The Data Leaders Digest – Third Edition
This Month in Data – Smarter Renewals, Better Questions, and Breaking the Cycle of Burnout
The best insights come from real conversations, with real leaders making tough calls, solving messy problems, and figuring out how to scale their data teams in the real world. That’s why we started The Data Leaders Digest.
Each month, we share what’s actually happening inside data teams, pulled from our podcasts, live events, meetups, and honest behind-the-scenes conversations.
No theory. Just practical takeaways you can use right now to make better decisions, avoid common pitfalls, and keep moving forward.
If last month’s edition focused on the immediate pressure of rising ETL costs, this month looks at the bigger shift: how smart data teams are protecting productivity, rebuilding trust and planning ahead, before the pressure hits.
In this edition, we break down Snowflake’s latest security updates, share highlights from our March 25 LinkedIn Live event on cutting ETL costs without disruption, and continue Season 2 of the Data Matas Podcast with honest conversations from MoonPay and IRIS Software.
We’ll also share frontline insights from the London Analytics Engineering Meetup at Monzo HQ, introduce three new resources to help with ETL renewals and strategy planning, and preview what’s coming next, including a powerful new podcast episode with Nik Walker from Co-op.
Product News / Announcements
Security Update: Snowflake Phasing Out Passwords by November 2025
Snowflake is rolling out key pair authentication – by November 2025, password-based access will no longer be supported for both human and service users.
Here’s what’s changing:
- Service users (like those used by ETL tools) must switch to key pair authentication or OAuth.
- Human users must switch to Multi-Factor Authentication (MFA) or key pair authentication
Why it matters: 🔒 Stronger protection against credential-based attacks ⚠️ Compliance-driven shift away from insecure access methods 🛠️ Required migration for all teams – especially ETL pipelines—before the deadline
📘 Read our Snowflake Key Pair Authentication Setup Guide
Why now? A quick reminder of what’s at stake
In 2024, the ShinyHunters breach exploited stolen credentials, some dating back to 2020, and affected major brands like Ticketmaster and Santander.
Third-party access through vendors like EPAM Systems exposed the gaps in relying on passwords alone.
LinkedIn Live Recap: How to Cut ETL Costs Without Disruption
On April 25, we hosted a LinkedIn Live session tackling one of the biggest hidden challenges data teams face today: how to reduce ETL costs without hurting delivery.
Our expert panel featured:
- Aaron Phethean – Founder, Matatika
- John Grubb – Head of Data & Business Systems, Pantheon
- Michael D. – Senior Director of Engineering, Imperva
Together, they shared hard-won lessons on managing cloud spend, fighting vendor lock-in, and avoiding the traps that can turn ETL migrations into productivity nightmares.
Key takeaways:
- Visibility comes first – track ETL costs like a product, not just a bill.
- Not every migration is worth it—test real workloads side-by-side (using Mirror Mode) before making a switch.
- The real cost isn’t just your cloud bill—it’s the engineering time lost to brittle systems and firefighting. Plan early—start 3–6 months ahead of renewal, not when the panic hits.
Missed the event? Watch it here: How to Cut ETL Costs Without Disruption: Proven Strategies from Data Leaders (LinkedIn Replay)
Prefer to read instead? How to Cut ETL Costs Without Disruption: Proven Strategies from Data Leaders (Full Article Recap)
Why we’re hosting these sessions: When Aaron first started working in data infrastructure, most of the content he found was either too technical or buried in sales messaging.
What didn’t he see? The kind of honest, grounded conversations happening behind the scenes—between leaders trying to manage costs, delivery pressures, platforms, and renewals in the real world.
That’s why we started this monthly series of LinkedIn Live events (and launched the Data Matas podcast):
To create space for real conversations with smart people doing the work. No fluff. No posturing. Just lived experience you can use to make better decisions under pressure.
If you’ve been tuning in, a big thank you. And if you’re interested in joining a future session or podcast episode, let’s talk.
How Matatika is helping: For teams approaching renewal or feeling the pressure to migrate, Matatika’s Mirror Mode lets you run your existing ETL setup in parallel with Matatika – zero disruption, no rework, no double billing. It’s the safest way to test, validate, and switch on your terms—not under pressure.
🗓️ Our next live event will be on Friday 23 May 2025 at 1:30pm
Lessons from the Frontlines of Data Leadership
📍 London Analytics Engineering Meetup @ Monzo HQ – April 10
On April 10th, I joined 239 data engineers, analysts, and architects at Monzo HQ for the biggest London Analytics Engineering Meetup to date—and once again, the room was buzzing with sharp takes, honest questions, and more than a few laughs.
First up was Daniel McNamara (Head of Analytics Engineering at Taptap Send), who sparked a lively debate: is analytics engineering really its own discipline—or just a new name for an old set of problems?
What followed was a refreshingly open discussion on how differently teams approach modelling and structure, and why data quality and trust are quickly becoming non-negotiables, not nice-to-haves. With more talk of data mesh and ownership models, it’s clear: teams are rethinking their stacks from the ground up.
One question that kept coming up: How should we measure ROI?
The best signals?
- Less code
- Lower cost
- More up-to-date data
- Better answers
Then came Bill Wallis (Sainsbury’s), flying the flag for DuckDB—and showing how it’s helping him simplify dev workflows and cut costs by running powerful SQL queries locally, without needing cloud infra just to test something.
It was practical, grounded, and exactly the kind of session that makes these meetups worth showing up for.
🎙️ Data Matas Podcast: Season 2 Continues
Season 2 of Data Matas launched with Jon Hammant from AWS breaking down how to scale cloud infrastructure without spiralling costs. Since then, we’ve continued the season with new conversations focused on how real teams are adapting under pressure—navigating change, balancing delivery, and building smarter systems that last.
Episode 2: Emily Loh, Director of Data at MoonPay
Emily L. leads MoonPay’s data team across engineering, science, and machine learning—and in this episode, she shares how crypto’s fast-moving landscape has forced them to rethink how data teams work.
Emily talks about shifting from reactive reporting to outcome-focused strategy, building flexible systems that can adapt fast, and applying her 20/40/40 resource model to balance business-as-usual, building, and research.
Key takeaways:
- The 20/40/40 model that helps escape the firefighting trap
- Why real innovation starts by saying “no” to low-value work
- How to apply AI to remove drudgery, not create more of it
- Designing for unpredictability—especially when the product may change next quarter
🎧 Watch & listen to the full episode
🔗 Prefer to read instead? Here’s the full article
https://www.matatika.com/crypto-inspired-data-strategy-what-top-teams-do-differently/
Episode 3: David Draper, Data Science Manager at IRIS Software
David Draper brings a fresh perspective on what it means to modernise data infrastructure—without hype. Drawing on his background in teaching and product, David shares how his team at IRIS is building modular, explainable systems that scale without breaking.
This conversation explores the value of quantum-inspired thinking in data: testing multiple approaches in parallel, protecting time for innovation, and making AI outputs human-friendly and business-ready.
Key takeaways:
- How to reduce migration risk by running systems side by side
- Ways to carve out time for experimentation inside fast-moving teams
- AI as a support tool, not a black box—why explainability matters
- A practical approach to building flexible, testable data systems
🎧 Watch & listen to the full episode
🔗 Prefer to read instead? Here’s the full article
🔗 Explore more episodes of the Data Matas podcast here
New Resources to Help You Plan Smarter ETL Renewals
This month, we released three new guides to help data leaders break free from reactive cycles and plan renewals with more control and confidence:
1.Why Data Migration Tools Fail – and How to Make Them Work
Why do so many migrations feel painful, even when you have the right tools? This article breaks down the real reasons—and how to shift from firefighting to future-proofing your systems. 🔗 Read the full article
2. Stop Fixing Data Pipelines – Start Driving Strategy
If your team is stuck fixing endless pipeline issues, you’re losing time you could spend driving real impact. Here’s how to build self-healing pipelines that free your team to focus on what matters. 🔗 Read the full article
3. Switch ETL Tools Without Risk – How Mirror Mode Makes Migration Safe
Migrating ETL tools can feel risky—downtime, double billing, broken pipelines. That’s why we built Mirror Mode, a risk-free way to test Matatika side-by-side with your current setup. No disruption, no pressure, no commitment until you’re confident.
What’s Coming Up Next?
🔜 Data Matas Podcast: Nik Walker (Episode 4)
📅 Releasing Thursday, 1 May at 10am
Data engineers don’t burn out from work, they burn out from pointless work. That’s the theme of our next episode, and Nik Walker (Head of Data Engineering at Co-op) doesn’t hold back.
In this honest and energising conversation, Nik shares how a culture of reactivity, unclear priorities, and dashboard overload can slowly break a team, and how to build one that thrives instead.
Why you should listen:
- Learn how to protect your team from burnout, not just outages
- Hear how Co-op is transforming legacy infrastructure without crushing its people
- Discover why trust, discovery, and meaningful work matter more than speed and shiny tools
🎧 Explore the full podcast series here
🔜 LinkedIn Live
📅 Join us live on Friday, 23 May at 4pm
How Do Your Costs Stack Up? A Ground-Level Look at Data Infrastructure Spend
Data teams are under growing pressure to reduce costs—especially with cloud bills rising, vendor renewals looming, and finance leaders asking harder questions. But where do you start when every layer of your stack claims to be mission-critical?
In this upcoming LinkedIn Live, we’re taking a practical, ground-level look at what drives data infrastructure costs, from storage and compute to transformation and reporting. Whether you’re in the middle of a renewal conversation or just trying to make your spend more transparent, this session will help you figure out what’s really costing you, and what you can do about it.
Our expert speaker is Ian Whitestone, the Founder & CEO at SELECT, the leading Snowflake optimisation and cost management platform. Ian brings deep expertise in warehouse tuning, query performance, and cost visibility across modern stacks.
RSVP link coming soon
We’re also lining up a second speaker with transformation and BI expertise to complete the full-stack view of cost drivers.
Stay tuned—it’s going to be a good one.
Found This Edition Useful? Pass It On!
If you found this newsletter valuable, why not share it with a colleague or your network? Forward it to your team, drop it into your Slack channel.
Follow Matatika on LinkedIn | Subscribe for More | Visit Our Website
Connect to Apps & Data now
Use Matatika to rapidly produce insights from more than 500+ apps and community sources
Data Leaders Digest
Stay up to date with the latest news and insights for data leaders.