%
Sprint 124 18 14.52%
T-Mobile 78 6 7.69%
Feature 801 84 10.49%
ATT 299 34 11.37%
Verizon 200 21 10.50%
Sprint 201 22 10.95%
T-Mobile 100 7 7.00%
Using your macros consistently you can roll up your data across the three campaigns and see for example if you were to analyze Carrier that ATT might outperform Sprint without having to check its performance for each individual campaign.
Think of mobile as a matrix of variables. All you’re trying to do is find the pockets of combinations that yield profitable traffic for you. Some campaigns require very tight buckets like Android > Sprint > HTC phones its really up to you and your goals. Typically the tighter you targeting the higher your ROI, but at the same time you will be sacrificing volume.
You can use this same strategy even if the network doesn’t have dynamic variables. To do this is a lot more manual since you will have to physically type in the targeting in the subid variables to get the same kind of matrix of data.
The key thing to remember is to stay consistent in how you use your macros. In other words if I assign JT_CAMPAIGN to variable 1 in JumpTap every campaign from then on should have that variable or else your data will start to look messy.
Daily To Do: Set up tracking, import your links, create links to be used for JumpTap.
页:
[1]