I want to tell you about the day I deleted our lead-scoring spreadsheet.

It was a Tuesday in November 2022, the same Tuesday, as it happens, that ChatGPT was released. I was a few months into building a regional marketing function for a global technology company entering Southeast Asia for the first time. I had a 25-person team. I had a CRM with custom fields. I had a sales pipeline I could filter by lead source, demographic, psychographic, and the average temperature in Jakarta on the day the lead was captured. I had a CAC that was high enough to be visible from the regional headquarters.

I had, in other words, the modern marketing operation. And it was not working.

The leads were coming in. The sales team was working. The conversion rate was, by industry standards, normal. The cost was high, but it was a high-intent market, and high-intent markets are expensive. The deck I gave to the regional leadership had all the right charts. The CAC trend line was bending in the right direction. The forecast was positive.

The only problem was that none of the people we were acquiring looked like the people who would actually complete the program and get hired. We were buying intent, and intent, it turned out, was the easiest thing in the world to fake.

The Number That Was Lying

I want to talk about CAC for a moment, because CAC is the most quoted and least understood number in marketing.

CAC stands for Customer Acquisition Cost. It is calculated by dividing your total marketing spend by the number of customers you acquired. It is the metric that every CMO reports to every board, that every growth team optimizes toward, that every agency promises to reduce.

It is also, in most cases, a beautifully precise measurement of the wrong thing.

When we reported a high CAC at the SEA marketing function I was leading, what we were actually measuring was the cost of acquiring a registration. The person had filled out a form. They had clicked a button. They had, in some cases, scheduled a call with our sales team. What we were not measuring — what no spreadsheet I knew could measure — was the cost of acquiring a graduate who got hired.

The distinction matters. It is the difference between a lead and a result. It is the difference between a marketing function and a marketing institution.

When we finally rebuilt the model around outcome, the numbers told a different story. The cost of acquiring someone who completed the program and received a job offer within six months was, in some cohorts, more than three times our reported CAC. We were not buying customers. We were buying churn.

The Workshop That Changed the Model

In February 2023, we ran a Saturday workshop. It was free. It was titled UpgradeKarir — Upgrade Your Career. We expected 40 attendees. We got 350.

The room was full of working professionals in their late twenties and thirties. They were not the “intent” audience our paid campaigns had been chasing. They were not the ones clicking on “Become a Data Analyst in 6 Months” banners. They were the ones who had been thinking about it for two years, who had colleagues who had done it, who had quietly decided that their current trajectory was not working and they needed to do something, but who would never, ever fill out a form.

We did not capture 350 MQLs that Saturday. We did not push them into a funnel. We did not score them. We did not add them to a nurture sequence. We had a conversation. We answered questions. We served coffee. We let them talk to each other.

By the end of the year, the WhatsApp group we created for that workshop had grown to several hundred members. By the end of the quarter, it was in the low thousands. By the end of the year, it was the primary source of qualified applications for the program — qualified, in the outcome sense, not the form-fill sense.

The conversion rate from community member to enrolled student was meaningfully higher than the conversion rate from paid lead to enrolled student. The cost per enrolled student from the community was a fraction of the cost per enrolled student from paid. The graduates from the community cohort had a higher completion rate and a higher placement rate, because they had been choosing us for some time before they ever clicked a button.

We had, without intending to, discovered a different model.

What Conversations Buy

I want to be careful here, because this is the part of the story that gets misinterpreted. I am not arguing against performance marketing. Performance marketing is a useful tool for buying intent at scale. I have spent the last twenty years building performance marketing functions. I know how to optimize them. I know what they can and cannot do.

What I am arguing against is the assumption that intent is the only thing worth buying.

Conversations buy something different. They buy trust. They buy education. They buy the slow accumulation of belief that happens when someone has been in a community long enough to see other people succeed. They buy the moment when a person who has been quietly considering a decision for two years finally raises their hand — not because they saw a banner, but because they trust the people in the room.

This is not a marketing tactic. It is a marketing institution. It is the difference between renting attention and building community. It is the difference between buying leads and building the conditions under which leads raise their hands voluntarily.

In higher education, this distinction is not optional. It is structural. A student choosing a university is making a multi-year, multi-hundred-thousand-dollar decision under conditions of intense uncertainty. They cannot test the product before they buy it. They cannot return it if it does not work. The signal they are looking for is not “this program looks good.” The signal they are looking for is “people I trust have made this choice and it worked for them.”

Performance marketing can produce the first signal. It cannot produce the second.

The Leadership Lesson

The hardest part of this shift, in retrospect, was not the marketing. It was the leadership.

When I took the conversation-based model to my own leadership, I was told — politely, reasonably, and with the best intentions — that conversations do not appear in dashboards. Conversations do not produce weekly reports. Conversations do not scale in the way the regional leadership needed them to scale. The board wanted a number. The number we had was a CAC. The CAC was going up.

What I had to learn, slowly and sometimes painfully, was that some of the most important work a marketing leader does is to defend a model of value that does not fit cleanly into a quarterly report. It is to say, in language a CFO can hear, that the cost of a community member who converts in month nine is not the same as the cost of a paid lead who converts in week two — even if the second number is lower.

The leaders who understand this are the ones who build institutions. The leaders who do not understand this are the ones who leave behind a trail of declining CACs and rising refund rates, and a marketing function that looks productive in the dashboard and exhausted in the hallway.

The Question I Ask Now

When I sit down with a marketing team now, whether at a university, a bootcamp, or a B2B enterprise, the first question I ask is not what is your CAC? The first question I ask is: what is the thing you are actually buying, and is it the thing you say you are buying?

If the answer is “we are buying conversions, and conversions are what we want,” then the next question is: conversions into what? A registration? An enrollment? A completion? A graduate who gets hired? A graduate who gets hired in a job that pays more than the cost of the program? A graduate who gets hired in a job that pays more than the cost of the program and stays in the field for five years?

Every additional layer of “into what” is a layer of truth that the spreadsheet is hiding.

The conversation-based model is not a rejection of measurement. It is a refinement of it. We still measure. We measure completion rates, placement rates, salary uplift, retention in field. We measure the things that compound over years, not the things that spike in weeks. We have rebuilt the dashboard. The CAC is still in the corner, but it is no longer the headliner. The headliner is: how many people, in this quarter, made a decision that changed the trajectory of their working life, and what did it cost us to be the reason they made it?

That number, when you finally measure it honestly, is the only one that matters.

I deleted the spreadsheet that day. I have not built another one like it since. The marketing functions I have built since then are messier, harder to report, and an order of magnitude more effective. The communities I have helped build — the career-upgrade community from the SEA expansion era, the higher education alumni networks I have helped establish, the partnership workshops I have facilitated, the open labs I now run — are the reason I sleep at night.

The spreadsheet was not lying. It was just answering a different question than the one we should have been asking.

I stopped chasing MQLs because I started counting the people behind them.