Analytics Club Podcast – Many Companies Collect Too Much Data – And Act Too Little Upon It

This post is a summary and English translation of a German-language podcast episode. Thomas Gerstmann from darive joined Florian Möllers E-Commerce Analytics Podcast (German Language) to talk about one of the most common – and most costly – mistakes in data projects: collecting data without ever consistently deriving actions from it.

🎙️ Listen to the original German episode here: Spotify | YouTube

Introduction

You've invested in Google Analytics. You have dashboards. You've connected your ad platforms. Your data infrastructure is solid.

And yet – somehow – your business decisions still feel like gut calls.

You're not alone. According to Thomas Gerstmann, data strategist and founder of darive, this is the dominant pattern in the mid-market: companies build beautiful data infrastructure and then park it in the garage.

The Problem: A Well-Trained Tech / Data Collection Arm – and a Neglected Analytics / Action Arm

Thomas uses a vivid analogy throughout the conversation: imagine a person with two arms. One arm – the tech/IT arm – is massively overtrained. Companies pour time and money into data collection, tracking setups, clean data models, and high-quality pipelines. The other arm – the one responsible for actually using and analyzing the data to extract data insights and make decisions – is completely neglected.

"I experience this constantly. The technical stack gets cleaned up, the data quality improves – and then nothing happens with it."  

This gap is especially pronounced in SMEs: companies with 50 to 500 employees who have the ambition to become data-driven, but lack the structures, culture, and rhythm to turn data into action.

The result? Dashboards nobody trusts. Insights that never get implemented. And eventually, the dreaded relaunch – where everything gets thrown out and rebuilt from scratch, because nobody maintained a culture of continuous iteration.

The Real Silos Aren't in Technology, They are in Marketers Heads

A common response to analysis paralysis is to invest in integration technology: "We just need to break down our data silos." Thomas pushes back on this.

"Silos aren't primarily technical. They're between people. Even if you technically integrate everything, you haven't integrated the thinking."  

His solution isn't a new tool – it's a monthly meeting. One table. Representatives from marketing, sales, customer care, social media, Google Ads. Each person brings their numbers. And suddenly Facebook knows what Google Ads is doing. Customer care understands what's happening in the funnel. The whole customer journey becomes visible – not because the data was magically unified, but because the people finally talked to each other.

The Right Business Questions Beat the Right Technology and Tools

One of the most powerful moments in the conversation comes when Thomas challenges the standard opening question in data projects: "What data do you currently look at?"

He argues this question is already framed wrong – it's too technical, too tool-centric. The better question is:

"What do you actually want to achieve? And how will you prove you achieved it?"  

This reframe changes everything. When a company starts from outcomes instead of from tools, they often discover that their measurement problem is far simpler – and more solvable – than they thought. Sometimes the answer to "how do we measure this?" is a tally mark on a piece of paper, not a six-month analytics integration.

Imperfect data that drives action beats perfect data that collects dust.

Want to go deeper into this? Read more about "asking the right questions" here.

The 90-Day Data Rhythm: Strategy → Experiment → Analysis

The most actionable part of the conversation is Thomas's core framework: a repeating 90-day cycle that any mid-sized company can implement.

Days 1–30: Strategy & North Star Metric Definition

Get clarity on what you're optimizing for. Define one North Star metric. Identify two or three concrete ideas to move it. Don't start everything at once – strip each idea down to its absolute minimum viable version.

Days 31–60: Experiment, Collect & Build

Launch the experiment. Make sure data collection. The goal isn't a perfect feature – it's the fastest path to a live test. Sometimes a developer can implement it in real-time during the meeting. Speed is the point. Collect data, observe, adjust.

Days 61–90: Analyze & Cultivate

Review what worked. Build a lean dashboard around the metrics that actually proved meaningful. Bring the whole team into the results conversation. Start building organizational muscle memory around this process.

Then repeat.

"You don't need statistical significance at every step. You need a rhythm. Last week: 3 calls. This week: 10 calls. That's a signal worth paying attention to."  

What AI Changes – and What It Doesn't

Thomas is direct about AI's impact on analytics work: the operational layer is already being disrupted. AI tools can now take a CSV, run regression analyses, calculate medians and distributions, and deliver actionable recommendations – all from a plain-language prompt, with the Python running invisibly in the background.

"You upload a file, ask a question, and within seconds you get a high-quality data analysis complete with predictive trends. The technical layer is becoming invisible."  

The implication for analytics professionals: the "Sherlock Holmes" detective work of finding anomalies in data will increasingly be done by AI. What remains irreplaceable is knowing which questions to ask – and having the organizational context to evaluate whether the answers make sense.

The companies that will win aren't the ones with the most sophisticated tech stack. They're the ones who've developed a culture of asking good questions, running fast experiments, and acting on what they find.

Where to Start: The darive Benchmark Check

If you recognize your company in any of this – too much data, too little direction, no clear rhythm – Thomas's recommended first step is the darive Benchmark Check.

It takes about 8 minutes (30 questions) and gives you a personalized assessment across the five dimensions that darive is built around:

1. Strategy – Do you have a clear direction?

2. Experiment – Are you testing and learning systematically?

3. Build/Development – Are you shipping fast enough?

4. Culture – Is data part of how your team thinks?

5. Rhythm – Do you have a repeating cycle that produces results?

You'll receive a detailed breakdown of where your biggest bottlenecks are – and that becomes the starting point for a conversation about where to pull the biggest lever in the next 30, 60, or 90 days.

👉 Read more here
👉 Take the free Benchmark Check

Thomas Gerstmann has 20+ years of experience in data strategy, advanced analytics, and UX. He's the founder of darive – a training and coaching program that helps mid-sized companies build the processes, culture, and rhythm to become genuinely data-driven. He's currently working as a digital nomad, having just completed a year-long journey through the Balkans.