Facebook starts offering its audience network tool to all advertisers and apps

Earlier this week, Facebook announced that they will start letting all advertisers and app developers use its Audience network tool against a fee. This means that any third-party app which host ads can now be powered by Facebook's targeting data. 

How does this work
Until now, an app developer who wanted to show ads in his app and earn money with it had two options. Option 1 was to employ a sales team and analyze its customer base and then create apps. This is a very expensive way and many apps can't use this option in its early days. Hence, option 2 was the preferred option so far, in which developers normally work with dependent ad networks who sell the ad inventory to the developer's app. These ad networks though have limited information about the identities of who they were reaching, and hence, the ads aren't targeted that well, don't resonate, and don't generate enough clicks or qualified leads to command high rates from advertisers.
This is where Facebook now comes in with its audience network. A developer is now able to install Facebook's SDK, which providers the developer immediately with an overview about each customer. The customer of the developer's app is not even required to be logged into Facebook by the time of using the new app, Facebook can work via phone ID to identify the person. Due to that, the developer will know immediately all about his customer and can adjust ads accordingly. The result is that this feature makes new apps way more attractive for advertisers and targeted advertising is possible from the very first app user onwards. 

Everyone can benefit
1) By offering this tool to all advertisers and app developers, Facebook positions itself very well. It is able to shift away from only offering ads on the Facebook page, which would eventually drive users away if the home wall is only full of ads. It additionally provides a good source for ads revenues and moves them even closer to Google.
2) The developer benefits because he is able to do offer targeted advertising from the very first customer on without the need of a sales team.
3) The customer of the new app benefits in the way that instead of being bombarded with mostly useless and ineffective ads, they might see ads which they are actually interested in.

Overall, I think Facebook did the right move in offering this tool now to everyone, as benefits are high for many stakeholders and potential downsides are very limited. 

A quick review of digital ad attribution models

Marketers are constantly seeking for the highest ad campaign performance and where to receive the highest ROI on advertising spent. As brand shift advertising spends to digital and attribution is key performance metric for media channels.  The related questions are what target strategies work? Which ads are driving conversions? Is one channel more valuable than another?

Below I will review the popular attribution models with their associated strengths and weaknesses.

1. Full Funnel Attribution

This classic model assigns values throughout the different stages of a consumer's experience, providing function into the role that each ad plays.   Consider the following stages: awareness, interest, consideration, preference and purchase.  There are two weaknesses with this model. First not all stages are equally important, awareness and closing the sales deserve the most credit. Second, each ad may serve multiple purposes such as awareness, influence brand preference and driving the closing of the sales.

2. Last-Click Attribution

This simple model gives the last advertising medium before purchase all credits for the sale.  Search market prefers this attribution because it avoids confusion. On a high level, this may seem logical. However, on a closer review the model has flaws. Specifically, the model doesn't account for previous advertisement have motivated the consumer to purchase the product in advance of the last click. The model fails to account for the process long term product awareness and interest or the evaluation that consumers go through.

3. Post-View Attribution

This model credits the last advertising channel that shows ad to the consumer who makes a purchase. The strength of this model is enable marketers to measure if viewing the ad actually resulted in sale. The weakness of this model is the similar to "Last-Click" model, as it encourages media partners to widespread to take credit for the conversion, even though a consumer may not have seen the ad.

4. Equal Attribution

An extension to the post-view model is where equal value assigned to every single ad placement. Branding campaigns are typically more likely to utilize an equal attribution model.  The weakness of this approach is assumed the quality of ads is created equal.

5. Fractional Attribution

The most sophisticated model which measures fraction attribution which shows duplication is occurring between channels, which help brands remove repetitive media who are not adding value.  Research shows that first impression is more important than subsequent impressions. Thus if an ad provider can deliver low duplication and strong first impression, then the ad has a higher value because it initiates consumer's awareness and play a important role for the conversion.

In my opinion, fraction attribution is the best model available and as close as we can get to practice full marketing funnel attribution.

The complexity of attribution lies in companies' varied objectives and targeted audience. Consequently, attribution models will continue to evolve. Certainly, marketer’s idea will shape the future models of attribution and smart people will help to better measure attribution of ad channels.

http://bit.ly/14gtg6e