Data overload is paralyzing brand strategy instead of improving it. With retail media networks expanding rapidly, the ability to turn massive amounts of shopper information into actionable decisions is the dividing line between market growth and wasted ad spend. Laura Weiderhaft, Director of Product at Crisp, joins the show to unpack how brands can stop hoarding metrics and start building scalable intelligence.

We sit down to dismantle the current state of retail media measurement and how teams are actually utilizing shopper data. We cover the danger of cherry-picking analytics, the OMG (Objective, Measure, Goal) framework for promotion planning, and how to properly flight budgets by Walmart week based on geographic sensitivity. Laura also reveals a highly effective philosophy on automation, explaining why marketers need to approach AI like software engineers by organizing tasks into specific "jobs to be done" rather than treating the technology like a basic search engine.

Managing all of this information is incredibly tedious, and the messy reality is that most teams aren't willing to be held accountable for the bad decisions currently hiding in their reporting. Listeners will walk away with a clear understanding of how to automate mundane data entry, the importance of aligning commercial and brand teams on a single source of truth, and why tracking secondary indicators is vital for long-term category share.

If you care about retail media execution, competitive promotion planning, and operationalizing AI, you’ll get a lot from this conversation. Please remember to subscribe and share this episode with your team to help us continue bringing practical insights to the forefront. What is the most frustrating, manual data task you are ready to hand over to an AI workflow?

More About this Episode

Mastering Shopper Data, Measurement, and Artificial Intelligence in Retail Media

We live in an era where data is being collected all the time and everywhere. Nearly every action taken online, and increasingly the actions taken offline, leads to some form of data generation. As a result, brands and marketers have access to more data than at any other point in history. However, having access to an overwhelming amount of data begs a critical question. Are we actually getting better at making decisions, or are we simply drowning in a sea of metrics?

The landscape of retail media is shifting rapidly. The way brands leverage shopper data, approach their measurement strategies, and integrate artificial intelligence will dictate who wins and who loses at the shelf, both digital and physical. To truly succeed, it is time to move beyond vanity metrics and start using data to uncover genuine shopper insights, build intentional measurement frameworks, and deploy artificial intelligence as a structured, scalable partner.

The Evolution of Shopper Data in Retail Media

For years, the gold standard of digital advertising resided within the closed ecosystems of major tech platforms. These platforms were undeniably effective due to their massive data collection capabilities. Advertisers could target specific audiences with incredible precision. Yet, the underlying mechanics often remained a black box. Brands could see that their campaigns were working, but they lacked visibility into how their audiences were evolving or what specific shopper behaviors were driving the trends.

Retail media networks have fundamentally changed this dynamic. Retailers like Walmart, Amazon, and Kroger are leading the pack by providing rich, scalable shopper data. Unlike traditional digital platforms, these retailers view brands as true partners. The goal is no longer just about optimizing ad spend. The goal is improving the overall shopper experience.

When you approach retail media, your first thought should not be about your advertising budget. It should be about the shopper. You need to understand what the shopper wants, how they make decisions, and how you can get closer to them. Elements like product assortment, pack size, item innovation, and pricing are foundational. Your retail media execution must serve as an extension of these core considerations. For example, Walmart is actively growing its market share and bringing in new shoppers every day. This presents a massive opportunity for brands to drive loyalty, but it requires a deep understanding of shopper data to execute properly.

Historically, retailers have taken different approaches to data collection. Kroger has long been considered best in class with its rewards card program, which naturally incentivized shoppers to identify themselves at checkout. Sam's Club has always relied on a membership model, ensuring a direct tie between the shopper and the transaction. Walmart has had to take a different approach, relying heavily on scale and advanced analytics to parse through massive transaction volumes, including a significant portion of cash transactions. Today, the sheer volume of data available through retail media APIs allows brands to look at store level granularity, track competitor price sensitivities, and plan inventory accordingly.

Consider a scenario where you are launching a new item, perhaps a new strawberries and cream flavor of protein powder. Instead of a broad, untargeted rollout, you can use historical shopper data to analyze similar product launches. You can identify exactly which specific stores have an audience primed for this flavor profile. This allows you to walk into a retailer with a highly informed recommendation for your new item launch, backed by scalable shopper insights.

Avoiding the Data Overload Trap

With so much data at our fingertips, it is incredibly easy to fall into a state of data overload. To navigate this, you must be absolutely clear about your business goals and identify the right path to achieve them. Are you trying to improve shelf execution and reduce out of stocks? Do you want to grow your business through trade up strategies? Or are you focused entirely on assortment expansion? Each of these goals requires tracking entirely different Key Performance Indicators.

One of the most common and damaging mistakes made by brands today is a lack of accountability to bad decisions. It is always much easier to tell a good story with data. When a campaign succeeds, everyone wants to take credit and highlight the metrics that point to a win. However, true growth requires radical honesty about what is not working. Cherrypicking data to paint a rosy picture ultimately harms the brand and the shopper. You must demand reporting that highlights failures just as clearly as successes. Data is most valuable when it challenges what the business already believes.

Furthermore, you must build flexibility into your strategy. If you lock in an annual budget without the wiggle room to respond to incoming data, you are setting yourself up for failure. You need leading indicators to tell you early on if things are moving in the right direction. Click through rates and early conversion metrics are excellent leading indicators, but you must have the operational agility to pivot when the data tells you to stop doing something.

Mastering Measurement with the OMG Framework

Measurement is often treated as an afterthought, scrambled together after a campaign has already launched. This reactionary approach degrades the quality of the insights and limits your ability to optimize. A highly effective way to structure your thinking is the OMG framework, which stands for Objective, Measure, and Goal.

Before a single dollar is spent, you must align on the objective. Is it a broad business objective, a specific campaign objective, or a tactical media objective? Once the objective is set, you determine exactly what metrics will be used to measure success. Finally, you establish the concrete goals or benchmarks you are trying to hit.

This level of intentionality is rare, but it is deeply necessary. For instance, if your objective is to capture shoppers searching for branded keywords, your measurement strategy should likely focus on trading those shoppers up to a higher price point item rather than simply selling them your opening price point product. Conversely, if your objective is acquiring new to brand customers in a highly price sensitive environment, you should feature your entry level items and closely track brand switching metrics.

You must also look beyond a single point in time. Relying on Return on Ad Spend as a north star metric is a trap. ROAS tells a very limited story. It does not account for the lifetime value of a newly acquired shopper, the loyalty they might develop over their next six purchases, or whether they add your product to a subscribe and save program.

True measurement requires a decision tree built on intentionality. You need to sit down with your team and ask hard questions. If this metric turns green, what specific action will we take? If this metric turns red, what creative are we swapping out? When you pre plan these "if then" statements, you transition from simply reporting on data to actively managing your business outcomes.

Harnessing Artificial Intelligence for Data Analytics

Artificial intelligence is the most transformative technology to hit the retail media space in years, particularly when it comes to data processing and reasoning. However, many marketers misunderstand how to use it effectively. They treat AI like a search engine, firing off one shot prompts and expecting brilliant, comprehensive business strategies in return.

To unlock the true power of AI, you need to start thinking more like a software engineer. This does not mean you need to learn how to write complex code. It means you need to adopt a systematic, process oriented mindset. AI is incredibly effective at writing and executing workflows, but it requires clear context, chunked tasks, and a defined goal. Big, monolithic processes confuse AI models. If you want a complex outcome, you must break the journey down into small, manageable steps.

Start by identifying the manual, repetitive tasks that drain your team's energy. If you hate pulling data from two different portals, dropping it into a spreadsheet, and formatting a weekly report, you should delegate that exact workflow to an AI agent. If you spend hours clicking through product pages to take screenshots for a competitive analysis deck, AI can automate that entire image gathering and slide building process in seconds.

As you get comfortable automating small tasks, you can begin designing AI personas based on specific "jobs to be done." You should not use the same AI agent for every task. You need one agent highly tailored for attribution modeling, equipped with specific rules about your category structure. You need a completely different agent focused on weekly analytics and execution. You need a third agent dedicated to building presentation decks. When you blend too many contexts into a single AI prompt, the quality of the output degrades significantly.

The most advanced teams are learning how to stack context over time. Instead of running a fresh Monday morning report in a vacuum, a well designed AI system will review your reports from the past month. It can remind you that four weeks ago, a specific item was struggling with replenishment, and prompt you to check if the supply chain has recovered.

This capability makes hoarding data an incredibly valuable strategy. In the past, saving every meeting transcript, Slack message, and daily log was pointless because no human could ever read through it all to find an insight. Today, you can feed a massive repository of unstructured data into an AI tool and ask it if a specific topic was mentioned during any meeting in July. The AI will find the exact timestamp in fifteen seconds.

Ultimately, artificial intelligence will not replace strategic thinking. In fact, as AI becomes more capable of handling low level data processing, human judgment will become infinitely more valuable. The teams that thrive will be the ones who use AI to do the heavy lifting, freeing up their human talent to tell compelling stories with the data, build deep partnerships across the retail ecosystem, and drive true intentionality in their measurement strategies.