The Core Tools Marketing Leaders Would Keep If They Started from Scratch
Architecture Shift: Marketing leaders advocate a five-layer architecture, challenging the "best-of-breed" approach for MarTech stacks.
CRM Alignment: CRM decisions focus on marketing-sales alignment, not just features, to ensure adoption and functionality.
AI Integration: AI is essential in execution layers, yet human input remains vital for strategic judgment and authenticity.
Orchestration Needs: Effective orchestration layers should minimize manual data exports, using tools like Zapier for edge cases.
Strategy First: Successful MarTech stacks prioritize strategic needs, ensuring tools are useful and regularly utilized.
Every marketing leader inherits technical debt. There’s the legacy CRM that was set up by a previous marketing director. Or, an automation platform the last CMO loved but no one on the team relied on. And, of course, the omni-present analytics tool that nobody quite understands but everyone's afraid to remove.
Imagine, if you could start fresh. With no fixed contracts, or data migration nightmares. And most importantly, without any of the “but, we've always used it" objections. I asked marketing leaders to paint us a picture of a future-state, an ideal world where they build from a clean slate, and they did not disappoint.
I heard from marketing leaders working as consultants managing multiple client stacks to CMOs building foundational systems to rebuild their marketing technology from scratch.
Want more from The CMO?
Sign up for a free membership to complete reading this article:
Have an account? Log In
The marketing leaders' responses coalesced around a five-layer architecture that challenges the "best-of-breed" conventional wisdom. Here's what they shared.
Core components of an ideal MarTech stack
1. The Infrastructure Layer
The answers split into two camps, first with the pragmatists who choose familiar, accessible tools, and architects who design around data flow first.
For Tricia Howard, CMO at Above Security, infrastructure isn’t the CRM at all. She shares how, as a startup foundational marketing leader, a zillion things vie for my attention."
Without a way to organize my tasks, things would get missed. Linear also allows visibility for my CEO to see what I’m working on and also streamlines his requests that come out from being in the office.”
Sonia Baschez, founder of the Meme Team podcast, takes a similar view, noting how much she relies on specific tools.
It’s not just project management, it’s my entire AI layer. Daily news monitoring, content drafts, meeting notes, client deliverables all run through one system. I’ve had separate tools for each of those things and it’s just a mess.
The pragmatist philosophy prioritizes what teams will actually use. A simple system everyone maintains beats a sophisticated system only specialists understand.
The architect camp flips the whole model. Laura Duckworth, former Head of Marketing at British Gas and now a fractional CMO, highlighting her stance.
If I were building a stack today with zero legacy ties, my core infrastructure would be Composable. I’d lead with Segment or PostHog as the CDP layer, because you need to own your data stream before it hits the platform.
Instead of the CRM serving as the source of truth, a data warehouse takes that role. The CRM becomes just another data source feeding into it. More technical upfront, but it solves the fragmented data ownership problem that plagues most stacks.
The split follows a predictable scale threshold. Under 100 employees, pragmatist infrastructure wins. For teams with 500+ employees with dedicated data teams, warehouse-first architecture takes over. The shift happens when manual data reconciliation becomes more expensive than building the warehouse.
2. The CRM / Operational Layer
The CRM conversation is less about features and more about organizational physics. The most common answer wasn’t about the best features or most integrations. It really comes down to having a tool that both marketing and sales will use. Robbie Ruuskanen, Marketing Director at ET Group notes that his "core infrastructure layer is HubSpot which acts as the central system of record."
Everything from campaigns, to forms, to automation and reporting feeds into it or pulls from Hubspot. I rely on it because without a single source of truth, alignment between marketing and sales breaks down fast.
CRM selection is an alignment decision disguised as a technology decision. Mauricio Acuña, Co-Founder at Impacto shares how “HubSpot is also our CRM of choice. It keeps us organized, tracks leads, and aligns our teams by combining marketing and sales data in one spot. It’s flexible, customizable, and gives us actionable insights to optimize our efforts.”
Heather Vaughn, Executive Director of Marketing at CI Design, cuts to why adoption is the only metric that matters, noting how it's about demonstrating value for the user.
The most sophisticated CRM fails if the sales team won’t log in.
Her honesty highlights a major sticking point. CRM decisions often have less to do with passion and more to do with path dependence, if your clients use it, your team knows it, and everything integrates with it, then it stays.
For larger organizations, Salesforce holds the same gravitational pull. And for smaller, relationship-driven businesses, Zoe Haugen, President at Haugen Holdings, uses Google Drive, leveraging Sheets and Docs, as her operational layer. Implementation discipline, she argues, matters more than CRM features.
3. The Intelligence Layer
Here's where answers diverged. You'll notice the split wasn’t between tools so much as it was between philosophies about what intelligence actually means.
Most leaders in the traditional BI camp use Google Analytics for web traffic, Looker Studio or Tableau for reporting, and SEMrush or Ahrefs for search intelligence. Heather Vaughn captures the standard:
Our intelligence layer is intentionally right-sized. Looker Studio, paired with Supermetrics to integrate data sources like Google, Meta, LinkedIn, HubSpot, and more, gives us a blended view of campaign performance across channels—which is where most of our clients’ most pressing questions actually live.
A growing cohort has moved to AI-assisted intelligence entirely. Maryanne Conlin, marketing strategy consultant, who shares that, "I still rely on Tableau and Adobe for reporting, but Copilot has been a game-changer for working with automation platforms—no more hours spent setting up campaigns, one well-written prompt will do it!"
My tech stack has changed—dwindled down to primarily Claude, Perplexity, and industry-specific LLMs or agents I’ve created to assist me in research, analysis and strategy development.
In a world where everyone has the same tools and the same data, qualitative intelligence that can’t be automated is where competitive advantage lives.
4. The Execution Layer
Every respondent mentioned AI in their execution layer. How they use it varied considerably. Several leaders run multiple models for different strengths. Nikola Baldikov, Founder at SERPsGrowth, takes the same approach at the model level, sharing that
We use different models for different tasks. ChatGPT and Gemini are good at detecting intent and helping us understand why someone would read a topic. Claude is better at long-form drafting and capturing tone and brand voice. It’s more natural in writing.
Others go deep on one tool and build around it. Sonia Baschez shares how she uses “Claude, running inside Thoughtful. I’ve built custom workflows for my specific use cases—podcast episode packaging, de-AI-ing copy, client context lookups. It’s not generic prompting. It knows my clients, my voice, my processes.”
Heather Vaughn describes how the most sophisticated users have moved beyond prompting entirely.
Claude is our primary AI tool, and we treat it as a strategic collaborator. We apply iterative refinement rather than accepting the first outputs. The project and knowledge structure have been valuable in long-term client engagements—they provide context across complex work, making thinking sharper over time.
Despite the enthusiasm, leaders identified clear limits. Kristin Steele, Founder at LaunchPad Creative highlights the human aspect as being instrumental.
The biggest challenge we see with AI is maintaining authenticity. While it’s a powerful tool for workflows and organization, marketing still requires human nuance, voice, and strategic judgment.
AI handles research, drafting, analysis, and structured tasks. Humans retain strategy, judgment, brand nuance, and anything where being wrong has reputational cost.
5. The Orchestration Layer
The final layer handles what your core stack doesn’t. And while Zapier was the near-universal answer for future-state tech stacks, the real question is how much orchestration marketers need.
Robbie Ruuskanen notes how they use a mix of automations to deliver on outcomes.
We use a mix of HubSpot automation and Zapier to help connect edge tools when we need. Its job is to remove manual handoffs and keep lifecycle movement clean—not to add complexity.
If you need 50+ Zapier workflows, your core stack isn’t doing its job. Orchestration should handle edge cases, not serve as your primary integration strategy.
Manual data exports are the universal failure signal. If someone on your team spends hours each week moving data between systems, the integration layer has broken down. Baruch Labunski, CEO at Rank Secure, describes what the ideal rebuild looks like:
In an ideal world, I would plan to rebuild my stack to offer more tighter integrations, focus on fewer tools, and build the stack around either HubSpot/Salesforce, implement a true first party data layer to BI, embed AI in the CRM and Analytics, and remove manual export workflows.
“
What’s Missing from These Tool Stacks
Did you notice what didn’t make most leaders’ rebuild-from-scratch stacks? Separate email marketing tools, standalone landing page builders, dedicated social schedulers, and specialized analytics platforms. Tricia Howard points out that,
The biggest point of ‘friction’ is how many tools there are out in the world now. Many tool differentiators are based on one or two unique features so you end up with several tools whose capabilities largely overlap, save those features. It eats up budget that could be used more effectively elsewhere.
A tool that delivers 70% of what a specialized platform provides, but integrates natively with your core stack, often wins.
The Strategy Question Nobody Asks First
So many issues can be solved from understanding the problem at hand. Often, it stems from asking the people doing the work. Labunski, highlights what should be obvious, and so often isn't.
With no strategy in mind, many MarTech stacks become bloated, overly complicated, and rely on temporary fixes to fill gaps. Marketers don’t even need 27 tools. They need 5 tools that actually do something useful.
Before asking which CRM to buy, ask your team what does qualified mean in commercial terms? What lifecycle stages matter for your business model? What attribution model will drive budget decisions? And, who owns data quality, and why should they care? Heather Vaughn, follows this up to note how the tools that count are the ones that are used daily.
The infrastructure question is really a maturity question. The right foundation is the one your team will actually use and build on.
Answer those questions first for yourself, then for your team, and the tool selection gets considerably easier.
What's next?
Sign up for a free account at The CMO Club and follow for more conversations with leaders navigating AI transformation.
