Introducing SimulateIDFA, The Solution For iOS10 Ad Tracking
IDFA is the abbreviation for identifier for advertisers on iPhones.
An Apple IDFA is somewhat analogous to an advertising cookie, in that it enables an advertiser to understand that a user of a particular phone has taken an action like a click or an app install. Also IDFA could identify the reach of ads, preventing the duplication of the same ad on the same user.
For the advertising industry, IDFA is the key attribution which helps the confirmation of ID for advertising network, DSP and DMP.
Changes to Apple’s mobile operating system, iOS 10, mean consumers will see less relevant advertising, thus depriving brands of valuable attribution data. When users opt to “limit ad traffic,” their IDFA will simply be represented as “00000000-0000-0000-0000-000000000000.” Ad tech companies will have no way of identifying those users — even when their only purpose of doing so was to ensure they weren’t serving them too many of the same ads (known as “frequency capping”), measuring whether ads led to sales, or preventing advertising fraud.
Actually the changes is the action of user’s and Apple’s regaining the control for their devices. Or,at least they could decide to view the ads or not.
According to a market research, currently 17% of the Apple iOS 10 users activate LAT. And as far as concerned, Apple owns many users which means 17% of them is not a small number.
Without IDFA, what are the new solutions?
IDFA is invented by Apple aiming to identify user’s phones for advertisers. It works as the Cookie but even better in iOS system. If IDFA is limited in such closed system, advertisers just get blind of the official data. Thus many ad tech vendors attempt to find the new tracking methods for advertisers in proper ways.
Methods replacing IDFA should have the following features:
1.Accurate & Long-term Tracking
IDFA provides an accurate and long-term tool for advertisers to identify users’ devices. When IDFA in limited, alternative solutions should meet these basic requirements.
Before iOS 10, Apple has limited the function of tracking iOS devices ID. It was when the inventor of OpenUDID developed a solution which was called OpenIDFA.
However, OpenIDFA could no longer identify accurately when users in China were hundreds of millions who has many same numbers. And value from OpenIDFA changed everyday making it impossible for long-term tracking. Since its launching, Open IDFA hasn’t updated for 3 years and could support nowadays needs.
The role of IDFA is not only for advertisers, but also for DSPs, DMPs and the third-party tracking services. Beside for apps, alternative solution should also be adaptable with other relevant platforms.
Therefore, solutions that failed to win the recognition of the industry could not replace IDFA, such as IDFV. IDFV is the the abbreviation of Identifier For Vendor, which helps developers to identify the devices. It could identify all the applications for the same developer account in the same device. But it could not identify different apps for different developers. Secondly IDFV was invented for developers which has no use for advertisers or other people. Thirdly, if user delete all the apps or reset, restart the device, the data would change accordingly.
In other words, IDFV is more adaptable for analyzing users’ in-app behavior. But when we talk about ads promotion, cross-app tracking, IDFA is better.
To preventing cheating is one of the important function of IDFA. Besides IDFA, it’s commonly suggested to use IP+UA, especially for the third-party because they may not have SDK to obtain IDFA.
But there are limitations in this solution. 1. IP address may be repeated. There could be millions of users in a same IP. Plus data of personal IP changes a lot. 2. Making a fake IP is not expensive which makes it easy to cheat. For performance based advertisers, the IP+UA has its limitations. App advertisers may not consider this one.
4. Recognized by the industry
In spite of methods that seems working, there have no other methods as perfect as IDFA.
When it comes to the choice of method, it should combines the opinion of advertisers, networks and third-party tracking system, forming a data chain is acceptable to all this parties. Only gaining the recognition of the industry, will the new alternative solution worked.
Since Apple’s absolute authority in the industry, other players have to adapt themselves according to Apple’s policies except giants owning completed account system such as Baidu, Alibaba and Tecent.
Different from Android’s open source which allows various tracking methods, tracking ads on iOS devices without IDFA may still have no perfect solution.
SimulateIDFA, New Way To Identify Devices
As mentioned above, if IDFA is limited, advertisers in iOS are searching for new method. Hence to realize goals of the advertisers, methods should equip with the above features.
That’s why Adxmi’s parent company — Youmi Technology is developing SimulateIDFA to help developer, advertisers and DSP platforms to identify devices in iOS and optimizing the performance and efficiency of their ads.
Inspired by OpenIDFA, SimulateIDFA develop a whole new identify solution which is more advanced than OpenIDFA and more adaptable in iOS system. In essence, SimulateIDFA is a new method abiding Apple developer’s policy and respecting user’s privacy.
Technically, here are the main features of SimulateIDFA:
1. Easy to operate
Data embed with the algorithm in App or SDK of SimulateIDFA would be as same as IDFA.The principle of generating is similar to OpenIDFA. MD5 value is divided into two parts,consisting of 32bits value.
The former part of the value only changes when system is upgrading. It contains information of system, devices, storage, coreServices and files updates,etc.
The latter part consists of adaptable values which only change when devices restart. It contains information of the time when system activated, country, language and device’s name.
2. Low repetition rate, High efficiency & Anti-fraud
For generation algorithm, repetition rate of SimulateIDFA in every ten million devices is one in seven trillion. Compared to OpenIDFA , the repetition rate is one in each ten million. In some extreme conditions, the rate of repetition is relatively high because the OpenIDFA algorithm involves the day time. Then the value of MD5 will change, making the value mutable every day.
But value in SimulateIDFA divides into two parts,. The former value of 16 bits changes when system upgrades. The latter part of the value changes according to user’s action, such as restarting the phone, modifying the device name, modifying the phone’s native language. Hence SimulateIDFA performs better than OpenIDFA in long-term checking.
For anti-fraud, SimulateIDFA equips with advanced value source and algorithm, which makes cheating spending more time and money, failing to modify value in a large scale and efficiently.
3. Recognized by the industry
Currently, many app advertisers have marked SimulateIDFA. As a new solution for iOS advertisers, this technology is extracted and improved from our platform’s daily anti-fraud techniques. Owning to its premier testing and launching in Github granted by MIT, SimulateIDFA has refined adaptability.
The development of SimulateIDFA enhances the credibility for the solutions, making the data more transparent. Also, we hope to create a more open mind among developers, building a new standard for China’s mobile advertising market.
“We are not assuming that SimulateIDFA would be accepted by the industry in a short period. What we do is to open a window for the rather closed industry in China, innovating and redefining technology the for mobile advertising.”says our technology team, “The open source project is accepted by advertisers, giving a chance to solve the problem of user tracking and data standard together.”
Apple may change its policy. The open source SimulateIDFA should also keep on track, making adjustment at the first time. From our point’s of view, data of devices from IDFA is the fundamental for mobile advertising . As developers of this industry, more emphasis should be laid on data analysis, mobile anti-fraud and ads optimization.
SimulateIDFA has been tested in some of our cases.
To see more about SimulateIDFA: https://github.com/youmi/SimulateIDFA