Brands: Take control of your advertising data!

With the coming demise of cookies and other online identifiers, there has been a lot of interest surrounding the increasing importance of first-party data for brands. Recognizing its strategic importance, brands are going to great lengths to increase their first-party data: Molson Coors recently released a direct to consumer zero-proof canned cocktail with the goal of increasing their first-party data. However, many brands, especially those who sell through retailers, may be overlooking an opportunity to capitalize on their existing relationships with consumers that is sitting right in front of them, which is through their advertising. Many of us have a cherished relationship with brands through our favorite ads. (My personal favorite is the Nike Fenway Park commercial which was even written about in the New York Times.) Despite the growing importance of first-party data and this existing relationship with consumers, many brands fail to take ownership of data generated by advertising and instead hand over the responsibility of advertising data to their agencies, their publishers, and their technology providers. Brands should continue building their first-party data strategies, and continue partnerships that can bring them valuable second-party data, but should capitalize on the opportunity of advertising data today.

What advertising data is available?

When we think about first-party data for brands, we often think of audience data. But building an audience is merely the first step in the advertising process; to fully capitalize on the opportunity presented by advertising data, brands should collect data generated throughout the entire process of running a modern ad campaign. For instance, during creative development, many different creatives and messages can be produced and pre-tested for consideration. Or, during planning, overlaps of the audience with a variety of publishers can be examined as a way to evaluate which partners most effectively reach the campaign’s desired consumers. When the time comes to run the campaign, that is when the scale of most campaigns leads to vast amounts of data for brands.

Let’s break this down a bit and start by looking at two critical systems: your ad server and your Demand Side Platform (DSP). If you are using a modern ad server with Dynamic Creative Optimization like Innovid, you should make sure to get data directly from your ad server since it will contain important information about the creative shown, far beyond just your traditional creative ID. Your DSP is important, because it surfaces all of the details about the auctions you participated in, both the wins and the losses. With your DSP, you should be able to get event level data, allowing you to garner critical insights about the performance of your campaigns. For instance, The TradeDesk has made it easy to get a stream of raw events through REDS (Raw Event Data Stream), which streams RTB activities at the log level for use by customers who operate their own analytics reporting systems. If you work with a publisher that uses neither your ad server nor your DSP, then brands should use their agency relationships to get as much data as possible or alternatively, it may be simpler to place an ad tag, if tagging is allowed.

The level of data, of course, varies significantly by platform. Some publishers, especially walled gardens, will typically only make aggregated data available, although some allow custom queries through a Data Clean Room. Some ad servers make logs available with identifiers stripped out, while others will make them available, only when the publisher allows it. The type and level of data available should be an important consideration for brands when choosing a tech stack and when planning campaigns. In general, brands should strive to get impression level data since it offers the most flexibility, such as being used for reach and frequency or multi-touch attribution.

Data is no longer the domain of digital advertising alone. TV, even Linear TV, has joined the data revolution. Companies like VideoAmp, Inscape, and iSpot have a wealth of data about when, where, and during what shows your ads are shown. This is in addition to more traditional TV data from the likes of Comscore and Nielsen.

Brands use measurement and verification companies of all types. Viewability data, brand safety data, and fraud data are all available, and can add valuable context to the performance of your campaign. Multi-touch Attribution (MTA) and sales lift can measure outcomes that result in an immediate sale, while further up-funnel advertising can be measured with attention metrics, brand lift, location lift, and search lift. All of these produce valuable data, and the granularity of data and the ease of storing it should be considered when choosing a measurement provider.

With the rise of data regulations, many companies are less willing to share data, especially person level data. There are many approaches to still getting data while reducing risk for the providing company. For example, they might remove identifiers like IP addresses and replace them with an approximate location. Many companies are moving data exchanges into Data Clean Rooms, such as those powered by Snowflake, allowing consumers to make custom queries of the data while only taking out aggregate insights without exposing granular PII.

What value can you get from the data?

Owning your advertising data allows you to join with other data in your first-party ecosystem. Breaking down data silos and bringing different types of data together unlocks new insights and use cases. Below I explore a few use cases to illustrate this data can be used to drive value within brands.

Some campaigns are intended to drive upper funnel metrics like consideration, but that does not mean brands don’t care about how the campaign translates to sales. Rather, the sales are borne out over the long-term, beyond the end of the campaign. Tracking upper funnel metrics over the long-term and tying those results to sales can help a brand understand what upper funnel metrics are most important for them.

In most organizations, creative and messaging performance between paid and earned or owned channels are not unified. This is a missed opportunity — understanding how consumers react to paid messages can give insight to engagement on your website. In addition, combining the performance of specific creatives and messages with social listening data can give data-driven creative teams valuable insights into creative development for future campaigns and organic social engagement alike.

Advertising is a major expense of brands, with companies averaging 11% of revenue for their marketing budget. As such, when CFOs need to cut costs, as they are looking to do now, marketing budgets need to get a close look. Unifying the data and making it available through a BI tool, ideally one the finance team is already using, can make it easy to find opportunities for savings. Understanding which channels have too much fraud, which channels are not adding incremental reach, and which channels are too expensive on a price-per-performance basis can help a business react when cuts need to be made.

How do you get started?

So if all of this sounds good, and you’re ready to get started in unifying your advertising data and realizing value from it. But how do you get started? The technical capabilities to do these types of projects internally are always in high demand, so you need a concrete plan that you have confidence can be achieved. Let’s walk through a step-by-step recommendation that allows you to execute on this vision:

1. Brands should select a data platform to serve as the foundation for the system you are putting together. The system should be scalable enough to handle data at the impression level, not just at the aggregate level. It should make it easy to ingest data from a multitude of sources, leveraging a robust partner ecosystem for ETL/ELT, a data marketplace, integration with clean rooms, and ideally, direct data sharing with your partners that eliminates the need for complicated data pipelines.
2. Identify what use cases are priorities that must be addressed first. From the wide variety of benefits of having this data, it is best to choose the one that is the most acute for your brand to tackle first, and that drives the most value for your organization Solving this will give credibility to the effort and prove the value of the data to your organization, allowing you to build the right foundation to scale forward. Once you pick the use case, you will identify what data is required to build it. It is important to note that the most successful organizations identify the first use-case jointly across marketing teams and their technical counterparts; organizational alignment is a catalyst for success.
3. Create an exhaustive list of the partners that power your advertising stack. This list should include your ad server, your DSP, important publishers, measurement providers, verification providers and much more. At this point, alignment with your agency becomes critical for success. Not only will they be hands-on-keyboard with your stack, they can also help you get better “data terms.” For the same reason your agency gets you better media prices through their buying power, they will be able to get partners to consider exposing more of their data.
4. Decide how to get data from each partner into the data platform of choice. To cover a lot of your ad stack quickly for aggregated data, a tool like those from Adverity, Funnel, or Supermetrics can make it easy to get started. They will take data from your accounts on platforms like Google, Facebook, Amazon Advertising and many more and put them in your data platform. These partners have hundreds of pre-built integrations with ad platforms, saving you valuable time and allowing you to focus on your specific taxonomy and naming conventions. In addition, having to maintain integrations with the changing APIs of these platforms would drain your resources.

Conclusion

Brands will no doubt continue to focus on their first-party data strategy. Given the wealth of advertising data and the value it can provide, brands would be well-served to make sure this data is not left out.