Case Study: How Coding Dojo Defied the Growth Plateau with Smart Automation and Process Overhauls
The story of how the revenue org broke through a growth plateau with smarter processes, automation, and seamless data piping between marketing and sales.
Coding Dojo had hit a wall.
They were a growth stage startup, bootstrapped. A team of 200. They were investing a quarter-million a month on paid ads, had hired a new head of sales, but it wasn’t showing up in revenue.
What happened next is a lesson in how “revenue architecture” - or rethinking data, process, and tools together — can transform your trajectory. I think most marketing teams are too-focused on the upper-funnel. That makes sense - it’s mostly within our control. But a lot of the leverage gets lost in the hand-offs from marketing to SDRs to AEs. By architecting the revenue funnel with a first-principles, aligned approach, we drove significantly more revenue from the same media spend. Below are the five specific optimizations we made.
The Problem: Stuck in the Growth Plateau
Coding Dojo had a solid product—offering high-ticket online coding courses priced around ~$15K. But despite investing more media spend and expanding the sales org, revenue was flat.
When I first started advising the CEO Richard Wang and mentoring the VP of Marketing Stephen Sinco, it became clear that something else was going on beneath the surface.
The marketing and sales teams seemed to be collaborating well, but in terms of real-time data and process, they were actually operating in silos.
This disconnect is a silent killer of growth.
According to a study by HubSpot, businesses with strong marketing-sales alignment achieve 208% more revenue from their marketing efforts. Not only that: “Organizations with good alignment not only achieve 27% faster three-year profit growth, they also close 38% more deals.”
The Complexity: When Leads Become a Dead End
When I spoke to the VP of Marketing, he told me about the volume of leads they were generating and the number of “application forms” being filled-out.
But here’s the rub: marketing had no visibility into what happened to the leads after that.
Meanwhile, the head of sales complained lead quality was declining, but couldn’t really show in any structured way that the quality had changed.
Another challenge: in this sort of business, the sales cycle could be weeks or months. That’s too long of a feedback loop for marketing to optimize on (we’ll come back to this problem later).
The Solution: The Power of Faster Feedback and Smarter Automation
The answer wasn’t in getting more leads or more applications. It was in identifying the moments that truly mattered in the customer journey, starting to track those moments in a structured way, and move users through those moments faster.
Optimization 1: Starting to Measure the Right Moments
The first key insight: nobody spends $16,000 without talking to a human being.
After a prospect submitted an application, they were invited to book a sales meeting packaged as an “interview.”
So our first order of business was to start measuring scheduled interviews and completed interviews.
Optimization 2: Tightening the Feedback Loop from Sales back to Marketing
The sales team (like every sales team 😉 ) was complaining about the quality of leads. So we worked with them to create a more rigid definition of what constituted a Sales Accepted Lead (SAL).
We introduced 4 qualification questions every rep was required to ask. To be considered a SAL, they had to meet 3 out or 4.
Next, I created a new CRM Opportunity Stages for Interview Scheduled, Interview Completed, and SAL.
To get the sales team using these new stages, we created hygiene dashboards and would chase-down anyone whose interview date had passed whose status hadn’t changed.
The result was a powerful new feedback loop: now instead of finger-pointing between marketing and sales, there was a shared framework for what “qualified leads” meant.
And instead of having to wait 2-12 weeks to know if a lead had become a customer, the marketing team could see within 1-2 weeks of lead creation how many leads had converted to SALs.
I also worked with the marketing team to overhaul their reporting dashboards.
We added Interviews Scheduled, Interviews Completed, and SALs to marketing’s dashboards. Previously, their success was measured by the number of Leads, Applications, Cost per Lead (CPL), and Cost per App.
But I stress-tested the new SAL field. And sure enough, the conversion rate to Customer was dramatically higher when someone was a SAL. So rather than Lead and Cost per Lead, SALs and Cost-per-SAL became the new north-star metrics for marketing.
It took a few months to retool the entire reporting system, but eventually, we could break down CP SAL for every channel (i.e. Paid Search), every campaign (i.e. Competitor Terms), adset (i.e. “General Assembly”), and even individual Google Ads keyword (i.e. “General Assembly tuition”).
Optimization 3: Massive Unlock: Better Data Feedback Loop for Meta Ads
This next bit resulted in its own double-digit improvement in revenue/CAC:
Instead of optimizing for Leads, we were now optimizing our campaigns for SALs. But this was happening through manually adjusting spend. We had a hypothesis: could we train the Meta algorithm to optimize for quality too?
Context: Normally when you drive ad traffic to your site, you’ll fire conversion event back to Meta when someone submits a lead form. Essentially you’re training the algorithm “Hey algo, you scored a point for that lead! Good job, go find more of those.”
But we don’t really care about leads, we care about customers. Why not train our friendly algo for something lower-funnel that’s more highly correlated to customer?
We did this using something called “server-side conversions,” triggering a conversion event for every Interview Scheduled, Interview Completed, and SAL.
There wasn’t enough weekly volume of SALs for the algorithm to learn from, so we told the algo to optimize for Interviews Scheduled.
Next, we designed a proper split-test:
For 50% of paid spend, we told the algo to optimize for Leads
For the other 50%, we told the algo to optimize for Interviews Scheduled
We had to run it for 6 weeks to get to stat-sig. But the results was massive: for the same spend, leads went down ~10%, Interviews Scheduled went up ~24%, and Customers increased ~12%.
Optimization 4: Speed-to-Lead to Increase Win-Rates
The upper-funnel was now firing on all cylinders. But we had a new mid-funnel problem emerging.
Prospects would click an ad, fill out the lead form and schedule an Interview all within minutes, but the next available Interview appointment might be a week away. And the further away it was, the worse the Interview show-up rates seemed to be. They’d dropped from 60% to 30%.
We couldn’t hire more sales reps, so we needed to find a way to make sure only the best leads made it to the sales team.
To address this, we implemented Chili Piper, a smart-routing tool that prioritized leads based on their qualification.
We created three distinct “lanes” to manage the flow:
Fast Lane: Highly qualified leads could request an immediate callback from a sales rep
Medium Lane: Moderately qualified leads could book an appointment within a few days
Slow Lane: Lower-quality leads were routed away from sales to a webinar instead
Next, we shifted to a two-step lead form. We added those BANT questions directly into the form. Now we could qualify SALs without talking to a person, and dynamically route SALs to the fast lane and non-SALs elsewhere.
We ran this one one as a split-test too, and found a 20% improvement in No-Show Rates.
Optimization 5: Meeting People “When” They Are
We made another discovery pertaining to show-up rates: we weren’t available “when” customers were thinking about us. This whole time the sales team had been working normal US business hours. But if you looked at the distribution of when leads were coming in, it was quite a lot of nights and weekends, when sales wasn’t working.
You could see this show up in the conversion rates to SALs by day-of-week and time-of-day.
After presenting this data, we managed to come to an agreement with our Director of Sales. We couldn’t expect existing hires to shift their schedules dramatically, but for all new sales hires we’d set expectations they’d need to work nights and weekends. Eventually we moved the entire team to shifts so we’d have coverage 13 hours per day, 7 days a week.
The Results: 2X Revenue Growth and an Acquisition
By improving the hand-off between marketing and sales, introducing automation, and speeding up lead management, Coding Dojo saw a dramatic improvement in revenue:
Lead-to-SAL conversion rate: Skyrocketed as leads were prioritized and engaged more quickly.
Time to first call: Moved the average time from Form Submission to Call Completed from 7 days to 4 days.
Win rates: Increased substantially. Taken together, all these changes resulted in more than doubling our revenue run rate, all while keeping the Customer Acquisition Cost (CAC) steady.
Right around this time, Coding Dojo began entertaining acquisition offers and eventually sold to Perdoceo Education.
Conclusion: More Leads Is the Wrong Answer. Growth Comes from Working the Leads Smarter, Together
What Coding Dojo teaches us is this: a marketing approach that focuses exclusively on upper-funnel is doomed to be inefficient. Todays marketers must be full-funnel.
We formed a cross-functional pod of marketers, sales folks, rev ops, and data analyst. We spent over two years optimizing the customer journey, shipping iteratively week after week.
By rethinking data, process, and tools we transformed Coding Dojo’s trajectory—not through launching new marketing channels or bringing in more leads, but through measuring shared metrics, working the leads smarter, and through marketing and sales alignment.