Outi Karppanen, Lead Marketing Industry Strategist at Supermetrics, on why collecting more marketing data is making teams slower, and what a real data foundation looks like.
The amount of marketing data flowing through Supermetrics has grown 230% since 2020. Not because marketers added a few new channels, but because every channel now spits out more rows, more granularity, more everything. And here’s the kicker: 56% of marketers told Supermetrics they don’t actually have time to analyze the data they’re collecting.
I wasn’t on this episode myself. David Welborn (DW), our CTO, filled in for me alongside Jessie. But when I watched it back, that stat stuck with me, because it describes almost every marketing team I talk to. Everybody has dashboards. Very few people have answers.
Outi Karppanen spent over a decade in media agencies, including a stint as a media strategist at Dentsu, before joining Supermetrics, the Helsinki-born marketing intelligence platform that started life in 2013 as a marketing ETL. She was a Supermetrics user long before she was an employee, and it shows. She talks about data like someone who has personally built too many reports at midnight.
Why is marketing data overload such a big problem?
The short answer: more data doesn’t mean more insight. It usually means the opposite.
“We were seeing this dilemma of marketers are just drowning in data and starving for insights.”
— Outi Karppanen, Lead Marketing Industry Strategist, Supermetrics
Outi’s fix is refreshingly simple. Stop grabbing everything just because you can. Start with your business KPIs, translate them into marketing KPIs, figure out which questions you actually get asked about marketing performance, and collect the data that answers those questions. Your CMO needs one level of data. Your creative optimizer needs another. Neither needs all of it.
“Don’t be data hoarders. You’re like a dragon sitting on top of your pile of data that you just don’t do anything with because you don’t have any time, and you’re just like, this is all mine, I want it all.”
— Outi Karppanen, Lead Marketing Industry Strategist, Supermetrics
I love that image. Data at rest has no value. It has to move and it has to be usable. Outi sees AI and agents helping here, not by doing anything glamorous, but by making data accessible to marketers who aren’t technical enough to wrangle a complicated visualization tool. Supermetrics was launching an agentic AI integration inside Google’s Gemini Enterprise (formerly Agentspace) the week of the recording, which is exactly that bet.
What does good marketing data governance look like?
This is the least sexy topic in marketing, and Outi says so herself. It’s also the one that decides whether anything else works.
“You have to have your data governance set, you have to have your naming conventions, your UTM settings, because if you don’t have that it’s just chaos.”
— Outi Karppanen, Lead Marketing Industry Strategist, Supermetrics
Her practical tip is one I’d steal immediately: use your data to monitor your data. Supermetrics customers build data quality monitoring dashboards that flag campaigns with missing naming conventions or broken UTMs in a big red box, so somebody can fix them before the data gets corrupted. Same idea for fraud. Supermetrics doesn’t filter anything out, it hands you the raw data, but that raw data is what lets you spot the weird traffic spike from a country you don’t advertise in, or a flood of sessions from one IP.
That one resonated with us. At BDEX we made a decision years ago to only sell quality data, and DW made the point on the show that identity data in particular is a jungle. There’s fraud, there’s junk, and sometimes people’s attempts to clean it up make it worse. You can’t judge quality until you can actually see the data.
“Now more than ever, the age-old saying garbage in, garbage out is very crucial. We’re starting to talk about utilizing AI and machine learning models, but what are you applying it to?”
— Outi Karppanen, Lead Marketing Industry Strategist, Supermetrics
She told a great story about almost booking a January trip to northern Thailand to escape the Finnish cold, only to find out at the last minute it would’ve been cold there too. Incomplete data, bad decision. Same thing happens when you model marketing performance on digital data alone and ignore your offline media.
How can AI help with marketing creativity?
Here’s where it gets interesting. Once you have clean, raw data, you can do things with it that go way beyond reporting.
Outi described using Supermetrics’ custom fields, which she calls “a chef’s pantry,” to prompt Gemini to label ad creative. Did the image have people? A dog? A 50% discount or a 20% one? An orange background or a green one? Label every ad, join it to performance, and suddenly you know that ads with dogs outperform ads without them. (Funny enough, our Chief Data Officer Michael has a nearly identical story about a mattress brand whose online sales tracked with dogs in the ads.)
“AI, when you analyze it, can tell you what, but it can’t tell you why. That’s where the limit is. You can’t go and ask the consumer why did you click this? Why did the dog work?”
— Outi Karppanen, Lead Marketing Industry Strategist, Supermetrics
She also raised a worry I share. Experienced marketers know to push back on AI output. Someone brand new to the industry, who never did the work by hand, may just accept whatever it says. Outi uses AI as a sparring partner and still makes her own calls. That’s the right posture.
Should marketing be data-driven or data-informed?
Outi draws a sharp line here, and I think it’s the most important idea in the episode.
“We call it data-informed decisions, not data-driven. Let it be your inspiration. But if you start just looking at data, start just looking at benchmarks, start using just AI, it’s all going to end up looking the same.”
— Outi Karppanen, Lead Marketing Industry Strategist, Supermetrics
Her cautionary tale: a big brand that chased short-term ROI, shifted budget from brand-building into tactical search and social with heavy discounting, and looked great quarter to quarter. Then a media mix model over three years showed the baseline sliding down right alongside the brand budget. They were slowly killing the brand and couldn’t see it in the short-term numbers.
That’s why she pushes for a holistic foundation that includes offline media, CRM, and sales data, not just last-click digital attribution. If you leave radio or podcasts out of your measurement plan, you’ll never learn they’re working for, say, truck drivers who listen all day.
DW had the line of the episode on this: put 20 people in a room to discuss a campaign without data and you’ll get 21 opinions. Bring data in and the conversation shifts from arguing over theories to deciding what to do next.
Is SEO dead?
A listener asked, and Outi’s answer was a clear no. It’s changing shape. With AI summaries answering questions before anyone clicks, the goal becomes getting your original thinking to show up as the default answer in ChatGPT or a Google AI overview. Original content wins, especially now that everyone is generating the same AI-written stuff.
The Bigger Picture
What I appreciated about this conversation is that nothing Outi said is flashy. Pick the right KPIs before you collect data. Enforce naming conventions. Monitor quality. Bring offline in. Let data inform you, but trust your gut on the brand.
That’s the unglamorous foundation every AI promise in marketing actually depends on. It’s the same thing we see in identity data every day: the companies that win aren’t the ones with the most data, they’re the ones with data they can trust and act on. Everything else is just a dragon sitting on a pile of gold.
Connect with Outi Karppanen on LinkedIn or check out Supermetrics, where you’ll also find her posts on the Supermetrics blog.
This article was adapted from an episode of Deconstructing Data, BDEX’s weekly podcast on data-driven marketing. Tune in live every Thursday at 4:15 PM Eastern on LinkedIn.
About BDEX: For companies that need clean identity data to power their products, BDEX offers unmatched quality and execution. Visit bdex.com and click “Talk to an Expert” to get started.
Video
Watch the full episode on YouTube: From Chaos to Clarity: Mastering Marketing Data in 2025
