The 5-Step Outbound Stack We Run for Every B2B SaaS Client

Every B2B SaaS client we work with at Understory runs on the same five-tool outbound stack: Clay for data, Claude Code for intelligence, Instantly for email, HeyReach for LinkedIn, and HubSpot as the system of record.

We standardized on these five because they function as one coordinated system rather than five tools sharing a Slack channel. This guide walks through how each layer works, where the integration points sit, and which steps cause the stack to break if skipped.

Why a Single-Channel Setup Stops Working

Single-channel outbound has a ceiling. Most B2B SaaS teams hit it once their pipeline targets outpace what email or LinkedIn alone can produce.

Fragmentation across vendors makes the ceiling lower. One team runs email, another runs LinkedIn, a third manages the CRM. The setup creates more coordination work, inconsistent messaging across touchpoints, and attribution that breaks at every handoff. Pipeline slips because no one owns the full prospect journey.

Running every channel through one workflow removes the coordination tax. The five tools below are the layers that workflow runs on.

Step 1: Clay, the Data Layer

Clay sits at the center of the stack as the enrichment and orchestration layer. It pulls multiple data sources into one interface and runs waterfall enrichment, querying providers sequentially until valid data comes back.

A single Clay table can pull company firmographics, tech stack data, verified emails, and hiring signals together in one workflow, replacing four separate tools and four separate logins.

After enrichment, Clay's AI columns run Claude to generate personalized email openers based on the enriched fields. A recent funding round becomes a specific first line. A CRO hire becomes a trigger-based angle. Those AI-generated variables, stored as fields like {{ai_opener}} and {{company_signal}}, pass directly into the sending tools downstream.

Note that full CRM sync may require Clay's higher-tier plans. Lower-tier accounts can still push to Instantly, but HubSpot sync options are more limited.

Step 2: Claude Code, the Intelligence Layer

Claude Code is where the stack gets smart. The pattern that works best is research first, write second.

Before Claude writes anything, it reviews company context, recent activity, open roles, and other available signals. That context is what makes the copy reference specific details instead of defaulting to generic openers.

We configure a CLAUDE.md file as a persistent context layer for each project. It defines company size, industry, geography, tech stack, funding stage, and target persona. Claude Code reads the file at session start and uses it as a baseline reference for research and copy-generation tasks in that session.

The output is a four-email sequence: a trigger-based opening thread, a short follow-up with a different angle, a new thread with social proof, and a breakup email. Each variation maps to the signal that triggered it. A company that just raised funding gets a different copy than one actively hiring SDRs.

Claude can also work alongside HeyReach to generate LinkedIn connection requests that reference one specific piece of profile data. The message writes to a custom field and adds the lead to a campaign without manual handoff.

Step 3: Instantly, the Email Engine

Instantly handles cold email sending and deliverability, with automated warmup, placement tests, blacklist monitoring, and auto-pauses to protect sender reputation.

Our ramp schedule for new inboxes is strict. We start low in week one and scale conservatively from there. Warmup happens before any cold sends go out. We don't skip this because deliverability is the most fragile part of the stack.

Enriched leads flow from Clay into Instantly campaigns with all their custom variables intact. The {{ai_opener}} Claude wrote in Step 2 plugs into the email template using Instantly's variable syntax, provided the field is mapped as a custom variable in your lead data.

A few non-negotiable technical prerequisites govern every Instantly campaign we run:

• SPF, DKIM, and DMARC authentication on every sending domain
• Plain-text emails with no HTML formatting, tracked links, or attachments
• No more than two follow-ups per prospect
• Spintax to create variation across sends
• Bounce rates monitored daily and held below industry-safe thresholds

These rules keep sending discipline. Full sequences with disciplined follow-up consistently outperform one-off sends.

Step 4: HeyReach, the LinkedIn Channel

HeyReach runs LinkedIn outreach with multi-sender rotation. Campaigns distribute activity across multiple LinkedIn accounts in one workflow, with anti-duplication logic that prevents the same prospect from getting double-contacted.

A typical sequence includes pre-connection warmup actions like profile views, post likes, and follows; the connection request itself; and a fallback sub-sequence if the request goes unanswered. The fallback keeps engaging the prospect through alternative actions instead of dropping them when the connection request stalls.

Cross-channel coordination happens here. HeyReach works alongside Instantly, so unanswered activity in one channel triggers the next step in the other. When a LinkedIn message goes unanswered, the lead enters an email sequence. When a cold email campaign finishes with no reply, the prospect gets added to LinkedIn outreach.

One timing detail matters: don't trigger the email fallback too quickly after a LinkedIn action stalls. Firing email immediately after a LinkedIn non-response lands flat. Build in a deliberate delay before escalating.

LinkedIn outcomes vary by audience and offer in SaaS specifically. The lift from adding LinkedIn to email is real, but expectations should be calibrated to your category and ICP rather than broad platform averages.

Step 5: HubSpot, the System of Record

HubSpot serves as the single source of truth for contacts, companies, deals, pipeline stages, campaign properties, and attribution records. Every other tool in the stack syncs data with HubSpot.

The most important configuration detail is simple. Contacts pushed from Clay should include a campaign name property. HubSpot attribution typically combines campaign associations, UTM parameters, and contact-level properties, so the campaign tag becomes the connective tissue that ties outbound revenue back to the right campaign.

The attribution dependency chain has four breakage points. First, Clay pushes a contact with a campaign tag. Second, a deal gets created and associated with that contact. Third, the deal reaches closed-won with an amount value. Fourth, HubSpot's attribution report filters by the campaign property to connect revenue back to the outbound campaign. A break at any step means outbound-sourced revenue shows up as unattributed.

HubSpot timeline coverage can also break if you rely only on native syncs. Instantly to HubSpot can leave gaps for sends, replies, bounces, and status changes. HeyReach to HubSpot can miss depth on connection requests and message content. We use OutboundSync as middleware to fill those gaps and stream more complete outbound activity into HubSpot contact records, with deals affected only when configured via HubSpot workflows.

HubSpot also feeds back into Clay. Existing CRM records are imported into Clay for re-enrichment, get updated fields, and pushed back with blank-field protection so clean data isn't overwritten. The loop closes.

The Numbers That Matter

Here's the consolidated view based on the available research:

The lift in meeting conversion from coordinated multi-touch sequences over single-channel email is the number that matters most. That's the gap between running these tools separately and running them as one connected system.

Run Your Outbound Stack with Understory

Building this stack is one project. Running it every day is the harder one. Keeping integrations healthy, tuning deliverability, refreshing enrichment data, and holding attribution clean across five tools eats the strategic time SaaS growth leaders should be spending on positioning and pipeline strategy.

This is the stack we run at Understory for every B2B SaaS client. Our Clay-powered outbound and GTM engineering service handles the full stack end to end, from list building and copy generation through deliverability, multichannel sequencing, and HubSpot attribution.

RemoFirst replaced their entire internal SDR team to run outbound exclusively through us. Yofi paused their program after we generated more qualified leads than their sales team could handle.

Book an intro call to see how this stack maps to your pipeline targets.

Frequently Asked Questions

Can this stack work with Salesforce instead of HubSpot?

Yes. Salesforce can replace HubSpot as the system of record without changing the upstream layers. Clay syncs directly to Salesforce, and OutboundSync supports Salesforce as a destination for streaming Instantly and HeyReach activity into contact and account records. The attribution dependency chain works the same way: contacts pushed from Clay carry a campaign property, deals associate to those contacts, and reports filter by the campaign field. The configuration surface inside Salesforce is different from HubSpot's, but the stack logic is identical.

Do we need Clay, or can we substitute a cheaper enrichment tool?

You can run the stack without Clay if you handle enrichment, AI personalization, and routing across separate tools. The trade-off is operational. Clay collapses waterfall enrichment, AI copy generation, and automated routing into one workflow. Substituting individual tools means rebuilding those connections and maintaining them as your data sources shift. For SaaS teams running consistent volume, the time saved usually justifies the spend. For low-volume programs, lighter enrichment can work, though you lose the AI column workflow that drives the personalization in Step 2.

How long does it take to stand up the full stack from scratch?

Setup typically runs four to six weeks if domains and inboxes need to be purchased, warmed, and authenticated from zero. The bottleneck is deliverability infrastructure, not software configuration. SPF, DKIM, and DMARC authentication takes hours, but inbox warmup runs on a multi-week ramp before any volume sends safely. Clay tables, Claude Code project files, Instantly campaigns, and HubSpot properties can be configured in parallel during the warmup window. Teams already on warmed inboxes can be sending coordinated sequences within two weeks.

Can a small SaaS marketing team run this stack in-house?

Yes, with the right ownership. The stack requires one person who owns deliverability and one who owns Clay table architecture and copy quality. Both roles can sit with the same person for low-volume programs, but they cover different skill sets at scale. The work that consumes the most time is not the initial build but the ongoing maintenance: inbox health, enrichment refreshes, attribution audits, and copy iteration. Teams without a dedicated outbound owner usually find the maintenance load is what pushes them toward an agency partner.