The Days After: What’s Next After UDID?

Mark Connon | 04.02.12 | 6:48 pm

The growth and maturation of mobile advertising depends in part on the continued ability of publishers, developers and advertisers to identify devices in the mobile app environment. Hence the mountain of blog posts on the topic once it appeared that Apple was actually rejecting apps using Apple’s UDID for this purpose. Dependency on UDID relates to how device identification enables relevant ads to consumers, analytics, frequency capping, conversion tracking, and/or app authentication; with Apple’s decision, the industry needs to develop effective and enduring solutions to ensure advertiser campaign success and the resulting publisher monetization needed to fuel the continued growth of mobile advertising.

Recognizing the import of UDID in the ecosystem, Apple’s decision to reject UDIDs can appear at first glance to be a setback to the industry. Indeed, it feels like jumping into cold water—planned and fretted—but when you splash down, the water is still dang cold.

120402_UDID_Mobile_PrivacyBut the splash-down is a just a temporary shock to the system, and fortunately the industry has been considering different solutions to provide a privacy-safe, effective, enduring, and ultimately better method to adapt to Apple’s decision.

Those companies that have grown up in mobile are very familiar with the foundational reality that mobile advertising (apps in particular) does not have third-party cookies—a mainstay of online—and depends on first-party data to enable targeting based on parameters such as demographic, psychographic, and location. First-party data passed on an impression enables privacy safe targeting without the need for device identification, and is a positive element to highlight in this context.  This “silver lining” however, should not be confused with an argument that device ID solutions are not necessary. Much like cookies, device IDs can and will serve other necessary purposes that will enable the ecosystem to flourish.

At Nexage, we have considered and analyzed all of the available solutions, and what is clear is that initially there will not be a single, or even dominant, substitute for UDID on iOS. We know that exchanges will play an important role as the industry innovates and evolves in this area, and Nexage’s goal is to enable solutions that maximize the value to buyers and sellers to drive liquidity (and commerce) through this transition period, and for the long term.

Most participants (ad networks, ad servers, analytics, mediation and exchange companies) have proclaimed their support for alternatives to UDID, and there is anything but consensus.  Here are some of the solutions the industry (meaning both individual providers and industry groups) is working on or have thrown their support behind:

  1. In-advertising solutions that create a trackable ID when a user clicks through to the advertised application, which is later matched to an ID created upon application download; used primarily for conversion tracking and attribution.
  2. In-app solutions where the app download generates a cookie by launching the browser upon download—the cookie is the ID—also used for conversion tracking.
  3. Device fingerprinting where a unique device identifier is created using device attributes (for each device).
  4. Using a hashed version of the device MAC address to replace UDID.
  5. The various open source solutions such as OpenUDID and ODIN.

Any or all of these may work—meaning that the industry is poised to overcome the deprecation of UDID—but it signals solution fragmentation in the near-term as different companies and approaches vie for attention and votes.

So what are we doing here at Nexage in response to recent events? First off, we will not embrace any one solution at this time, but in fact have to adapt to the decisions of the participants in our exchange, both buyers and sellers. Along those lines, our iOS SDK will support the passing of a hashed version of the device’s MAC address, which is driven by many of our developer customers. This is not to the exclusion of other solutions that, depending on the purpose and method of collection (emphasis on privacy), are very viable solutions as well. Certainly the industry group platforms such as ODIN and OpenUDID hold much promise. The role of the exchange is in part to take fragmentation and turn that into liquidity by accommodating the demands of its buyers and sellers. We embrace that role while being an active participant in the development of sustainable solutions for device identification.

Ultimately, the industry should depend on a solution 1) where its existence is created by those with a pure interest in the privacy-safe growth of mobile advertising and 2) which is purpose built to address the requirements of the mobile advertising ecosystem. It’s tempting to conclude that this is a problem that needs to be addressed/solved by the device/OS manufacturers, yet it’s a device manufacturer’s decisions that have led us to where we are today. Apple will always be a lighting-rod on matters related to privacy—largely because of their size and influence—not necessarily their actions. Apple’s decision on how to respond to the pressures around UDID was certainly influenced by the impact their decision could have on their business as a whole. I would suggest that for as long as iAd represents $100M of their $127B in annual revenues, the weighting which the impact of their decisions has on mobile advertising will be proportionate. This is not a criticism as much as it is a simple reality.

Regardless of the chosen solution of our customers, the industry must also strive to mirror the transparency, notice, choice and opt-out capabilities that exist online today. We do think the industry inevitably needs to rally around a solution that:

  1. Does not make publishers or buyers need to stand up a technology shop to succeed.
  2. Does not create privacy risk, such that the solution is inevitably unsustainable.
  3. Does not create artificial advantage to one provider at the expense of market liquidity.
  4. Produces results.

We know it is not “if” we develop an effective and enduring solution set, it iswhen. We look forward to working with all of you as we build out this great industry.


Tags: , , , , ,

Related Articles