Welcome to Spark, a newsletter from Vivace. We curate and publish the most interesting thinking and ideas from our community on themes ranging from business and finance to culture and creativity. Send pitches and feedback to [email protected]. We’d love to hear from you.

Unlock your potential with our comprehensive Business Analyst Course at https://www.apexvirtualeducation.com/courses/business-analyst-course/

Hello {{ First Name | Spark reader }},

There are very few people I trust more than Josh Mendelowitz when it comes to marketing technology, data, and measurement. He's one of the smartest marketers I know and a longtime friend of Vivace.

In this week's Spark, Josh explores what AI means for one of marketing's most enduring tools: the scorecard. His conclusion is one I wholeheartedly agree with. While the technology is changing rapidly, the fundamentals of good marketing measurement remain remarkably consistent.

I hope you enjoy it.

All the best,

—Joel

Lately, I’ve found myself asking a question I never expected to be asking: What would Marek do?

Marek Pasternak was a friend, a colleague, and one of the smartest technical marketers I’ve ever known. He had a rare ability to understand both the mechanics of marketing technology and, more importantly, how to make it useful. He taught me about paid search, data, dashboards, and much more during the years we worked together at LSEG. When I later found myself in need of some technical expertise at Marqeta, Marek was one of the first people I called.

Marek passed away unexpectedly last year. I know his friends and family still miss him dearly.

If Marek were here today, I’d probably ask him what he thinks about AI and what it means for measuring marketing performance. I can’t ask him that question directly, but I can think about the lessons he taught me and what I think he might say. This article is inspired and dedicated to him. 

Everywhere you look, the value of marketing is being questioned. Whether we are in a full-fledged crisis or simply at an inflection point, one thing is clear: marketers can no longer take the value of marketing for granted. We need to consistently demonstrate how marketing is contributing to the commercial performance of the business.

And there remains no better tool to make this case than the humble marketing scorecard.

At its simplest, a marketing scorecard is a consistent and regular view of marketing performance. It brings together a core set of KPIs, tracks them over time, and compares performance against an agreed benchmark. But its real value goes beyond reporting numbers. A good scorecard helps the business understand what is driving performance, where things are headed, and what marketing should do next.

When done well, it gives executives a clear snapshot of marketing performance while creating a shared language for what marketing is delivering. Over time, that consistency can build something even more valuable: trust in marketing’s ability to measure, explain, and ultimately influence business outcomes.

The way we create these scorecards is changing. AI and increasingly agentic workflows are making it easier to collect, clean, analyze, and present marketing data. The mechanics of producing a scorecard are becoming significantly easier.

But as technology changes, the fundamentals of good measurement become even more important.

1. Data is never perfect, but we have to start somewhere 

While I am sure there are some data-first organizations out there with beautiful infrastructure, squeaky-clean data hygiene, and robust governance, most companies struggle with data gaps and tech debt to varying degrees. It’s tempting to look at the inaccuracies, inconsistencies, and inaccessibility of our data and think that pulling together a marketing scorecard isn’t feasible or, at the very least, is not worth the squeeze. Yet, there is no perfect data setup, and it’s almost always messy. 

So you have to start somewhere. In fact, I’ve found that, more often than not, the very act of reporting on the data serves as a self-fulfilling mechanism for improving it. The more you show the data, the more the respective teams feel compelled to clean up their parts of the data ecosystem. This is something Marek did better than most. He would start by hacking together dashboards using manual CSV exports rather than automated data feeds, just to show proof of concepts that serve as a starting point to get buy-in. 

That lesson feels particularly relevant as AI begins to automate more of the work involved in data preparation. AI agents can help identify inconsistencies, enrich records, and automate some of the manual work involved in maintaining clean CRM and marketing data. But automation doesn’t remove the need for good underlying data, and it certainly doesn’t mean we should wait for perfect data before we start measuring. 

2. Be consistent but adaptable 

If a marketing scorecard is going to build trust, people need to know what they’re looking at from one reporting period to the next. That means agreeing on a core set of KPIs, a reporting cadence, and a consistent way of presenting the data.

Are you reporting monthly or quarterly? Are you measuring against targets, year-on-year performance, or another benchmark? Which metrics are always going to appear? However you answer those questions, the important thing is to establish a methodology and stick with it.

The consistency builds trust in the process. Business priorities shift and evolve, and so too should your scorecard. That is encouraged. But changing the approach too often or too flippantly undermines the message and can erode trust. 

This is particularly important as AI makes data more accessible across an organization. One of the risks of democratizing data is that everyone starts creating their own version of the truth, using different definitions, methodologies, and sources. AI can help solve this by embedding agreed definitions and methodologies directly into the agents, playbooks, and workflows that people use to access and analyze marketing data.

The goal isn't to make everyone look at exactly the same report. It's to make sure that, however the data is sliced and presented, everyone is working from the same underlying definitions and methodology.

3. Think audience-first, always

One of the easiest mistakes to make with marketing reporting is to build the report you want to see rather than the report your audience actually needs.

I’ve certainly been guilty of this. When you’re building dashboards and scorecards, it’s natural to think about how the data is structured, how the different reports connect, and how you can make the system as efficient as possible to maintain. But that can lead you to optimize for the person creating the report rather than the person consuming it.

I remember doing exactly this while working with Marek. I was trying to build a suite of connected dashboards and thought linking everything together would make the experience easier. Marek pushed back. His view was that different teams needed their own dedicated dashboards, tailored to the information and decisions relevant to them.

He was right. The connected approach might have made things easier for us as the creators, but it would have made things harder for everyone else. It added noise and forced people to navigate information that wasn't relevant to them.

The best scorecard isn't necessarily the one with the most information. It's the one that gives the right audience the information they need, in a format they can actually use.

This is another area where AI can make a significant difference. One of the biggest constraints on creating tailored reporting has traditionally been the time and resources required to produce and maintain multiple versions. With AI, a single underlying scorecard could be used to generate different views for different teams, regions, or levels of the organization, while maintaining a consistent underlying methodology.

The opportunity isn't to give everyone more data. It's to give everyone the right data, in the right context, for the decisions they need to make.

4. It’s about measurement AND insights 

It’s tempting, when producing a marketing scorecard, to focus on what happened: leads increased, pipeline declined, conversion rates fell. But reporting the numbers is only the beginning.

As a CMO once told me, “Don’t just report the news.”

A good scorecard should help explain what is happening, why it is happening, and what we should do about it. What’s working? What isn’t? What has changed? What might explain that change? And, most importantly, what are we going to do next?

This is the point at which marketing measurement becomes marketing insight. The scorecard stops being a regular performance report and starts becoming a tool for shaping the strategic direction of the function.

This is also where AI could fundamentally change the art of the possible. AI can analyze large volumes of structured and unstructured data far faster than any individual marketer. It can identify patterns, surface relationships that might otherwise be missed, and bring together information from multiple sources to provide a richer view of what is happening.

But AI-generated insight is only as useful as the context behind it. Sometimes, a change in conversion rate can be significant. But other times, it might be explained by a change in audience, market conditions, or something happening elsewhere in the business. The more context we can give our AI systems, the better the questions they can answer and the more useful the insights become.

The opportunity isn't simply to automate the reporting of what happened. It is to spend less time producing the report and more time understanding why it happened and what we should do next.

5. It was never about the dashboard

For years, the marketing dashboard was the middleman between our data and our scorecard. Data came in from the warehouse — or, more realistically, from a patchwork of integrations, ETL connectors, and spreadsheets — and the dashboard gave us somewhere to bring it all together.

And, for all its frustrations, it felt like a revelation.

One place to visualize performance. Filters to slice the data. Charts to explore trends. A way for marketers to self-serve rather than constantly ask someone else for a report.

Of course, let’s be honest. It was also often buggy, slow, and cumbersome. We spent countless hours refreshing dashboards, changing filters, and waiting for them to inevitably crash, before exporting everything into Excel because the dashboard never quite told the full story.

As someone who has spent years building, maintaining, and using marketing dashboards, it’s almost painful to admit it: the writing is on the proverbial dashboard wall.

The traditional marketing dashboard is going away.

And I think that’s a good thing.

Because it was never really about the dashboard. The dashboard was simply a way to make data accessible and digestible, so marketers could understand performance and make better decisions.

AI and agentic workflows are beginning to remove that intermediary. Instead of navigating through a predefined set of charts and filters, we can increasingly ask questions in plain language, request a particular view of the data, and have an agent generate the analysis or report we need. It can bring together additional data sources, provide context, and even challenge the assumptions built into the way we have traditionally reported performance.

The implication isn't that marketing scorecards disappear. It’s that the interface through which we access them becomes much more fluid.

The key lesson here is not to mistake the tool for the outcome.

The dashboard was never the goal. The aim was always to create better understanding, better decisions, and ultimately, better marketing.

Looking ahead

There’s a version of this article that takes a much more pessimistic view of what comes next. One that worries that increasingly capable AI and agentic workflows will make much of the work marketers do today redundant, and make it less important for marketers to understand their own data.

I’m not convinced. When I think about how Marek would approach this moment, I don’t imagine him standing on the sidelines worrying about what AI might replace. I imagine him trying the latest tools, figuring out what they could do, finding where they fell short, and working out how to make them useful. I think that’s the mindset marketers need now.

AI is changing how we collect data, analyze performance, and communicate insights. Some of the tools and processes we have relied on for years will disappear, if they haven’t already. Others will evolve beyond recognition.

But the fundamentals haven’t changed. As marketers, it is incumbent upon us to understand the data we have, know its limitations, build trust in the numbers, put the audience first, find the deeper insights, and never lose sight of the ultimate goal. I can’t say for sure what Marek would do - but I think he’d be up for the challenge.

Josh Mendelowitz is a senior B2B marketing leader with experience spanning product marketing, demand generation, digital marketing, marketing technology, and analytics. He is currently SVP of Demand Generation at BNY, where he focuses on driving growth across its Global Payments & Trade business. Previously, Josh held marketing leadership roles at Marqeta and LSEG/Refinitiv, where he built teams and programs across search, paid media, marketing technology, and digital growth. He has spoken and written on topics including SEO, paid advertising, and marketing technology. Josh is passionate about the intersection of technology, data, and marketing — particularly how marketers can use emerging tools to work smarter while staying focused on commercial impact.

Thanks for joining us this week. Anything we missed? Something we should include next week? Send us your shout-outs and strong opinions to include in next week’s edition at [email protected]. 

Spark is a production of Vivace, a global B2B creative studio and consultancy that helps businesses drive meaningful brand and commercial impact. Get in touch if you’d like to chat with any of the team. Have a great week ahead.