What is Snap Advanced Conversions?
Snap’s Advanced Conversions is a data collection, analysis, and reporting process that provides quality advertising services and conversions reports while respecting user privacy and the increasing restrictions around data tracking. Its strategy and how it works make it especially promising compared to other solutions (or the lack thereof) currently offered up by platforms that previously relied on third-party data for advertising.
Advanced Conversions was officially launched in 2021, but its features and capabilities are being regularly developed and updated. Despite its infancy, it has already gained some controversy and is often labeled as Snap’s method of “bypassing” iOS data tracking opt-outs. The notion that Snap Advanced Conversions “skirts around” data opt-out and tracking rules is found throughout recent media, but it doesn’t paint a very accurate complete picture of this new feature and what it actually does.
For investors, advertisers and users alike, it’s important to get an idea of how Advanced Conversions works. This article is a good starting point if you are one of the above, and will cover:
- A simple summary of what Snap Advanced Conversions is
- A straightforward explanation of how Advanced Conversions works
- Important upcoming developments and updates
- A deep dive into why these developments are important to keep an eye on
How Advanced Conversion Works, in Snap’s Own Words:
Snap explains Advanced Conversions in three steps:
- Obfuscating impression and conversion data so it can no longer be traced to an identifiable account or user
- Combining this data together and sorting them into groups that are useful for ad optimization and reporting using cohort analysis
- Repeat steps 1 and 2 on a daily basis. Reports are delayed by 36 hours to prevent real-time attacks and make the results even less traceable to any particular user or account.
It sounds straightforward, but to make things more clear, let’s break those three steps down further in a more simple way.
How Snap Advanced Conversion Works, Complete Breakdown:
The best way to understand how Advanced Conversion actually works is to compare it with traditional ad optimization and reporting to understand what it’s attempting to avoid and solve. The goal of every digital advertising platform is to provide businesses with ad placements that result in clicks, conversions, and sales. Data is key to doing this.
So now, the question for digital platforms like Snap is: How can we give results to businesses only data can bring, but without the invasion of privacy that ad optimization and reporting previously relied on upon?
Step 1: Make personally identifiable data from Snap and the advertiser unusable
Snap Advanced Conversions solves this by taking the focus away from exact, user-identifiable data and focusing instead on the meat and potatoes of what advertisers were looking for in the first place: groups and patterns in data that are useful for insights so they’re not showing ads to random uninterested people, they know where their ads are going, and what gets clicks.
To do so, Advanced Conversions relies on two sets of data: Impressions data (first-party data from Snap) and Conversions data (first-party data from the advertiser). This data includes identifiable information by nature — most first-party data does. However, Snap obfuscates both of these datasets separately so that:
a) Impressions data does not give the advertiser any identifiable information of Snap users,
b) Conversions data does not give Snap any identifiable information of the advertiser’s conversions and customers, and
c) Identifiable information from both sets cannot be combined to use in ad optimization and reports, or be accessible to any outside companies.
What does “obfuscate data” mean, and how does it ensure personal privacy?
The user data that makes you identifiable as a person or by your device are things like email addresses, phone numbers, and IDFAs (i.e. a tracking number for Apple iOS devices specifically for advertising). Obfuscating data means the process makes these pieces of information unintelligible and unusable so that they are not joined or used in any way in the next step.
It’s the data equivalent of black bars redacting sensitive details on a document before handing it over to someone else. This obfuscation prevents Snap from ever learning what its users are doing on other sites or apps, and the same for advertisers. Because the data sets are obfuscated before analysis is ever done, potential bad actors can’t simply “decode” this obfuscated data in analyses and reports either, as the information was never a part of the equation when making them and was not obfuscated as one unified dataset.
Step 2: Combine the remaining information to find useful patterns, groups, and traits
The now-obfuscated datasets are now combined and organized into different groups with common characteristics. These common characteristics are extremely important and helpful in creating the same conversion reports that advertisers and businesses are familiar with, and can reliably give advertisers insight into how well their campaign performed.
It can’t and won’t, however, give them any insights at a user level. In other words, it’s all aggregate data that paints a good picture of the advertisers’ audience on Snapchat for important campaign metrics without ever handing out any names.
You can think of it as a buyer persona — it’s accurate, detailed, based on heaps of information about real people, and you can use it to get real results when trying to reach them. You can even compare it to the customers you know firsthand, but you can’t pick it apart to figure out the specific people behind all these informative insights.
Step 3: Continue this process every day, with reports delivered to the advertiser 36 hours after the fact.
This delay is purely for the sake of privacy, making it harder to try and draw connections between real-time user activity and insights in conversions reports. The delay is also a security measure against any outside entities attempting to do the same thing if attacks were to occur.
Current Limitations of Snap Advanced Conversions
Advanced Conversions was updated a few months after its 2021 launch to allow advertisers to see a small portion of real-time analysis, but it takes 36 hours to receive the entire report from the Advanced Conversions process. This 36-hour timeframe isn’t entirely accurate though, as Snap recommends waiting ~72 hours before drawing any conclusions.
This is the system’s biggest drawback for advertisers as of now — the delay is not only frustrating but also leads to confusion for those not used to it.
For example, the delay can cause CPX (any “cost per ___” metric) to initially look much higher than what the advertiser’s max bid was even after the full 36-hour delay. The real metrics become more accurate after a few days of daily analysis and full reports, but any advertiser that doesn’t read this note on Snap’s information on Advanced Conversions might reduce their max bid in reaction and hurt their own campaign in the process.
If the advertiser found any relief in the lowering CPX metrics a few days later, they might be a bit disappointed to see that their returns lowered with it as well — ROAS also suffer the same initial over-projection for the first few days and become more accurate in the same 72-hour timeframe.
This also makes any A/B testing difficult, at least if you’re trying to do so in a timely manner. You can’t rely on any estimates or modeling figures either; the system currently does not provide any.
All of the above is why Snap quickly decided to include a snippet of the real-time analysis for advertisers while they wait for full reports to kick in. Still, it’s hard to ask advertisers accustomed to real-time reports from third-party data to be patient, especially when platforms with tons of personal first-party data don’t have to ask them to.
Upcoming Updates and Developments
Snap keeps their description of upcoming developments fairly vague: Advertisers should expect a “more holistic, real-time view” of results — but no word on exactly when other than early 2022, and they leave this note with “More information to come.”
What we do know is that they’re intending on addressing Advanced Conversions’ one major flaw of delays, and we know roughly how they plan on doing that: conversion modeling, to provide those estimates the process currently doesn’t have.
Though Advanced Conversions is a novel process of reporting that is far from perfect now, it has a good track record of delivering on promised launches and updates thus far. If we use that credibility to make any guesses of this near-future update, it’s safe to say that whatever month coming up the update is released for Advanced Conversions, it will be exactly what is promised; the solution to the system’s few growing pains, and perhaps Snap’s next big win.
Why You Should Keep An Eye on Snap’s Developments
Traditional advertising on most digital platforms relies heavily on third-party data in order to find the most suitable audiences and track important aspects of campaign performance like impressions, conversions, and click-through rate (CTR). This caused issues for many major platforms as the usage of cookies and other data tracking methods began to decline. It’s such a pillar of discussion in the digital economy, you’d have a hard time not reading about it in business media.
The lack of sufficient data for advertising due to opt-out policies like the iOS14 update resulted in losses in ad revenue, plummeting share values, and general frustration from the businesses relying on these platforms for their ad campaigns.
The consensus is to abandon its usage completely in favor of first-party data, but this is yet another problem: It’s nowhere near as widely available.
Some platforms automatically have plenty, like Amazon or TikTok, due to how the platforms work and the permissions given by users in order to access them. However, a typical social media platform does not. In other words, there is no way for companies like Meta or Snap to provide the same quality of advertising by simply swapping third-party data out with first-party data in their traditional reporting and optimization systems.
They can’t rely on the third-party data they already have either. Personal preferences, interests, and even basic information like emails are subject to change, which makes most third-party data unreliable unless it’s fresh and regularly updated.
For companies in this predicament, there is no standard alternative solution in place to save their advertising, a.k.a their main source of revenue and what makes the platforms free to use. As a result, most are opting (read: scrambling) to internally develop their own solutions according to their resources and requirements.
Due to the nature of first-party data and the changing social media landscape, it’s unlikely that there will be one widely accepted method for ad optimization and reporting for quite some time, if ever. But whichever company manages to figure out the best way to provide the reliable advertising that businesses grew accustomed to will win big as we approach the inevitable death of third-party data advertising — so long as advertisers and investors aren’t waiting for too long, that is.
According to Snap, they won’t have to; unless they insist on waiting for the metaverse.