> ## Content Index > Fetch the complete content index at: https://agustin-gutierrez.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # New Data, Old Questions - URL: https://agustin-gutierrez.ghost.io/new-data-old-questions/ - Published: 2026-08-13T08:03:30.000Z - Updated: 2026-08-20T13:20:40.000Z - Description: New measurement can improve old questions. Its greater value may be revealing questions the organization could not previously ask. - Author: Agustin Gutierrez Bazaco - Tags: Commercial Measurement **New measurement should not only improve the questions an organization already asks. It should also reveal questions that its previous measurement system could not express.** ## Principle > **Measurement should be designed around decisions, not available data.** New observability should support existing decisions without being constrained by the organization's existing way of seeing the business.*.* When we present a new In-Store Retail Media dashboard, one of the questions I hear quite often is: **Can we add the KPIs we already use?** It is a perfectly reasonable question. In many cases, it is exactly the right place to start. Most retailers are not beginning from zero. They already have POS data, loyalty programs, traffic measurement, promotion reporting, category dashboards, e-commerce analytics and years of experience working with data. They know which metrics matter internally. They know what gets discussed in meetings. They know which numbers trigger action. So when a new measurement capability appears, the natural instinct is to connect it to the way the business is already understood. But there is a second question worth asking: **What can we understand now that we could not understand before?** That is where things become more interesting. ## New data enters an existing measurement culture A new measurement system never really enters an empty organization. It enters a culture that already has its own language for performance. Some KPIs have been used for years. Historical benchmarks exist. Teams know what a good number looks like. People may be accountable for improving it. Budgets, targets or incentives may depend on it. That continuity is useful. It creates trust. It makes new information easier to adopt. And it prevents every new technology from inventing its own language for the business. But continuity can also become a constraint. A new capability can be translated into the existing measurement system so quickly that nobody stops to ask whether it should also change the way the business is being understood. The company gets new data. Perhaps better data. The dashboard becomes richer. But the questions remain almost exactly the same. A new measurement capability can end up doing little more than giving us better reporting of the same old representation of the business. ## Familiar KPIs are not the problem I do not think retailers should abandon existing KPIs every time a new capability appears. Quite the opposite. Sometimes the best thing a new system can do is measure something familiar more accurately. Estimated audience becomes measured audience. Manual reporting becomes automated. A fragmented metric becomes consistent across stores. That is useful. Sometimes the new capability goes a step further and connects things that previously lived separately. Media exposure can be connected to transactions. Traffic can be connected to category behavior. Delivery can be connected to conversion. The existing picture becomes richer. But every now and then, something else happens. The new capability makes a question possible that the organization could not really ask before. That is the point I find most interesting. Not because the previous questions were wrong. But because the organization can now see something that was previously outside its practical field of view. ## The conversation changes when something new becomes observable I see this in conversations with retailers. A meeting can begin with questions about the dashboard. Can we add this KPI? Can we reproduce the metric we already use? Can we break this down by store? All sensible questions. Then, once the conversation moves into what the system can actually observe, the questions often start to change. Can we compare exposed and non-exposed shoppers? Can we understand what happened before the transaction? Can we see whether the effect was incremental? Can we understand category impact rather than just sales of one promoted SKU? Can we identify conditions under which the intervention works better? At that point, the technology is not simply improving an existing report. It is beginning to expand the decision space. That does not mean every new signal deserves to become a KPI. Far from it. A capability is not valuable simply because it produces more data. But new observability can reveal commercial questions that were difficult to formulate when the underlying behavior could not be seen. ## Yesterday's measurement shapes today's questions There is another reason this matters. The way an organization measured the business in the past influences how it evaluates new measurement today. That is natural. If a retailer has spent years managing a certain activity through impressions, reach or Share of Voice, those metrics come with context. People understand them. They can compare them historically. They know who owns them. They know what action normally follows. A new signal has none of that. So the instinct is to translate the unfamiliar into something familiar. That makes adoption easier. But if every new capability is forced into the old language, the measurement system can become conservative without anyone consciously deciding that it should. The danger is not that the organization rejects new data. It may embrace it enthusiastically. The danger is that new data is allowed to improve the existing view of the business, but not to challenge it. ## In-Store Retail Media makes this tension visible This is one reason I find In-Store Retail Media such an interesting environment for measurement. For a long time, the natural language of the medium was media delivery: playouts, impressions, reach, frequency, Share of Voice. Those metrics are useful, and they are not going away. But once exposure can be connected to behavior, transactions and incrementality, the questions can change. The conversation no longer has to stop at: **How much media did we deliver?** It can move towards: **Did that exposure actually change behavior?** Or: **Under what conditions did it create incremental commercial value?** The store itself has not changed. The screen has not changed. What has changed is what the organization can observe about what happens around them. And that can change the way the organization understands the same commercial environment. ## Continuity matters. So does discovery. This is where the balance becomes important. Existing measurement systems should give new capabilities somewhere to land. Retailers need continuity, comparability and governance. A new platform that ignores all of that is unlikely to be very useful. But the existing measurement culture should not become the ceiling. A mature system should be able to support the decisions the organization already knows it needs to make while remaining open to questions that were not previously practical to ask. Sometimes new measurement will simply produce a better answer to an old question. That can be valuable. Sometimes it will connect evidence that used to sit in separate systems. Also valuable. And sometimes it will change the question itself. Perhaps that is the most interesting outcome. ## The real test of new observability When a new measurement capability appears, it is tempting to ask: **Which new KPIs can we add?** I think there is a harder question: **What can we understand now that our previous measurement system could not represent?** Sometimes the answer will be modest. The new system may simply make an existing measurement more accurate, consistent or easier to use. That is still progress. But sometimes the answer opens a different way of thinking about the business. A behavior that used to be invisible becomes observable. Two signals that used to live separately can now be connected. A commercial effect that used to be assumed can now be tested. And a question that previously had no practical answer becomes worth asking. The danger is not that organizations keep asking old questions. Some old questions are still the right ones. The danger is that new data arrives and nobody stops to ask whether the questions should change too. Because sometimes the real value of new measurement is not a better answer. **It is a better question.** --- *This essay is part of the In-Store Retail Media Framework under Commercial Measurement. New readers can begin with* [*Start Here*](https://agustin-gutierrez.ghost.io/start-here/) *or explore the* [*Principles*](https://agustin-gutierrez.ghost.io/principles/) *that guide the Journal.*