Is it time for a comprehensive MD exam?
Your marketing performance depends on the health, strength, speed, and agility of your Marketing Data (MD).
Data does not just inform your marketing decisions anymore, it executes them. The bar is rising fast. For many brand marketers, their survival depends on major changes to their data diet as everyone scrambles to prepare their capabilities for the unprecedented conditions impacting our marketing ecosystem, including AI, privacy, clean rooms, e-comm, retail media, fragmentation, etc. This transformation is already underway, and it is apparent that the gap between winners and losers is widening based on who captures these opportunities versus just mitigating the risks.
The hesitation, of course, is the cost. The reality is that AI, privacy, fragmentation, e-comm, etc., are unavoidable and will require massive change and investment no matter what. This may be the ideal forcing function to proactively build a marketing data foundation that serves all your marketing channels and activities with one common view of your customer, improving internal cost synergy and external customer relevancy.
Acknowledge you need help
Are you confident your data is smart enough to independently make marketing decisions for you? Does your data orchestrate coordinated and consistent actions across your fragmented marketing playing fields? Are you keeping up with your customer needs? Do you beat your competitor to opportunities?
If not…. Stop kicking the can. No more siloed solutions. Do not automate mediocracy. Be honest about our Marketing Data readiness. Invest in a healthy future. Schedule your MD Exam!
See a specialist and get a second opinion
This is not just a data assessment. Or just a marketing assessment. This is a marketing data capabilities and results assessment.
You will need your data gurus and marketing channel experts involved. You should also include an industry expert who has seen most marketing use cases, industry data sources, media channels, best practices, costly mistakes, and other situations that your internal experts may not have encountered.
This is not like going to a podiatrist for your feet, a cardiologist for your heart, or a psychologist for your mind. This is like going to a physician who specializes in preparing an athlete’s entire body for a marathon. Likewise, you don’t want someone who just specializes in digital advertising, trade promotion, or social media. You want someone who specializes in using data and analytics to deeply understand and effectively influence consumer behavior across all your marketing use cases.
If you have someone like that, get her/him on this project! If not, find someone (I know a guy)
Align on a holistic marketing scope
Your assessment team should start by aligning on all the marketing activities, channels, platforms, and use cases your marketing data needs to support. It should encompass all your marketing and customer touchpoints. Do not rush through this step. Too many organizations have data-starved marketing activities that wish someone had spent more time thinking this through.
For each marketing use case, inventory its purpose, objectives, vendors, data requirements, audience capabilities, targeting methodologies, typical performance, rough costs, chronic issues, etc. Be fairly specific with anything related to data requirements (e.g. Retailer shopper id level transaction data vs. Sales data). The will be tedious at the beginning, but it gets easier as you go when (hopefully) patterns emerge.
In addition to gathering details on your current capabilities, gather future desires and major frustrations too. This is where your best ideas, cost savings, synergies, and avoided mistakes will come from.
Example Marketing Use Case Feedback:
Assortment – “Our advertising messaging needs to better emphasize the audience choice drivers that spawned our product features and earned us category shelf space.”
Price – “I wish I knew when a shopper’s price sensitivity causes them to downgrade to a lower-price alternative, upgrade them to a better value, or leave the category.”
Trade Promotion – “We need to get better at aligning our in-store merchandising and promotions with your retail media advertising.”
Digital Advertising – “Our campaign brand audience definitions are completely different whenever I have to scale beyond addressable media to contextual or non-addressable media.”
Retail media – “We cannot support any more retailer media networks unless we find scaling efficiencies and translate learnings across retailers.”
Out of Home – “We need to find a way to systematically map your digital brand audience recipes to geographies, stores, parking lots, commuter paths, dayparts, or moments we can buy.”
Voice of Customer – “I don’t understand why our SEO/AIO marketing does not leverage what consumers are literally telling us every day.”
Factor in historic complications
If marketing wasn’t swamped enough keeping up with consumer technology, evolving media channels, and budget efficacy debates, along comes the biggest industry disruptions since the internet. Unlike the internet, which largely brought incremental capabilities, consumer privacy, AI, and e-commerce are effectively taking capabilities away and forcing radical changes in the way we do business.
These massive paradigm shifts impact your data strategy more than anything. Your data needs to be used everywhere, but the best practice is not to move your data anywhere. The best data about your customers is behind walled gardens and inside clean rooms. Consumers want to keep their behavior private, but still want you to be behaviorally relevant. AI is amazing until it isn’t, and you cannot let that ever happen. Affordable reach requires contextual targeting without always knowing the media’s content. Easy peasy.
Gallows humor aside, you need to deeply assess the new requirements, risks, and opportunities associated with each of these paradigm changes. This is not optional for anyone.
Example paradigm questions
AI – Is your data AI-ready for confident, business-ready use cases? Have you injected human domain expertise into your data so your AI can find it without requiring heavy human oversight that degrades its speed and cost advantages? (more here)
Consumer privacy – Have you audited your data collection and usage against the growing list of state-specific consumer privacy and sensitivity laws? Have you seamlessly enabled alternative data where necessary without degrading the rest?
Clean rooms – Do you have a single version of your brand audience truth that can be efficiently mapped, harmonized, and leveraged into dozens of retail, media, and agency clean rooms? (more here)
Id-less audience targeting – Is your contextual targeting and id-less DSP bidding optimized to known brand purchasing behaviors? (More here)
Retail media – Does your data strategy support sharing insights and aligning tactics across retailer and non-retailer marketing tactics?
Perform a comprehensive data exam
After you get a handle on the marketing scope and challenges, dive into your data. Do not start with the data assessment first. It is much easier to scope and honestly critique your data requirements with the marketing use cases top of mind.
Marketing data scope varies by vertical, but most have the following data types. Be sure to catalog and describe the data content for what it is and not where it came from (e.g. customer data, not Salesforce data).
People - Shoppers, consumers, households, etc.
Audiences – Segments, cohorts, etc.
Places - Retailers, merchants, stores, points-of-interest, etc.
Products –UPCs, SKUs, ASINs, brands, manufacturers, categories, ingredients, etc.
Time – Fiscal year, promotion calendars, reporting week, dayparts, etc.
Geographies – Markets, postal areas, retail trade areas, ad zones, etc.
Media – Sites, URLs, apps, TV programs, genre, content classifications, etc.
Moments – Occasions, life events, life stages, local events, etc.
Consumer Sales – Market level, retailer, store, shopper, segment, purchase panels, etc.
Merchant Sales – Shipments, direct store delivery, inventory on hand, etc.
Media consumption – Clickstream, media panel, bid logs, etc.
Voice of Customer – Product reviews, social engagement, survey, search intent, etc.
Marketing mix – Campaigns, media spend, reach, frequency, performance results, etc.
Additionally, spend extra time on the attributes, hierarchies, and taxonomies embedded in each data source. These data attributes are more important than ever in the world of hyper fragmentation and data sensitivity, where cross-silo data will increasingly need to be linked, mapped, and reported at some level of relevant aggregation.
Run some lab tests
If data is the lifeblood of marketing, you need to test it against common impediments. Each test needs to be done with marketing use case requirements in mind. For example, deterministic data use cases like audience targeting typically require high scale to be practical, but may not require super high precision relative to highly modeled or demographic proxy audiences. That said, probabilistic look-alike and act-alike audiences do not require large scale for training, but depend on highly representative data.
Common data quality tests, with examples
Accuracy – How reliable are the data elements in general? e.g. This URL content classification is publisher-supplied and often suspect.
Bias – Do the data source’s collection methods artificially limit or skew to a subset of the desired universe? e.g. This receipt data skews to deal-seeking shoppers.
Scale – Is there enough data to practically accomplish the goal or task? e.g. There is not enough stable data to do year-over-year analysis.
Permissible Use – Are there usage restrictions or cost implications with this data? e.g. This panel data cannot be used for deterministic audience targeting.
Sample – Are there enough observations to credibly act on the data? e.g. This data can only support large national brands.
Completeness – Are there too many gaps in the data to enable analysis on the absence or volume of behavior? e.g. Several of the top retailers are missing in this data so you cannot assume they do not buy your brand.
Consistency – Are the data definitions stable over time? e.g. Category definitions changed frequently, so do not use for trending analysis.
Representativity – How well does the data represent national behavior and demographics? e.g. Purchase data is missing Southeast retailers, need to account for that in look-alike modeling.
Connectivity – Does the data have the ids and codes necessary to link across internal and external data sources? e.g. Not all sales data has UPCs necessary to align with our corporate category definitions.
Chart an actionable diagnosis
At this point, you should have a tangible and impartial assessment of your data’s ability to meet current and future marketing imperatives. This will help you align, prioritize, and justify your data strategy efforts to improve speed to market, cross-marketing synergy, cost efficiency, and competitive differentiation.
Every diagnosis and treatment plan will be different, but the common marketing data treatment plans include:
Harmonize disparate data across your unavoidably fragmented marketing to enable knowledge sharing (product categories, audiences, measurement results, etc.)
Infuse human domain expertise into the marketing data to ensure AI finds and leverages it for confident automation
Maintain leverage of behaviorally based target audience definitions when mitigating the loss of addressable targeting methods
Consolidate or eliminate redundant data to fund other critical data gaps and competitive differentiators
Inject attributes, taxonomies, and other ‘connective tissue’ across your data to enable linking of increasingly disparate and asymmetrical data
Listen to your customers better by injecting the ‘why’ behind what they do into your data and resulting actions
Take control of your customer knowledge as third-party data is increasingly restricted beyond walled gardens, privacy clean rooms, and paywalls
Facilitate customer message consistency via shared audience intelligence despite increasingly fragmented marketing channels and platforms
Ultimately, this marketing data assessment should result in you defining and aligning on your marketing data vision. Your NorthStar. A destination you will never reach, but a destination that all your marketing investments should proactively strive to attain. Not a big bang, but every chance you get. This is not possible without a thoughtful assessment of where you are today and a proactive game plan on where you collectively want to be. Without it, everyone’s scattered marketing efforts will continue to be selfishly focused on their own immediate needs resulting in shiner silos that drain marketing dollars and erode your customer relationship.
Do not blow this rare opportunity to reset your market position by being faster, stronger, and healthier than your competitors.