Which tools sit where in the modern GTM stack? Which replace each other, which complement? Here's the complete breakdown per layer — Clay, Cargo, Common Room, HubSpot, Salesforce, n8n, and the rest — with concrete choice criteria.
The most common question from founders building their GTM stack: "do we need Clay if we already have HubSpot?" The answer is almost always "yes, because they serve different layers." But that answer only works if you understand what those layers are and what each does.
In this article I break down the modern GTM stack into six layers, explain which tools dominate per layer, and give a choice framework per layer. For context also read What is GTM Engineering?.
The six layers of a modern GTM stack
Seen from above, a complete GTM stack consists of six layers. Each has a specific purpose and specific tools. Some tools serve multiple layers, but no tool does all six well.
- Layer 1: Identity & Visitor. Who visits your site? Who's in your market?
- Layer 2: Data & Enrichment. Clean, enriched prospect data.
- Layer 3: Signal & Intent. Detection of buying signals.
- Layer 4: Orchestration & Automation. The "glue" between everything.
- Layer 5: Engagement. Outreach (email, LinkedIn, calls).
- Layer 6: CRM & System of Record. The single source of truth.
Important disclaimer: not every company needs every layer. Seed-stage starts with 2-3. Scale-up through Series B typically has 4-5. Only enterprise has all six. Don't build everything at once.
Layer 1: Identity & Visitor
Goal: identify who visits your website (anonymously) and which companies are active in your market.
Dominant tools:
- RB2B: US-focused, identifies anonymous visitors at person-level. Doesn't work in Europe (GDPR).
- Vector / Albacross / Leadfeeder: EU-friendly, identifies companies (not persons) based on IP. GDPR compliant.
- Common Room: Combines visitor identification with community engagement tracking.
When yes: When you have high-volume site traffic (1,000+ visitors/month) and want to see which ICP accounts look without converting.
When no: Too early in early-stage with <500 visitors/month. The noise-to-signal ratio is then too low.
Layer 2: Data & Enrichment
Goal: enrich prospect data to a usable level. Email addresses, phone numbers, firmographics, technographics.
Dominant tools:
- Apollo: All-in-one (data + sequencer). Strong in US, decent in EU. Good entry-level.
- Cognism: EU-focused. Strong compliance. Pricier than Apollo.
- LeadMagic: Email discovery specialist. Cheap. Good for waterfall.
- Findymail: Email verification + discovery. Pay-per-success.
- Datagma: Mobile + alternate contacts.
- BuiltWith: Tech stack data.
The key insight: choosing one tool works poorly. A waterfall (see The waterfall enrichment stack) delivers 2-3x better coverage at comparable cost.
Layer 3: Signal & Intent
Goal: detect when an account is "in market" — via signals.
Dominant tools:
- Common Room: Cross-channel signal detection. Strong in community engagement signals.
- Trigify: LinkedIn-specific signals (job changes, engagement).
- Bombora: Third-party intent data (research activity in your category).
- G2 Buyer Intent: Intent on your category via G2.com.
- UserGems / Champify: Specifically champion-job-change detection.
When yes: When you have a defined ICP and want to prioritize outbound on "in market now" versus "fits ICP."
When no: Pre-PMF. First validate your product, then add signals. Used too early it muddies your signal-to-noise.
Layer 4: Orchestration & Automation
Goal: the "glue." Workflows moving data between tools, applying rules, running AI agents.
Dominant tools:
- Clay: Specifically for GTM data orchestration. Strong in waterfalls, AI personalization, enrichment workflows. Per Clay's own GTM Engineering manifesto this has become the central layer for many teams.
- Cargo: Programmable data engine. Strong in dev-friendly workflows and complex transformations. Pricier, more powerful.
- n8n: Open-source workflow automation. For teams wanting self-host and custom integrations.
- Make (formerly Integromat): No-code automation. Cheaper alternative to n8n with SaaS convenience.
- Default: AI-native workflow tool, strong in agent orchestration.
Which to pick?
- Start with Clay if you primarily do enrichment + outbound automation. Fast results, low learning curve.
- Add n8n or Make for automations outside GTM data (internal processes, customer onboarding, CS workflows).
- Consider Cargo when your team has developer resources and wants to build complex data flows.
- Default for teams working deeply with AI agents.
Layer 5: Engagement
Goal: the outreach itself — email, LinkedIn, calls.
Dominant tools:
- Instantly / Smartlead: High-volume email outbound. Strong deliverability features.
- Lemlist: Personalization focus, video personalization.
- Apollo (sequencer): All-in-one with data layer.
- Outreach / Salesloft: Enterprise sales engagement.
- LinkedHelper / Trigify: LinkedIn automation (mind TOS risk).
- Aircall / Dialpad: Call infrastructure.
Key choice: stack or monolith. Apollo does everything but mediocrely; Instantly + Smartlead specialize in email and are better at it, but you must integrate more.
Layer 6: CRM & System of Record
Goal: the single source of truth for all accounts, contacts, deals, and activities.
Dominant tools:
- HubSpot: Dominant in NL and European B2B SaaS. Strong in marketing-sales-service alignment. Pack of Nodes — my second brand — specializes in HubSpot implementations.
- Salesforce: Enterprise standard. Configurable infinitely. Heavier maintenance.
- Pipedrive: SMB-focused. Simple, fast.
- Attio / Folk: New generation, modern UI, strong in early-stage.
How to choose?
- HubSpot: Best-fit for 80% of B2B scale-ups in NL/EU. Broad ecosystem, reasonable price for growth.
- Salesforce: When you have 100+ employees, complex processes, and RevOps resources to manage it.
- Pipedrive: For pure sales organizations without marketing automation needs.
- Attio: For early-stage with technical founders wanting programmable CRM.
The three GTM-stack archetypes
In practice I see three archetypal stacks working for B2B scale-ups in 2026.
Archetype 1: The Clay-first scale-up
For scale-ups of €2-10M ARR with strong outbound focus.
- CRM: HubSpot Sales Hub Pro
- Enrichment: Clay (with LeadMagic + Findymail + Apollo waterfall)
- Engagement: Instantly + Smartlead
- Signals: Trigify + Common Room
- Orchestration: Clay + n8n
Monthly cost: ~€2,500-3,500. Strong in outbound, weaker in inbound attribution.
Archetype 2: The HubSpot-centric scale-up
For companies wanting to align marketing + sales + service.
- CRM + Marketing: HubSpot Marketing Hub Pro + Sales Hub Pro
- Enrichment: Clay (smaller usage) + Cognism direct in HubSpot
- Engagement: HubSpot sequences + Lemlist
- Signals: Bombora intent in HubSpot
- Orchestration: HubSpot workflows + Make
Monthly cost: ~€3,500-5,500. Strong in full-funnel attribution, less flexible for custom workflows.
Archetype 3: The PLG / data-engineered scale-up
For product-led growth companies centering product data.
- CRM: Attio or Salesforce
- Data warehouse: BigQuery or Snowflake
- Product analytics: Amplitude or PostHog
- Orchestration: Cargo or dbt + n8n
- Engagement: Customer.io + Apollo
Monthly cost: ~€4,000-7,000. Requires developer resources, but delivers disproportionate value for PLG companies.
What does NOT belong in your stack
Tools I regularly see costing money without delivering value:
Too-early all-in-one tools. Apollo, ZoomInfo "because it does everything." Result: mediocre performance per layer, expensive subscription. Better: specialized tools per layer, connected via orchestration.
Expensive ABM platforms for early-stage. 6sense, Demandbase — great for enterprise, overkill for seed/Series A. Start with Common Room or a simple first-party setup.
Tools without clear integration path. A tool without API or webhook doesn't belong in a modern stack. No exceptions.
Generic AI tools for specific GTM tasks. "We use ChatGPT for leads." OK for ad-hoc, fails at scale. Use tools specifically designed for the problem.
How to build your stack — in what order?
The right order for a new scale-up:
Phase 1 (months 1-3): Set up CRM correctly. Lifecycle stages. Base properties. Source tracking. No fancy tools. Lay this foundation first.
Phase 2 (months 4-6): Add enrichment layer (Layer 2). Start with Clay + 2-3 providers. Outbound deliverability foundation. Sequencer (Instantly or Smartlead).
Phase 3 (months 7-12): Signal detection (Layer 3) and orchestration workflows. First AI agents for research and personalization.
Phase 4 (months 13+): Visitor identification (Layer 1) as volume supports it. Advanced attribution. Product data integration for PLG components.
Not the other way around. Too often I see companies starting with visitor and intent tools while their CRM is still chaotic. That's parking sports cars on dirt.
The engineering layer beneath the stack
Finally: the stack itself is just tools. What connects it is GTM Engineering. The waterfall you build in Clay, the signal detection you configure in Common Room, the workflows you write in n8n — that's the work that makes the difference.
Tools change, layers stay. Those who understand the layers can always reselect when a tool becomes outdated. Those who blindly bet on one tool are hostage to one vendor's product roadmap.
Read also Build vs Buy: when to hire a GTM Engineer for the choice between DIY and outsourcing.