- Headline result
- Tickets -75%
- Client
- Sweden's largest motorcycle education platform
- My role
- Operations Lead
- Scope
- 25,000+ students, 130,000+ annual visitors, WordPress + Shopify + ActiveCampaign
- When
- 2018 - now
The problem
When I took over operations in 2018 the platform already had thousands of students and a lot of moving parts. Email deliverability sat around 65%, so a third of every campaign never reached an inbox. Open rates were around 30%. Support ate most of the week: 663 tickets that year, most of them the same handful of questions about access, login and course delivery.
The tool stack had sprawled. Every question about the business meant logging into a different tool. Traffic in one place, sales in another, email in a third. Onboarding was partly manual, so a student who bought at 11pm waited for a human in the morning.
Nothing was on fire. It was the slower problem: a business doing well in spite of its systems rather than because of them.
What I did
I run the back-end as Operations Lead, so this is not one project. It is eight years of fixing the thing that costs the most, then the next thing.
Email first. I ran the CRM migration from GetResponse to ActiveCampaign end to end: list hygiene, segmentation rebuilt from scratch, every automation re-created and tested before the switch. Then onboarding and course delivery. Purchase → access to the course portal → welcome sequence → customer nurture, all automated. Shopify, LearnDash, ActiveCampaign and GoHighLevel wired together so a student gets access within minutes at any hour.
I built and manage the LMS on WordPress with LearnDash. I rebuilt the checkout flow. I consolidated the stack, which cut tool spend 60%. For support I built FAQ content and automations around the questions that kept coming back, so the ticket count dropped year after year while the student base grew.
In 2026 the work moved to data and AI. I built a Supabase warehouse that pulls GA4, Search Console, Shopify, ActiveCampaign, GoHighLevel, SamCart and the old analytics history into one place, with a private Next.js dashboard on top. I also re-platformed the 48-page email sales funnel from WordPress to Next.js as an exact replica, with GDPR consent and per-lead tracking built in and a monitor that checks all 48 pages every five minutes.
How it works
Fix deliverability before anything else
Every email improvement is fake if a third of sends never land. The migration started with list hygiene and the real opt-in history, then segmentation, then automations. Deliverability went 65% → 93%, which is the only reason open rates could double afterwards.
Automate the purchase to access path
A student who buys gets portal access, the welcome sequence and the right tags without a human in the loop. That one change took 10+ hours a week off the team and removed most access tickets at the same time.
Answer the recurring question once
Tickets were counted and categorised, not just handled. The same five questions kept coming back, so each one got a fix at the source: an FAQ, an automation or a clearer page. 663 tickets in 2018 → 167 in 2025.
One stack, not ten tools
Every tool had to earn its seat. Overlapping tools were cut or merged into the ones that did the most. Tool spend fell 60% and there is now one place to look for each answer.
Own the data
The 2026 warehouse reads every platform through read-only APIs into Supabase and the database refreshes its own rollups. The dashboard shows traffic, sales, email, SEO and ads in one view. A Google Sheets mirror gives the team the numbers without a login.
Results
- Email deliverability 65% → 93%, open rates 30% → 65%
- Support tickets -75%: 663 a year in 2018 → 167 in 2025
- Checkout conversion +18%
- Tool spend -60% after consolidating the stack
- Onboarding and course delivery automated, 10+ hours a week saved
- Funnel emails now open at 68% with 40% clicks
- 7 tools and 12 years of history in one live dashboard, refreshed automatically
- 48-page sales funnel re-platformed to owned code with a 5-minute outcome monitor
What this means for your business
Most online education businesses I see have the same shape: good product, growing list, a stack that grew by accident and a founder answering the same support question for the fifth year. None of it needs a rebuild from zero. It needs someone to own the back-end, fix the most expensive thing first and keep going. If your deliverability, support load or tool sprawl looks like 2018 did here, tell me where it hurts most and I will show you how I would fix it. Happy to grab a call to look at your setup.
Tools
- ActiveCampaign
- GoHighLevel
- WordPress
- LearnDash
- Shopify
- Zapier
- n8n
- Supabase
- Next.js
- Claude Code
Skills
- Operations Manager
- Marketing Operations
- Email Marketing
- Email Deliverability
- CRM Migration
- Marketing Automation
- LMS Management
- Customer Support Automation
- Checkout Optimization
- Business Process Automation
Questions
How do you improve email deliverability for an online course business?
Start with the list, not the copy. Clean it, confirm who actually opted in, remove dead addresses and segment by behaviour. Then rebuild the automations so people get emails they expect. On this platform that took deliverability from 65% to 93% and open rates from 30% to 65%.
How do you cut support tickets without hiring more support staff?
Count and categorise every ticket first. A handful of questions usually make up most of the volume. Each one gets a fix at the source: an automation, an FAQ or a clearer page. Here that cut tickets 75% between 2018 and 2025 while the student base kept growing.
What does an Operations Lead do at an online education platform?
Everything between the marketing and the student. CRM, email, the course platform, checkout, support, reporting and the automation that connects them. I run it hands-on and I build a lot of it myself, these days in Claude Code.
Is a custom dashboard worth it for a small business?
If you log into more than three tools to answer one question, yes. The warehouse here pulls 7 tools into one view and refreshes itself. In the first weeks it surfaced a 2 to 3x organic recovery opportunity nobody had seen because the data lived in separate tabs.