TeamWhat We DoCase StudiesToolsBook a consultation →
CASE STUDY. SYNAPTIQ · APPLIED AI. B2B · DEMAND-GEN LOOP

Signal.Compounded.

Synaptiq sells AI implementation to CTOs and Chief Data Officers. Fragmented channel testing. Anonymous traffic. This is how iExcel turned four disconnected efforts into a compounding demand loop.

Engagement
Multi-phase · Ongoing
Scope
Paid · MOFU · RevOps · SEO · SDR
Buyer
CTO · VP AI · Chief Data Officer
Signal loop
Paid → MOFU → RevMethods → HubSpot → SDR
Neural network abstract — signal texture
NEURAL LAYER · SIGNAL TEXTURE
>// initializing demand loop for Synaptiq · applied-AI · CTO-buyer segment    channels=[paid, mofu, seo]   identity=RevMethods   spine=HubSpot   handoff=SDR   state=compounding
// The Squint Test · 3 receipts from the loop
Buyer-intent segment engagement lift
Generative-AI buyer-intent lists opened and clicked at twice the rate of broader AI-topic lists in the same test.
23.4%
Peak vertical click rate
The mid-market legal segment clicked outbound sequences at 23.4% — more than double the account's own 11% blended average.
11×
Peak search CTR vs. published B2B average
One high-intent bottom-funnel ad group converted at 28.57% CTR — WordStream's published B2B search average is 2.41%.
Chapter 01 · The Setup

A CTO won't tolerate noise.

The Synaptiq buyer is not a marketing target. They're a CTO, a VP of AI, a Chief Data Officer — someone who has already seen every framework deck, every "AI transformation" landing page, and every LinkedIn message that opens with "I saw you're working on GenAI…". They evaluate vendors like they evaluate model outputs: with skepticism, with a reference set, and with a very short patience for anything that can't be defended on the technical merits.

Synaptiq's product is applied AI — strategy, implementation, and engineering delivered to enterprises building custom ML systems. The engagement began where most demand-gen engagements begin: a lot of activity, not much accountable pipeline. Paid search running on inherited keyword lists. Paid social firing against loose personas. A newsletter that mailed on schedule but wasn't tied to a nurture ladder. A website that took anonymous traffic in and let it leave anonymous.

The problem wasn't ambition. The problem was that none of the channels were talking to each other — and none of them were feeding a system a CTO would actually respond to. What the business needed wasn't more spend. It was a loop.

AI model interface — technical buyer context
Chapter 01 · The technical buyer
Chapter 02 · The Signal Test

Paid isn't the answer. Paid is the diagnostic.

The first workstream wasn't a scale plan. It was a diagnostic. Applied-AI is a low-volume, high-consideration category — the buyer isn't Googling "buy ML consulting" at 2 p.m., they're comparing partners against a shortlist their engineering team already built. Paid channels can't manufacture that demand. They can only expose it and tell you what the buyer is willing to click on.

We rebuilt the paid stack as an experiment matrix, not a budget line. Every campaign was tagged, every conversion event fired through GTM, and every source landed on the HubSpot contact record with campaign, medium, and content intact. Every keyword and audience cluster earned its lifespan by the quality of what showed up in the CRM after — not by CPM.

What the diagnostic returned wasn't a magic keyword. It was a map: which channels were feeding real technical buyers, which were feeding lookalike noise, and which were quietly delivering the highest-intent traffic no one on the team had been watching. One bottom-funnel ad group, once isolated from a rogue keyword dragging the average down, cleared 28.57% click-through — evidence the technical buyer was already searching, and the diagnostic just had to stop hiding it.

11×
Peak Search CTR · Vs. Published B2B Average
One high-intent bottom-funnel ad group converted impressions to clicks at 28.57% — more than 11x the 2.41% average B2B search CTR WordStream publishes for the category.
Circuit board macro — the diagnostic layer
Chapter 02 · The diagnostic matrix
Chapter 02 · Continued

The paid-media test matrix.

Every channel earned its budget line by what it produced downstream in HubSpot — not by what it looked like in the ad platform. Named tests, real conversion definitions, source integrity from click to contact record.

// GOOGLE ADS
Non-branded technical search. High-intent queries around applied AI, ML implementation, and platform-specific engineering. Filtered against a role-negative list so ads didn't burn budget on students and unrelated searchers. Every click tagged with source, medium, campaign, and landing intent — so HubSpot inherited context, not just a hit.
// LINKEDIN
Role-targeted paid social. CTOs, VPs of AI, Heads of Data Engineering — narrowed by firmographic segment, not just job title. Testing thought-leadership content against direct offer variants to learn what the technical-buyer segment was willing to click on before they were willing to talk.
// META
Lookalike + retargeting. Not a primary acquisition channel — a warming layer. Anyone who touched the site through paid, organic, or newsletter got re-approached with lower-friction education content so the second touch was cheaper than the first.
// UTM DISCIPLINE
Source integrity, end-to-end. Every paid destination URL built from the same UTM schema. Every conversion event mapped in GTM. Every HubSpot lifecycle transition source-tagged at capture. No channel is allowed to claim credit it didn't earn.
Chapter 03 · The MOFU System

Newsletter is not a channel. It's a nurture ladder.

The middle-of-funnel was where the engagement earned its compounding. A CTO who clicked a paid ad twelve weeks ago is not going to remember Synaptiq's name — unless something has been quietly, credibly showing up in their inbox in the meantime.

We built the MOFU as a system, not a slot. The newsletter became the nurture spine — technical, opinionated content aimed at the buyer's actual problem set, not at their industry. Cold outreach ran alongside it, structured against role-segmented lists deliverability-verified before every send. Content assets — playbooks, teardowns, applied case notes — dropped into the sequence at the point where a specific segment was most likely to convert on them.

The delta between cold-first-touch reply rates and nurture-touch reply rates is where the ROI of the whole system lives. The nurture ladder — the newsletter, the sequenced follow-ups, the segmented content drops — is what made the top-of-funnel worth funding at all.

Segment-level testing sharpened the ladder further. A controlled webinar-promotion send split a Generative-AI buyer-intent list against a broader "Other AI Topics" list — the buyer-intent segment opened and clicked at the rate. That signal reset how every subsequent list was built. It also surfaced which verticals were worth the deeper cut: the mid-market legal segment clicked outbound sequences at 23.4%, more than double the account's own 11% blended average, confirming a tighter ICP — not more volume — was the lever that mattered.

Buyer-Intent Segment · Engagement Lift
In a controlled webinar-promotion test, the Generative-AI buyer-intent list opened and clicked at twice the rate of the broader "Other AI Topics" list — the signal that reset how every subsequent list was built.
Code on dark screen — MOFU content layer
Chapter 03 · The nurture ladder
Chapter 04 · The Signal Loop

Four inputs. One spine. A loop that closes.

The reason a demand-gen engine compounds is because every channel deposits into a system the next channel can draw against. This is the Signal Loop as it runs today — paid, MOFU, SEO, and RevMethods visitor-ID all feeding a HubSpot spine, and every named account routing cleanly to the SDR desk. The loop closes when the SDR conversation feeds back into content, and content re-enters the top of the loop.

01

Paid Search + Social

// google · linkedin · meta

02

MOFU Nurture

// newsletter · cold email · content

03

SEO + Content

// technical foundation · editorial

04

RevMethods · Visitor ID

// anonymous → named accounts

// HubSpot Spine

The signal.

Lifecycle · attribution · scoring

05

SDR Workflow

// routing · sequences · handoff

06

Sales Conversations

// discovery · qualification · pipeline

07

Attribution Read

// what fired · what closed · what to fund

08

Content Re-Entry

// sdr objections → new mofu assets

// The loop closes when SDR conversations feed back into content — and content re-enters the top of the loop
Code on screen — the signal layer
THE READ
"A CTO won't tolerate marketing noise. But a CTO will read a good technical newsletter. That's the whole game."
— Engagement Recap · Synaptiq × iExcel
Chapter 05 · The Spine

HubSpot as infrastructure, not a mailer.

None of the loop above holds up if HubSpot is a glorified contact list. The engagement's operational center — the reason paid, MOFU, SEO, and visitor-ID can all deposit into the same system — is the HubSpot lifecycle and data layer we rebuilt underneath it.

Lifecycle stages were re-scoped against how a technical buyer actually moves through evaluation: Subscriber → Lead → Marketing-Qualified → Sales-Qualified → Opportunity → Customer. Every stage tied to a property, a source, and a next-action rule the SDR desk could run against. Custom properties captured firmographic and technographic context — company size, cloud stack, model maturity — so a lead wasn't just a name, it was a qualification signal.

Reporting rebuilt on top. Multi-touch attribution didn't magically appear — it was engineered through the same UTM discipline that governed the paid stack. Dashboards were sized for the CTO-buyer economics: MQL cost trended monthly, SQL rate, pipeline aging. The contact database was normalized against the rebuilt property model — closing the duplicate-record and source-attribution gaps that had been masking true list quality — so every campaign fires against clean context, not a merge conflict. Leadership stopped reconciling spreadsheets and started reading a single view.

Server room / data center — the spine layer
Chapter 05 · The data spine
The engagement, in one line
"Turn fragmented channel testing into a compounding, accountable, multi-channel demand-generation loop for a technical-buyer market."
— Multi-Phase Engagement Recap · Synaptiq × iExcel
Chapter 06 · SEO + Content

The compounding half of the loop.

Paid is the diagnostic. SEO is the accrual account. If paid channels expose what the buyer is willing to click on, organic search is what makes that click cost less every quarter — and eventually, cost nothing at all.

The SEO workstream ran on two lanes. The technical foundation — crawl coverage, indexation hygiene, schema, Core Web Vitals, and internal-link architecture — was audited against Semrush visibility and shipped as a rolling backlog against the buyer's actual query set. The editorial lane was built to the specifier's discipline: applied case notes, model teardowns, and implementation guides — the kind of content the buyer would forward to a colleague, not the kind that reads like a bid deck.

Every published asset was UTM-tagged, indexed by segment, and instrumented as an entry point into the same nurture ladder the MOFU system fed. Search became a feeder — not into a form, but into the loop.

Fiber network abstract — content compounding
Chapter 06 · The accrual layer
Chapter 06 · Continued

Traffic without a landing doesn't count.

None of the upstream diagnostic work pays off if a visitor never reaches a place to act. A session-replay audit (FullStory) on Synaptiq's own site found the drop-off wasn't happening at the top of the page — it was happening between the fold and the form.

On desktop, 85% of sessions reached the page's median fold, but only 33% advanced past the primary call-to-action zone. On mobile, the CTA sat below the fold by default. A high-traffic broken link was pulling a meaningful share of sessions into a dead end entirely.

None of that shows up in a channel report. It only shows up when someone watches the session. Fixing it was the unglamorous work that made every upstream channel's traffic worth more.

33%
Desktop Sessions · Reached The CTA Zone
A session-replay audit found 85% of desktop visitors reached the median fold, but only 33% advanced past the primary call-to-action — the finding that reset the homepage's conversion path.
// The Conversion Read
30%
Click-To-Open Rate Across Outbound Sequences

Of everyone who opened an email, nearly one in three clicked through — proof the segmentation was landing with the right buyer, not just the right inbox.

Chapter 07 · SDR Enablement

The handoff is where most engines quietly break.

An SDR who receives a lead with no context does what a rational SDR does: sends a generic email and moves on. A demand-gen loop that produces context-rich leads and then hands them to a workflow that strips the context is a loop that leaks at the last possible moment.

We treated the SDR workflow as a first-class part of the system. Every marketing-sourced lead landed in HubSpot with the source, campaign, content asset, RevMethods company signal, and lifecycle history the SDR needed to open a conversation that didn't sound like a template. Routing rules moved named accounts to the right rep. Sequenced follow-ups were structured — not scripted — so the first touch respected what the buyer had already engaged with.

The measurable win wasn't a specific touch cadence. It was the fact that the SDR desk stopped treating marketing leads as a separate lane. Every conversation had upstream context. Every objection got captured. Every objection became the next MOFU asset. The loop closed.

Structured architecture — the handoff layer
Chapter 07 · The clean handoff
Chapter 08 · The Stack

The measurement layer under the loop.

None of the campaign work above holds up without a stack that leadership can trust. This is what stayed on — from the first paid diagnostic through the SDR handoff, every workstream instrumented for source integrity and lifecycle continuity.

// CRM SPINE

HubSpot

Lifecycle stages rebuilt against the CTO-buyer motion. Custom properties for firmographic + technographic context. Multi-touch attribution engineered end-to-end.

// VISITOR ID

RevMethods

Anonymous website traffic resolved into named accounts. Company-level signal enriched onto the HubSpot record — turning ambient visits into workable pipeline.

// TAG LAYER

Google Tag Manager

Every conversion event fires through GTM. Every UTM lands intact on the contact record. Every developer handoff includes the tag map so nothing ships blind.

// ANALYTICS

Google Analytics 4

Standing GA4 property tied to source attribution on the contact record. The site's read of what channels are actually working matches HubSpot's — no more disagreeing dashboards.

// PAID SEARCH

Google Ads

Non-branded technical search running against a role-negative list. Every campaign tagged, every conversion instrumented, every keyword earning its lifespan on CRM output.

// PAID SOCIAL

LinkedIn + Meta

Role-targeted LinkedIn against CTO / VP AI / CDO segments. Meta as the warming and retargeting layer. Both wired into the same UTM schema.

// SEO

Semrush

Weekly visibility reporting against the applied-AI query cluster. Competitor keyword footprint benchmarked and re-run each cycle so the editorial calendar has a defensible reason to exist.

// DELIVERABILITY

List Verification

Every outbound audience — every newsletter cohort, every cold sequence — verified before it fires. Invalid, catch-all, and role-based addresses stripped so infrastructure stays clean.

// SDR WORKFLOW

Sequenced Routing

Marketing-sourced leads routed to the right rep with source context intact. Sequenced follow-ups structured against upstream engagement — not generic templates.

Ambient tech interior — the compound layer
THE COMPOUND
"Every channel deposits into a system the next channel can draw against. That's what compounding actually means."
— Multi-Phase Engagement Recap · Synaptiq × iExcel
Chapter 09 · The Read

The playbook for an applied-AI demand engine.

If you sell technical services to CTOs, VPs of AI, or Chief Data Officers — and you're looking at a marketing surface that spends real money on channels that don't talk to each other, a HubSpot instance that behaves like an inbox, anonymous traffic you can't act on, and an SDR desk that treats marketing leads as a separate lane — the Synaptiq engagement is the shape of what to do about it.

Diagnose with paid. Compound with MOFU. Ground everything in a HubSpot spine you can actually attribute against. Identify the accounts already on your site. Close the loop with an SDR workflow that inherits context instead of stripping it. Then let SEO turn the whole system into something that costs less every quarter.

Most agencies sell the campaign. iExcel builds the loop underneath it — and stays through the quarters that make the loop compound.

Whiteboard architecture — the read
Chapter 09 · The read

Diagnose the channels. Then close the loop.

A paid stack you can attribute against. A MOFU ladder that earns second touches at a fraction of the first. A HubSpot spine that behaves like infrastructure. An SDR handoff that inherits context. Book a 15-minute sanity check on your setup — we'll tell you what's leaking and what to do about it.

Book a 15-min sanity check →
Book →