The same demand generation framework ran at two organizations in two years: Uplift Education, a 23,000-student charter school network, and iPEC, one of the largest ICF-accredited coaching certification companies in the world, 25,000+ graduates generating $500M+ in coaching revenue. Two different verticals, the same architecture, both built from nothing.
A full demand engine from nothing: HubSpot automation, intent-based scoring, multi-channel acquisition across three revenue streams, and a B2B intent pipeline built with n8n and ZoomInfo. No inherited infrastructure.
6,913 MQLs, 1,557 meetings, 316 deals, 5.95x ROAS — all maintained through a 58% mid-year budget cut. The same architecture ran at Uplift Education the year before with the same results.
The same architecture ran at two organizations in two different verticals, coaching certification and K–12 education, both built from nothing. The trigger logic, scoring model, sales alignment, and spend discipline aren't tied to what's being sold. They transfer to any considered-purchase funnel.
iPEC had strong product, an engaged community, and a $3.3M enrollment goal. Marketing infrastructure to hit it: none.
This wasn't the first build. The year before iPEC, I ran the same playbook at Uplift Education, a 23,000-student, 20-campus charter school network with a $20M+ per-pupil funding target tied to enrollment. HubSpot automation built from scratch, behavior-based lead scoring, dynamic audience segmentation. Result: 4,000+ enrollment applications and 75% student re-enrollment rate. Different product, same system.
The iPEC marketing function wasn't one campaign type. It was three distinct businesses running in parallel, each with separate buyers, offer structures, sales motions, and campaign cadences.
High-consideration, long-cycle enrollment product. Buyers research 3–6 months before committing. Webinar-anchored campaigns, behavioral lead scoring, admissions handoff, and multi-touch attribution across 3–4 cycles per year.
Direct-to-coach assessment tools sold via Shopify-style checkout. Two SKUs (Self $60, 360 $119) plus two courses ($210/$299). Promotional flash sales layered onto the enrollment calendar, each with dedicated email, social, and landing page campaigns.
Corporate workshop program targeting HR leaders and L&D buyers. $1,500 live virtual sessions for 12–50 attendees. Supported with a dedicated sell sheet, B2B landing page on ipeccoaching.com, and a dedicated sales rep pipeline. Separate buyer persona, separate campaign motion.
Every piece of the campaign was designed to trigger from behavior, not schedules, not gut feel.
Three behavioral segments. Three distinct nurture tracks. Automatically routed by HubSpot based on real interaction signals.
AI-assisted lead scoring tied to behavioral thresholds, not time-in-funnel. Score hit means sequence enrollment and BDR task, simultaneously.
Five channels running in parallel, all tied to meetings booked and deals won, not cost per click.
The B2B side had no inbound. Tanya was working a cold list. I built her a better one. n8n scraped job boards daily for companies posting Leadership Development or L&D Manager roles. Those postings are a reliable signal: the company is actively investing in people development, which is the same budget that funds coaching workshops. ZoomInfo enriched each hit with firmographics, org depth, and the right contact to call. When a company cleared the size and industry threshold, n8n dropped a HubSpot task with the job posting attached as context. Tanya's queue went from a spreadsheet to a prioritized list where every row had a reason to reach out.
A single quarter of results isn't a system. I built this to run at Uplift Education and then brought it to iPEC with minor adjustments. A new vertical or product line could be onboarded to the same architecture in a few days.
Vertical, use case, offer, channel mix, and measurement plan in one doc. Co-authored with sales before launch, not distributed after.
Behavioral signals defined before campaigns launch, not added post-hoc. Each threshold maps to a specific sequence entry and BDR action.
Sales in the room. Pipeline contribution as the metric: not MQLs, not CTR. Every channel decision made with field signal from reps.
Which signals predicted conversion, which didn't. Reweighted scoring model after each cycle. Attribution agreed with RevOps before launch.
Late-cycle CTP enrollment campaign, March 2026. A webinar-anchored, multi-channel push across paid search, email, social, and landing pages over a 3-week enrollment window.
The behavioral segmentation built in HubSpot is why this happened: each contact got the right message at the right moment, not a broadcast.
Static ad creatives produced across concurrent campaigns running alongside the Elevate webinar, each mapped to a distinct audience segment and funnel stage.
The reporting layer I built to connect every channel to closed deals. Captured March 22, 2026, three weeks into the late-cycle enrollment window.
Mid-year. Without warning. Pipeline commitments unchanged.
The wrong response: reduce spend evenly, maintain breadth, hope for the best. That burned two weeks before I stopped it.
Ran funnel conversion analysis by channel. Which dollars were producing closed deals, not MQLs, not meetings, deals.
Cut Pinterest, reduced Meta, concentrated spend on Google Ads and top HubSpot sequences. Leaned into organic + partnerships as zero-cost layer.
Weekly optimization reviews became non-negotiable. Every channel decision justified by pipeline data, not media habits.
"Coaching certification" was our highest-spend keyword at QS 3 with $8,600 in spend. Identified that improving to QS 7 would cut CPC by ~64%, saving $5,500 annually at existing volume.
Against a $3.3M enrollment program. 4-person team. $300K annual budget, cut to $126K mid-year.
This system wasn't built for one industry. It ran at Uplift Education, a 23,000-student school network, and at iPEC, a global coaching certification company: two different verticals, the same architecture, both built from nothing. The mechanics don't care what you sell. Here's how each component transfers to a new business.
| Component | What I built at iPEC | The transferable principle |
|---|---|---|
| Intent Scoring | Custom intent pipeline: job board monitoring (n8n + crawl4ai) and ZoomInfo enrichment feeding scored HubSpot tasks to the sales rep, with the triggering evidence attached. | Score on behavior, not time in funnel. Whatever signals predict a purchase in your business — page paths, content downloads, product usage, renewal windows — feed the same trigger model. |
| Multi-Stakeholder Routing | Three buyer types (individual candidate, L&D sponsor, admissions decision-maker) on separate HubSpot tracks triggered by role signal. | Most considered purchases involve more than one decision-maker. Each role gets its own track, sequenced by where they sit in the decision, instead of one broadcast to everyone. |
| Sales Handoff Logic | Score threshold hit = a sales task created in HubSpot with context. No manual queue, no lag between signal and outreach. | Signal hits the threshold, the system creates a sales task with context attached, automatically. No manual queue, no lag between intent and outreach, in any CRM. |
| Spend Decisions | 58% budget cut: pulled conversion data by channel and stage, cut channels that generated MQLs but not deals, concentrated on what closed. | Cut what generates activity but not revenue, concentrate on what closes. The discipline holds whether the budget is growing or gets cut 58% mid-year. |
| Sales Alignment | Weekly pipeline reviews with Admissions: every channel justified by meetings booked and deals won, not MQL volume. | Marketing and sales in the same weekly review, measured on pipeline contribution rather than vanity metrics. Every channel decision made with field signal from reps. |
| Content Architecture | Replay outperformed invite (15.74% vs 0.5% CTR): on-demand content sequenced by behavioral signal, not a broadcast calendar. | On-demand, behaviorally-sequenced content beats broadcast for any considered purchase. The right message at the right moment, triggered by what each contact actually did. |
Map where the funnel leaks. Identify what marketing contributes at each stage versus what sales absorbs manually. Establish an attribution baseline before touching spend.
Define the scoring model and segment tracks with sales. Map the behavioral signals that predict a purchase in this business to CRM properties. Wire trigger logic to the score thresholds.
Run one campaign end to end against a real target segment. Measure by pipeline movement and conversion, not impressions. Prove the model before scaling it.
Edward Chalupa · Demand Generation & Marketing Operations · Dallas, TX