---

Scaling Web and Digital From Conversion Turnaround to Enterprise Pipeline Engine

Scaling global web and digital demand generation through a $50M → $450M ARR growth journey

The situation

Smarsh, a PE-backed compliance technology company selling into financial services, government, and other regulated enterprise buyers, was scaling fast, organically and through repeated acquisition. Web and digital had to grow from a corporate asset into a measurable pipeline channel, without breaking under the complexity of global sites, 2,500+ domains, and long, multi-stakeholder sales cycles.

My role

I owned global web and digital demand generation for nearly nine years: strategy, execution, budget, and the number.

Phase 1: Fix the foundation (first 6 to 9 months)

The initial focus was fundamentals: rebuilding conversion paths, tightening landing page experiences, aligning campaigns to pages, and using analytics to find where traffic was failing to convert. That optimization phase produced a clear step-change:

• Doubled website conversion rates and tripled average time on site, with significantly reduced bounce rates
• Cut average CPA from ~$1,200 to ~$400 in Google Ads and from ~$2,200 to ~$800 in LinkedIn Ads

Phase 2: Scale the engine

Those early gains became the foundation for a much larger demand engine, built over the following years:

• A web estate run as a revenue channel, not a brochure. Global websites and 2,500+ domains consolidated, governed, and instrumented so every visit could be traced toward pipeline.
• $3.4M+ in annual paid media across Google Ads, LinkedIn, retargeting, and ABM (plus a $1.2M martech portfolio) reallocated continuously based on attribution and pipeline coverage.
• Full-funnel digital programs across paid search, paid social, ABM, and lifecycle, planned and deployed against pipeline coverage targets agreed with sales, not impression goals.
• Attribution leadership trusted. Reporting built in GA4, Looker, Salesforce, and HubSpot connecting spend → conversion → pipeline → revenue, reviewed with executives on a standing cadence.

The results

• ~$700M in digital-influenced pipeline across tenure, including $127M in 2025 alone
• Supported the company’s growth from ~$50M to ~$450M ARR
• Web established as one of the company’s largest and most measurable pipeline channels
• Repeatedly absorbed acquired brands into the architecture without losing pipeline continuity

A note on measurement: because this work spanned years of company growth, acquisitions, and changes in traffic mix, I separate the early optimization metrics from the long-term scale metrics. The conversion, engagement, and CPA improvements reflect the initial turnaround period; the pipeline and ARR figures reflect the later scale of the engine.

What I’d do differently

Push personalization and intent-data activation earlier. The infrastructure could have supported account-level web experiences a year or two before we prioritized them. That’s pipeline we left on the table.

---

Building an outbound demand engine from zero

The situation

A data-driven marketing company needed an outbound demand generation function where none existed: no sending infrastructure, no campaign system, no CRM architecture to track what outbound produced, and no honest way to model what outbound could contribute to revenue. I owned it end to end, from the first inbox to pipeline reporting.

What I built

• The full outbound operating system: sending infrastructure, deliverability controls, and cross-campaign suppression, scaled from 19 inboxes at launch to 64 once I identified send capacity, not daily cap settings, as the real scaling constraint.
• A multi-ICP campaign architecture across 8 ICPs, AI-driven with human approval: audience strategy, research, per-email sequencing, and QA, every campaign in co-pilot mode with review before send.
• Data-quality discipline as a gate, not a suggestion. I traced a failing campaign’s 33 to 64 percent bounce rate to disabled email verification that had been misread for months as a persona and messaging problem, then made list verification a hard requirement before any send.
• CRM architecture and reporting that separates influenced from sourced pipeline, replacing activity metrics with numbers leadership can trust.
• An outsourced BDR calling program on top: phone-data infrastructure, routing audits, reporting dashboards, and reply-and-call coaching for a five-rep pod.

The results

In the first 90 days from first send:

• $204K in influenced pipeline, built from zero.
• $7,300 of pipeline per positive reply. The engine converts engagement into pipeline efficiently, which told me volume, not efficiency, was the next lever to pull.
• 12 meetings requested and 331 accepted connections, a warm layer that does not show in reply metrics but banks contacts to work later.
• Two campaigns proved the model when sourcing and messaging aligned: one at a 1.5 percent reply rate, one with excellent sourcing health. That pattern is what the rest of the program now replicates.

What I’d do differently

• Enforce data verification as a hard gate from day one. The single biggest drag was a data-quality failure hiding as a messaging problem. I caught it, but later than I should have, and it cost weeks of misread results.
• Consolidate sooner. I let campaign count sprawl past 20 before collapsing it to 5 or 6 and cutting message length roughly in half. Fewer, sharper campaigns beat more campaigns, and I would hit that discipline earlier next time.

---

ABM That Sales Actually Used: Pipeline in Regulated Enterprise Accounts

The situation

Regulated enterprise buyers do not respond to volume marketing. The target set was the top 100 global banks: the longest cycles, the largest buying committees, and the most risk-averse stakeholders in the market. Legal, compliance, and IT all sit on the buying committee, and any one of them can stall a deal on its own. The job was to build ABM that created pipeline in named accounts and that sales trusted enough to build their own plans around.

My role

Account selection frameworks, sales alignment, and the digital execution layer: paid social, retargeting, web conversion paths, and pipeline measurement.

What I built

• Account selection with sales, not for sales, priority accounts agreed jointly, so the target list was a shared commitment rather than a marketing artifact
• Layered digital coverage on priority accounts: paid social, retargeting, and tailored conversion paths mapped to committee roles
• Progression measurement: account engagement tracked through to opportunity creation and deal progression in Salesforce, the metric sales cared about
• Regulated-buyer messaging: compliance-literate content that survived scrutiny from risk-averse stakeholders
• Persona-mapped campaign architecture: separate demand gen tracks for legal, compliance, and IT, with assets matched to each persona at each funnel stage, so every committee member met content built for their specific concern instead of generic brand messaging.

The results

• Enterprise funnel velocity increased: MQL to Closed Won cut from 31 months to 23 months for enterprise accounts, an 8-month acceleration in the longest, most complex deals.
• Engagement and pipeline progression inside some of the most demanding accounts in the world, including multiple of the largest global banks.
• Sales adoption: account plans built around the program’s engagement data, the signal that ABM was trusted, not just delivered.
• Measurable engagement-to-opportunity progression in top-tier enterprise accounts.

What I’d do differently

Tighten the influenced-pipeline definition from day one. ABM attribution invites skepticism; the stricter the definition, the more credible the number.

---

Growth that survives a privacy complaint

The situation

Modern demand gen runs on tracking, data enrichment, and consent. In regulated markets, that surface area is a liability the moment it is not handled correctly. At a data-driven marketing company, a formal third-party privacy complaint alleging the website ignored cookie opt-outs escalated to the CEO within hours. In parallel, a new state data-broker deletion law was days from a hard compliance deadline. I owned the technical side of both.

What I did

• Reproduced the reported failure independently in clean sessions with full network captures, producing stronger evidence than the original complaint and pinning the root cause: the consent banner recorded visitor choices, but no tracking tag actually read that consent, with tags injected through three uncoordinated paths.
• Wrote the formal internal findings: a tag-by-tag remediation plan with named owners, the two-part fix (block first, then gate on consent), and a post-fix verification protocol.
• Rebuilt the setup on a single consent platform with modern consent-mode signaling, replacing the parallel configuration that caused the failure, and owned the verification testing.
• Caught a silent failure the scans missed: the new consent gate was likely blocking the privacy request form itself, which meant legally required deletion requests could be failing quietly. I forced a pre-deadline test and moved the deletion channel off the marketing CRM entirely.
• Readied the CRM for the data-broker deletion law: a shared suppression list, suppression checks at every enrichment point, and deletion-to-suppression integration so deleted consumers cannot be re-added through data enrichment.

The results

• A clean, defensible consent architecture that legal, the CEO, and an external complainant could all stand behind. The accept-path verification scan passed with zero violations across 982 pages.
• Deletion-request intake fixed before the deadline, closing a silent-failure risk that could have meant legally required requests going unanswered.
• Growth infrastructure, from tracking to enrichment to CRM, that operates inside the compliance line by design rather than after a fix.

For buyers in financial services, government, and compliance, this is table stakes. Get it wrong and the growth engine becomes legal exposure.

---

Case Notes

M&A web and demand integration. Repeatedly integrated acquired companies (sites, domains, redirects, tracking, paid traffic, SEO equity, conversion paths) into a global web and demand architecture without losing pipeline continuity. If you’re PE-backed and acquisitive, you know exactly why this matters.

Outbound data-quality diagnosis. Outbound diagnosis discipline: before changing messaging, I verify deliverability and data quality first. Root-cause analysis over rewrites; kill decisions made on evidence, not hope. This diagnostic order has repeatedly located the real constraint where creative changes had failed to.

CRM lifecycle and reporting hygiene. Rebuilt lifecycle stages, field architecture, and reporting so that pipeline numbers meant the same thing to marketing, sales, and the board.

---

What This Means for You

I am not offering a framework I read about. I am offering the systems I built and ran inside one of the fastest-scaling companies in its category, adapted to your stage, budget, and team.

Interested in discussing how I build pipeline engines? Let’s connect.