Naming the squeeze honestly
The pressure you feel is structural, not personal, and it comes from two directions at once. From below: AI now produces serviceable copy, visuals, campaign variants, and reports in seconds - the exact production tasks that junior and mid marketing roles were staffed for - and every marketing team in Malaysia is quietly recalculating headcount against that fact. From above: marketing budgets are the first cut in every downturn, a seniority-independent exposure the profession has always carried. The result is the fastest-shrinking middle of any white-collar lane except perhaps administration. The role-level assessment lives in will AI replace marketers. But run the audit on what AI did not automate and the exits appear: strategy, positioning, user understanding, judgement about what will actually move humans - the thinking layer of marketing - got more valuable, exactly the pattern we document across every profession in AI is the new Excel. The three routes below are three ways of carrying that surviving layer somewhere it compounds.
Route one: the marketer-who-builds (start this weekend)
The lowest-commitment route is also the most immediately underrated: stay in marketing, add building skill, and become the hybrid every team promotes first. The marketer who ships their own landing pages instead of queueing for developer time, automates the weekly reporting deck, builds the UTM-tracking tool, or wires the lead-capture flow end-to-end is operating at a different level from their peers - visibly, measurably, and with receipts at appraisal time. AI collapsed the cost of this layer from years to weeks: the free tutorials plus a few weekends of curiosity covers the first landing page, and the free trial’s six projects formalise it. This route also functions as reconnaissance for the other two: you will discover, cheaply, whether building energises you enough to go further.
Route two: product - marketing’s natural promotion
Product management is, uncharitably but accurately, marketing’s thinking layer given authority over what gets built: understanding users, defining positioning, prioritising against data, and shipping things that move numbers. Marketers convert into product more naturally than almost anyone - the common Malaysian path runs through growth or product-marketing roles at tech companies (RM 4,000–7,000, faster ladders than agency-side) and sideways into PM within a year or two. Two accelerants: genuine technical literacy (a PM who can prototype and read the codebase conversation is a different species - which is why even this route includes the building layer), and portfolio proof - a side project where you identified a user problem, shipped something, and grew it beats every certificate in the PM interview. The full role map, including how product compares to the other seats, is in which tech career fits me.
Route three: the full switch - and why marketer-developers punch above their weight
The complete route: the standard 400–600 focused hours part-time over 6–9 months into the RM 6,000–9,000/month AI-capable junior band (the data) - which matches or beats most marketing-manager salaries at entry. Here is the part the generic switching guides miss: marketing instincts are a career-long edge in development. Developers who understand users build better products; developers who understand distribution build products people actually find; developers who can write and present become the public face of teams. And for the one-person-product path - building your own small software business - the marketer-who-codes is arguably the single best-equipped profile alive: most solo products die of no-distribution, the exact disease you are vaccinated against (the playbook is in how to build a SaaS with AI). Whichever route you choose, the first step is identical and free: a week of real building to locate your energy. Marketers know better than anyone that the landing page matters less than the conversion - this is yours.
Build the portfolio a marketer should build. Whichever route you take, your projects should showcase the thing that makes you different rather than imitating a computer-science portfolio. Strong shapes: a small tool that solves a marketing problem you personally have (a UTM builder, a campaign-reporting dashboard, a competitor-monitoring script), a landing page you built AND then optimised with real traffic data - the second half is what no bootcamp graduate can show - or a small product where you did the whole loop: found the problem, built it, marketed it, and can talk about conversion numbers. That last one is close to unbeatable in junior interviews, because it demonstrates the full builder profile most candidates are missing half of. Attach the numbers; you are the one switcher in the pile who knows how to present results.
