<img height="1" width="1" style="display:none;" alt="" src="https://dc.ads.linkedin.com/collect/?pid=88527&amp;fmt=gif">

Google Ads Data Hub

Analytics By: Luke EllisJan 15, 2021
Why Google’s Ads Data Hub (ADH) could be the answer to advertisers’ attribution and targeting prayers

With the media industry’s continuing focus on user privacy, Ads Data Hub (ADH) ­­is Google’s answer to privacy-safe attribution and targeting. In Google’s words, it’s a data warehouse used to perform customised analysis to better direct advertising towards specific business objectives.

ADH supplies a clean room environment in which advertisers can continue to use their own modelling capabilities and data science within Google’s walled garden. This means that none of Google’s data can be exported out, but advertisers are potentially able to access more Google data for attribution and measurement than previously. Google says the tool can help advertisers unlock insights and increase effective optimisations, leading to better advertising efficiency.

Why is it important in the industry?

In the past, advertisers built their attribution strategies around the DoubleClick ID, Google’s cookie-based identifier that ties together user information across its properties. Previously, you could upload raw log data for each click, view, or event, including user ID, from Campaign Manager 360 (CM360) to Google BigQuery.

When GDPR came into effect, Google limited the use of the DoubleClick ID effectively breaking independent attribution for advertisers in Europe.

ADH is the workaround, future proofing the analysis of data in a compliant way. It allows advertisers to transparently evaluate the effectiveness of display advertising whilst making sure they stick to all privacy requirements. User IDs can no longer be downloaded from CM360 but they are available at Ads Data Hub. So, this is now the only way to complete user level analysis and allows bespoke analysis on Google data.

Why should advertisers get excited?

The bespoke analysis carried out in ADH can help advertisers understand their audiences better. But the most exciting aspect is in the activation side. For example, when you’ve found something interesting, you can use the analysis you conducted to build audiences powered by the user ID, then activate that in Display & Video 360 (DV360) either through targeting or suppression. So ADH is effectively taking the role of a hybrid DMP / CDP.

When comparing audience building in ADH and GA360, ADH is faster and more robust as it includes more complex filters and constraints. ADH also requires only a single SQL query, whereas GA might need multiple teams to set up changes.

When reporting on ad campaign data, the SQL request quickly provides up-to-date information without needing to store ad data in your project. Updating is also an automatic feature, with all collected data available for reports.

ADH also allows more flexibility in the look back window, so you can go beyond the standard reporting tool look back of ninety days. 

How does it work?

Ads Data Hub is an API in Google BigQuery. It isn’t an independent data storage, but it links two BigQuery projects, yours and Google’s. ADH uses a data clean room to join the log level data using the user id, allowing for complex analysis.

You can configure Ads Data Hub to key event-level data and to a unique user identifier (from the advertiser, passed in a privacy safe way). Both of these are sent to Google when users interact with your ads. After working with Google to set this up, you can write queries that join data in the Google-owned Cloud project with the data that you store in your Cloud project.

Your data remains in your Cloud project throughout the entire process as Google only operates on the data that you explicitly reference in your SQL queries. Joins are privacy-centric and are keyed to a unique user identifier of your choosing. This ensures that end-user data remains secure.

Results from the queries you run using Ads Data Hub are written to BigQuery datasets in a Cloud project that you own. You can combine your data within your Cloud project with Google’s data to evaluate the efficacy of your ads.
This allows you to bring in your own data, whether that is in or out of the Google ecosystem, such as CRM data or a list of emails. This helps you better understand your first party data by comparing it to the Google ID, to determine where you are getting unique reach. It also allows you to link all clicks and conversions to impression data and see how each campaign has affected the conversion.

What are the complications for advertisers?

One complication is the query results given by Ads Data Hub are aggregated to 50 users, which means that each table row must contain data about 50 or more users. You cannot drill down to one specific user. This restriction was created to comply with data privacy standards. From an analysis standpoint this is sufficient, it doesn't allow you to measure things like customer lifetime value which require a one-to-one relationship. 

Another restriction is ADH only starts collecting data once you set it up for the advertiser. So, you have to build your database over a time period. Also, you can’t run queries that are similar consecutively in short succession. So, you can’t run a frequency analysis over 10 days, then run the same analysis over 9 days.

A key restriction for digital marketers is the lack of an intuitive UI, you can’t just log into ADH and play around. You need an analyst or data scientist to write the SQL. Google has introduced templates with standard queries, and an easier to use operating system may be on the plan. However, the hardest bit will always be understanding what you want to analyse, what you want out of it, and the limitations. For example, there is currently no Google Analytics or Search Ads 360 (SA360) data available. The work around for this is complex and would a be problem for most advertisers.

How do advertisers overcome the limitations?

To overcome these limitations, advertisers will need to cultivate a close relationship with the Google cloud team, or work with an agency that has established this. We have extensive experience with this technology only because we built our proprietary DSP (since deprecated) on the same infrastructure, so gained an understanding how to store, use and visualise the data. We have needed this knowledge and our relationship with the Google Cloud Team to overcome a number of complications when using ADH for clients.

What’s been done so far?

Over the last few months, we’ve been getting a better understanding of what’s possible and working with clients to generate new and exciting insights.

We began by linking all our advertiser accounts to ADH to store three bits of analysis for each: time to conversion from impression, time between conversions, and cross campaign analysis.

We followed this up by working with a large retail client to find the path to conversion for its customers. Specifically we are looking at the interactions between search and display to find out if users are more likely to convert if they see a search ad or a display ad and how this changes at retail category level. Based on these results we are learning how best to change the campaign set up and the creatives shown.

We have recently, started pushing audiences directly into DV360 from ADH. This began with remarketing audiences, such as one time and low frequency converters. And now, we are testing ways in which we can use ADH to replace the current Data Transfer files from Google, as these are due to be withdrawn sometime this year.

What is the future of its use?

The most advanced campaigns on Google Marketing Platform will soon only be achievable through an extensive use of ADH and GCP.

This could include:

-  Custom Bidding optimisation powered by ADH scripts.
-  Bespoke analysis, eg. cross-channel
-  Pushing audiences directly into DV360 or custom algorithm

It’s possible we might even see incrementality results showing the percentage of conversions that were received as a direct result of a campaign. However, due to the technical nature of these tools, most advertisers will need a specialist to help maximise their use.

Becoming more reliant on the Google ecosystem and not being able to do independent attribution or measurement has disadvantages, especially when trying to understand the full media plan. But the value of being able to connect marketing activity to Google’s universe of data sets and business outcomes far surpasses the value we were able to get previously from the user-level log files. Advertisers now need to learn how to use the new tools and adapt their strategies to the new insights and principles.  

Having said that, it's worth bearing in mind that ADH is not immune to industry changes, but by using it advertisers are taking an important step toward where analytics is heading.

If you would like to get some help with targeting and attribution, please feel free to get in contact

Get monthly programmatic insights into your inbox