Conversions, orders, and revenue reported by media partners like Facebook, Google, and others are often contradicting and not a solid basis for data-driven decision-making. The reported data is usually incomplete, inaccurate, and has significant gaps. Here’s why:

1. The tracking implemented by media partners only sees a small part of the user journey.

2. Privacy features from operating systems and browsers heavily limit the tracking of media platforms and apps (e.g., Facebook conversion pixels in iOS 14).

3. Attribution numbers become inflated because media platforms run their dedicated attribution models that are generally not comparable across platforms. Some platforms consider ad views (not clicks) as fully adequate touchpoints. Thus the numbers cannot possibly add up: When summing up the reported orders of each media platform you are working with, you usually end up with way more sales than you had in your store.

4. There is an inherent conflict of interest in media platforms reporting attribution numbers as they indirectly impact their profit. Some media platforms like Facebook even model conversion numbers, meaning the reported data is not measured but calculated through non-public models that are a black box to advertisers. 

Admetrics Data Studio is an all-in-one marketing data warehouse, user journey tracking, and attribution solution. The reports paint a solid picture of how your paid and non-paid marketing activities contribute to your growth.

The fully privacy-compliant user journey tracking hooks into your existing consent mechanism and doesn't require any coding skills to set up. 

Once the user journeys are tracked, Admetrics proprietary attribution models are able to accurately attribute the sales to each touchpoint, channel, and campaign.

The orders and revenue numbers reported by Admetrics will generally be lower than those reported by media platforms like Google or Facebook.

This is mainly because:

1. Admetrics Data Studio uses click-through attribution, which only logs a touchpoint when an ad is actually clicked. Many media platforms like Google and Facebook additionally use view-through attribution, which attributes orders to ads that were displayed to users (but never clicked).

2. Admetrics Data Studio sees the complete user journey and is a single truth source. Our attribution models are not attributing orders to multiple ads - instead, the order is fairly attributed to a single or multiple touchpoints depending on the selected attribution model.

An example: When a user clicks a Facebook ad, then a Google ad, and ultimately makes a purchase, Facebook and Google will both claim/attribute the conversion. As Admetrics knows all of the touchpoints of the user's journey, our attribution models can attribute the order correctly. In the case of last-touch attribution, we would attribute it to the respective Google campaign; in the case of multi-touch attribution, both would get a portion of the revenue attributed.

We strongly believe that an experimentation-focused approach is key to sustaining a successful marketing organization. Marketing teams that master testing and learning can outlearn their competition by delivering better products and more relevant ads and experiences and ultimately drive growth. Transforming your team to have an experimental mindset and building a culture of experimentation is the first step.

If you are interested in more details about building a culture of experimentation, check out this article:
How to Establish a Data-Driven Culture of Experimentation and Supercharge Growth

Always-on experimentation is an agile process that enables companies to iterate, optimize, and innovate faster by continually learning from ever-running marketing experiments. Implementing always-on experimentation can tremendously impact how businesses conduct their marketing practices and drive growth.

Learn more about how agile marketing teams leverage always-on experimentation in this article:
How Winning Teams Leverage Always-on Experimentation

No, even if you currently do not run marketing experiments, you can still leverage Quantify to make better data-driven decisions.

Quantify will help you understand the credibility of your performance data and make better decisions faster while eliminating guesswork and the dependency on data scientists.

With all the possibilities that Quantify provides, there are still common-sense limits. For instance, comparing click-through rates of large advertisements with those of text links might not make sense. While Quantify will deliver results in this case, the usefulness of these results will be limited.

• Complicated to setup
• No test peeking
• Data hungry
• Limited amount of variations in each test
• Results are abstract and indirect
• Dependencies on expert resources

Learn in this article why you should move away from A/B significance testing:
7 Reasons to Move Away From A/B Significance Testing

Quantify uses Bayesian statistics, which is in many ways superior to other approaches. Quantify models probability distributions for all metrics and calculates credible intervals. A credible interval specifies the range that includes 95% of all probable values. Quantify also provides credible intervals for actual lifts, so it can tell you how much better a variation performs compared to other variations. These tests can be run both on-off or continuous, even while changing the test's underlying variations during the process.

Quantify will automatically draw meaningful conclusions and highlight winners and losers as soon as the results are credible enough.

We explain the basics of Bayesian statistics in this video: Watch Video.

If you are interested in our Bayesian statistics engine's mathematical theory and methodology, please check out our whitepaper: Download Whitepaper.

Admetrics Data Studio provides a growing number of integrations with selected media partners. If your data source is unavailable now, please reach out.