Provocateur:

The trend toward the democratization of marketing data demands that marketers embody the role of citizen data scientists. Marketers need to become savvier around data and analytics in general. More specifically, to maximize marketing impact, they must learn to leverage information to reach and engage consumers at every touchpoint along their cross-channel, multi-device journey. And, learn to use data from that multi-touch approach to better understand their effectiveness by audience and at a granular level.

One of the foundational elements of a marketer as citizen data scientist — some-one who can derive insights from cross-channel, multi-device data — is defining and using a marketing data taxonomy. Using a data taxonomy, marketers can better analyze the impact of each channel and tactic, track changes in performance, and make more effective decisions around optimization. A data taxonomy also will improve the quality of marketers’ data standardization

practices and business processes, as well as support their data integration framework across all inputs.

While there is no one right way to define a taxonomy — each aligns with a company’s lines of business, specialized channel teams, and campaign objectives — the configuration of a taxonomy directly impacts marketers’ ability to deliver impactful recommendations that map to a business’s unique needs.

The best taxonomies align with an organization’s specific structure and terminology, so marketers can more effectively optimize campaigns and increase ROI faster. Marketers looking to deliver powerhouse advertising and marketing should follow these four steps to build their company’s taxonomy.

1. Speak a universal language

Marketers may “own” the taxonomy, but it should derive from their organization’s shared vernacular. So, when defining your business’s taxonomy, you must gather input from all your media channel, regional, and product line teams. This should be a formal, streamlined, and repeatable process (because taxonomies will need to evolve) to engage those teams effectively — not only for gathering feedback to define the taxonomy, but also for ensuring that the standard terminology you create is then used across the business. The latter is critical for consistent analysis, reporting, and planning.

2. Go big

You’re building a marketing data “dictionary,” based on elements from across the organization, which means you need a lot of input. Compile an organized wish list of dimensions based on feedback from each channel, region, or product line manager that you need to analyze, including factors such as media planning, key performance indicators (KPIs), and any specific reporting requirements. To encourage consistency — and retain your sanity when defining and updating the taxonomy — provide each manager with a template they can use to assemble their list. Without a template you risk getting disorganized or incomplete lists of terms and dimensions, which will leave you with reams of unmanageable data to try to sort through to build your taxonomy.

3. Streamline

Using the completed wish-list templates as a guide, you may want to start grouping dimensions with similar objectives in order to ensure simplicity and focus. Otherwise, you may end up with hundreds of dimensions that make analysis much more complex. For instance, a paid search channel manager may want “keyword” as a dimension, while a display channel manager may want “placement” as a dimension. Since they are both dimensions that will be used to measure the effectiveness of a campaign, you can consolidate both terms into a dimension called “placement/keyword.” Whether that dimension is understood as a placement or keyword depends upon the channel being analyzed.

4. Apply generously

With a streamlined list of dimensions that comprise your overall taxonomy, build out specific taxonomies for each channel or region or product line. Below is an example of a simple taxonomy that a hypothetical brand uses to compare the effectiveness of various paid search campaigns and ad groups executed during the holiday season. This hypothetical brand has defined its taxonomy with the terms, “Channel,” “Providers,” “Campaigns,” and “Ad Groups.”

• Step 1: List the channel name (e.g. paid search).

• Step 2: Add the vendor or provider’s name(s) (e.g., Google AdWords).

• Step 3: List the first dimension to be tracked (in this example, the dimension could be the campaign name, such as “Holiday_Campaign.” This will allow the brand to measure the effectiveness of its holiday campaigns on Google AdWords).

• Step 4: List the ad group names (in this example, the brand groups its keywords by holiday and the first ad group is “NewYearsEve_Keywords”).

Now, the brand has a basic taxonomy that standardizes how data around its campaigns is organized — empowering the marketing team to easily compare results and optimize performance.

As a marketing leader, your goal is to maximize the impact of your team’s marketing. As a citizen data scientist, you can help your team achieve that objective by building a marketing data taxonomy — one that maps to your business’s unique needs. With a standardized approach to structuring your data, you’ll be equipped with the actionable insights you need to drive performance and deliver stronger business results. Without it, you’ll limit your ability to optimize marketing, and miss valuable opportunities to drive customer actions. 

​

To make the most of multi-touch attribution and harness the data-driven insights that sophisticated marketing demands, lay down the necessary foundation of building a taxonomy. Read more

About the Author

Visual IQ Product Marketing Manager Moira Freeman manages the go-to-market process for product releases, ensuring that product enhancements are communicated effectively to internal teams, current customers, and prospects through training and collateral.​

You can find Moira on LinkedIn.