I was brought in as Head of Marketing at a leading Singapore EdTech provider offering SkillsFuture-subsidised courses. The mandate was simple, the scope was massive: grow revenue. Twelve months in, we'd hit $9.8M in annual revenue, acquired 9,947 leads, closed 5,298 sales, earned 22 media placements, and generated $791K in PR value.
This isn't a story about luck or viral campaigns. It's the story of fixing broken attribution, rebuilding the funnel from scratch, and learning which channels actually drive revenue versus which ones just look good in spreadsheets.
When I arrived, the first thing I did was ask the team, "What's our cost per sale?" The answer was uncomfortable silence. They knew their cost per lead. They knew their click-through rates. But nobody was tracking what actually matters: revenue per pound spent.
This is where the real problems were hiding.
Attribution was a mess. Teams were counting course registrations as conversions, not actual payments. Someone would download a guide (that's a lead), then maybe click an email three weeks later, then fill out a form, then think about it for a month, then finally pay. By then, the team had no idea which touchpoint actually drove the sale.
Paid channels were running in silos. Google and Meta teams weren't talking to each other. There was no funnel mapping from awareness to purchase, just ads pointing to a homepage. We were pouring money into campaigns with no idea which ones worked and which ones were burning cash.
"A cheap lead that doesn't convert is just waste with a lower unit cost."
We started from first principles. Revenue is the only metric that matters. Everything else is a proxy.
First, we rebuilt attribution from scratch. We implemented UTM parameters across every single channel, tracked the full funnel from first click to payment, and created a conversion window. We could finally see: this Google Search click eventually became a sale. This Facebook impression led to a website visit, then an email, then a purchase. This offline event led to three direct sales two weeks later.
Second, we separated campaigns by intent. Brand campaigns serve a different purpose than course-specific campaigns. Awareness campaigns shouldn't be judged on conversion rate, awareness campaigns should be judged on brand lift and downstream conversion. We split the budget accordingly.
Third, we built audience tiers. Cold audience: people interested in online learning but not yet aware of us. Warm audience: people who visited the site but didn't convert. Hot audience: people who added a course to cart but didn't finish checkout, people who filled out 80% of a form then abandoned.
This simple framework meant we could stop throwing everyone the same generic message and start talking to people where they were actually at in the buying journey.
Google Search was our highest-intent channel. People searching "SkillsFuture course" or "Singapore data analytics course" were already in buying mode. We bid aggressively on course-specific keywords and less aggressively on broad learning keywords. Cost per sale on Search was our lowest across all channels: roughly SGD $215 by year-end.
Meta drove upper funnel. We stopped trying to convert people on Facebook. We used it for awareness, for reaching people interested in upskilling, for retargeting people who had visited the site. We measured success by cost per website visit and downstream conversions, not by immediate form fills. When we stopped expecting Meta to do the heavy conversion lifting, we actually got better conversions out of it.
Offline was invisible before we measured it. Roadshows, corporate partnership events, and talks at co-working spaces were happening, but nobody was tracking the revenue. Once we added tracking, we found offline drove 30% of our annual sales. The average sales cycle from a roadshow was 17 days. These were high-quality customers, not just high-volume leads.
PR amplified everything else. 22 placements across CNA 938, Lianhe Zaobao, Business Times, and MoneyFM. When someone sees your brand in credible media, then sees your ad, conversion rate goes up. We didn't measure PR by media value alone. We measured PR by how it reduced cost per lead across paid channels. Once we had earned media coverage, our cost per lead on Google dropped 22% because trust was already there.
These aren't hypothetical improvements or projections. These are the actual results across 12 months.
The cost per sale reduction mattered more than the volume. We could have acquired way more leads. We could have spent twice as much on Meta awareness. But 33% efficiency gain while growing revenue 10x meant we weren't just buying growth, we were building a sustainable model.
Track cost per SALE, not cost per lead. A cheap lead that doesn't convert is waste with a lower unit cost. It's still waste. If you can't connect a lead to a revenue outcome, you can't make smart channel decisions.
Google and Meta serve different stages. If you're using them the same way, you're optimizing one and losing money on the other. Google is for intent. Meta is for reach. Use them accordingly.
Offline still converts. Every marketing conversation in 2024 is about TikTok and AI and attribution. But roadshows, talks, and events are still some of your highest-intent channels. Track them. Double down on them.
PR amplifies paid. Paid alone is expensive. Paid plus earned media is sustainable. Your audience is more likely to convert on an ad if they already trust you because they saw you in the news.
Scale isn't about working harder. It's about measuring right and allocating to what works.
If you're running paid ads, you're probably not tracking cost per sale yet. Start there. Connect your ad account to your CRM or payment system. Give it two weeks of clean data. Then look at which channels and campaigns are actually profitable, not which ones look good on dashboards.
If you've already done that, look at your offline channels. Are you tracking webinars, events, partnerships? If not, you're ignoring your best-converting channels.
If you want someone to audit your specific page, campaign structure, or attribution setup, that's exactly what we do in paid strategy reviews. Not generic advice. Your specific numbers.
Get a scored, actionable report on your exact page - not generic tips. Delivered within 48 hours. From SGD $99.
Get a Page Audit →