I optimize paid media toward the business outcome behind the metrics.

I'm Andrew Fonseka. At Aegon I manage a seven-figure annual paid media budget across five platforms. Before that I ran nine agency accounts at once, and I now consult for 10+ clients across three countries.

Performance Marketing Specialist Surrey, BC

Portrait of Andrew Fonseka
  • Seven-figureAnnual paid media budget, across Google Ads, Meta, LinkedIn, DV360 and Amazon DSPAegon · U.S. and Canada
  • +15%CRM-qualified leads in three months, at flat spendAegon, in-house

The stack behind the results.

Grouped by the job each tool does, from buying the media to proving what it returned.

Paid media

Search, social and programmatic

  • Google AdsSearch, Shopping, Performance Max, Display, YouTube
  • Meta Ads
  • LinkedIn Ads
  • DV360
  • Amazon DSP
  • Microsoft Advertising
  • Google Merchant Center
  • TikTok, Snapchat and Reddit Ads

Tracking and measurement

Getting real outcomes back to the platforms

  • GA4
  • Google Tag ManagerIncluding server-side tagging
  • Offline conversion imports
  • Enhanced Conversions
  • Meta Pixel and Conversions API
  • LinkedIn Insight Tag
  • CRM integration

Reporting and automation

Dashboards leadership reads, and fewer manual hours

  • Power BI
  • Looker Studio
  • HubSpot
  • Airtable
  • Excel and Google Sheets
  • Google Ads scripts
  • n8n and Zapier

Web, CRO and SEO

The pages the ads land on

  • A/B testing
  • Landing page optimization
  • Shopify
  • Lovable
  • Microsoft Clarity
  • SemrushOn-page and technical SEO

Three accounts, judged by what the business counted.

In-house at Aegon, independent for Effat University, and agency-side for a gold-loan lender. In each one I found the outcome the business counted, made it measurable, and moved budget toward it.

Putting a seven-figure budget where customers come from.

Company
Aegon, financial services
Role
Digital Marketing Specialist, in-house
Markets
U.S. and Canada
Platforms
Google Ads, Meta, LinkedIn, DV360, Amazon DSP
Timeline
January 2025 – present

The problem

Five platforms, three funnel goals, two countries. Platform dashboards reported leads, but none of them showed which channels produced leads that sales actually qualified.

50% lead generation25% brand25% awareness

The signal

HubSpot knew which leads sales qualified. Google Ads and Meta did not, so their bidding treated every lead the same.

The decision

First, measure the real outcome. I fed HubSpot lead outcomes back to Google Ads and Meta through offline conversion imports and Enhanced Conversions, backed by server-side tagging and the Meta Conversions API.

Then, move the money. With qualified-lead data per channel in the Power BI and Looker Studio dashboards I own and present to leadership, I reallocated and re-paced budget toward what worked, changed bid strategies and tested creative.

Alongside that, I restructured campaigns and used HubSpot lead-stage data to improve landing pages, building page variants in Lovable and briefing the web team on larger rebuilds.

The outcome

  • +15%CRM-qualified leads in three months, at flat spend
  • −13%Cost per qualified lead in the same three months
  • −12%Blended cross-channel cost per lead, over six months
  • −11%Google Ads cost per lead across Search, Display, Performance Max and YouTube
  • 2.5% → 2.8%Paid media conversion rate after campaign and landing page work
  • −8%Programmatic CPM on DV360 and Amazon DSP
Reporting and automation

Co-built an Airtable budget-pacing dashboard fed by platform APIs, now used across the marketing team, and n8n and Zapier workflows routing Google Ads data into Power BI. Built Google Ads scripts that save about 12 working days a year of manual campaign setup.

The learning

The platforms can only optimize toward what they are told. Feeding them qualified-lead outcomes came first; every budget, bid and creative decision after that was judged against it.

Results shown as percentages only. Absolute budgets, lead counts and platform data are confidential under NDA, and all advertising met financial services marketing regulations.

Twice the sign-ups on a quarter less budget.

Client
Effat University, Saudi Arabia
Engagement
Independent, via agency partner
Channels
Meta Ads + HubSpot CRM
Timeline
March – June 2026

The problem

Effat University was paying SAR 80 for every sign-up through a basic Meta setup run by its previous agency: no real targeting, and nothing telling Meta which leads mattered.

The signal

Past academic and enrolment data showed which programs and student segments had high demand and high enrolment. So I rebuilt the account around one question: which students actually sign up and enrol?

The decision

  1. Target who enrolsI used past academic and enrolment data to find the programs and student segments with high demand and high enrolment, then A/B tested them against each other.
  2. Qualify at the formI rebuilt the lead form so it screened for real intent before a lead ever reached the admissions team.
  3. Teach Meta what a good lead isMeta Pixel on the site, and HubSpot lead outcomes synced back into Ads Manager, so delivery optimized toward sign-ups instead of form fills.
  4. Move budget with the calendarWhen the intake closed I paused retargeting and moved its budget to fresh prospects for the next intakes.

The outcome

  • 500 → 1,050Student sign-ups, up 2.1× on the previous agency's campaign
  • SAR 80 → 29Cost per sign-up, down 64%
  • 63% → 81%Qualified leads that went on to sign up
  • −25%Budget versus the previous agency's campaign

The learning

Cheaper leads were the side effect. The goal was sign-ups.

Backing the channel with the better leads.

Client
Gold-loan lender, Sri Lanka
Engagement
Kites Global (agency)
Channels
Meta and Google Ads
Ad spend
US$25K
Timeline
2023

The problem

A lender offering gold loans, which are loans secured against gold jewellery, competing with Sri Lanka's national banks for the same borrowers on a US$25K ad budget.

The signal

Meta was bringing in higher-quality leads than Google, at a lower cost per lead.

The decision

Lead quality and cost set the split: 60% of the budget to Meta and 40% to Google Ads, the only two channels we ran. The client reported funded loans back as a weekly total, and that total, not platform lead counts, was the number I reported against.

60% Meta40% Google Ads

The outcome

  • 259Funded gold loans from US$25K in ad spend
  • ~US$97Ad cost per funded loan, not per lead
  • US$400K+Loan value originated (client-reported)
  • 16×Loan value originated per ad dollar

The learning

A cheap lead only counts if it turns into a loan. Lead quality set the split; the weekly funded-loan total was the scoreboard.

More from my independent work.

  • Metro Vancouver · B2B home services · Paid search−14%Paid search cost per lead, through search term restructuring, expanded negative keyword lists, landing page rebuilds and budget shifts to high-intent campaigns.
  • DIMO Academy · Technical education · Google Ads+11%Lead conversion rate, by rebuilding Google Ads audiences and campaign structure, feeding CRM lead outcomes into optimization and redesigning landing pages.

I learned paid media where every dollar had to prove itself.

I started in 2022 at Enfection, a Colombo agency, building remarketing audiences and rebuilding the pages where visitors dropped off. At Kites Global I ran nine client accounts at once across financial services, retail, beauty and education. Weekly client reporting taught me early that results have to show up in the client's numbers, not just the ad platform's.

An MBA brought me to Vancouver. Since January 2025 I have managed a seven-figure annual paid media budget at Aegon across Google Ads, Meta, LinkedIn, DV360 and Amazon DSP for the U.S. and Canada business units. Since January 2026 I have also taken on 10+ independent engagements in Canada, Saudi Arabia and Sri Lanka.

Three countries, one lesson: a lead count is where the work starts. Growth comes from optimizing to what the business counts.

Work and education

  1. Sep 2019 – Jun 2022BSc (Hons)Management and Digital Innovation · University of London
  2. Jul – Dec 2022EnfectionDigital Marketing Intern · Colombo
  3. Jan – Aug 2023Kites GlobalDigital Marketing Strategist · Colombo
  4. Aug 2023 – Sep 2024MBAUniversity Canada West · GPA 4.00 / 4.33
  5. Jan 2025 – nowAegonDigital Marketing Specialist · Langley, BC
  6. Jan 2026 – nowIndependent consultantCanada · Saudi Arabia · Sri Lanka