The certification industrial complex had a good run.
For a decade, the pitch worked: take this course, pass this exam, get this badge, get hired. Bootcamps, LinkedIn Learning certificates, Google certifications, AWS badges — the market was flooded with credentials, and for a while, employers paid attention.
They're paying less attention now.
Not because skills don't matter. Because certificates stopped being signals of skill and became signals of free time. Anyone can pass a multiple choice exam with enough practice. Fewer people can show you something they actually built.
In 2026, the question that's replacing "what certifications do you have?" is simpler and harder: show me what you've made.
What Changed
Three things happened simultaneously that broke the certification model.
AI made credentials easier to fake. You can now pass most online certification exams with AI assistance. Employers know this. The ones who don't yet will figure it out soon. When a credential can be obtained without demonstrating genuine understanding, it stops carrying information.
The talent market got more competitive. Remote work globalised hiring. The candidate pool for any given role expanded dramatically. When everyone has the same certifications, certifications stop differentiating. You need something that can't be templated.
Hiring shifted toward demonstrated output. The companies hiring aggressively in 2025 and 2026 — in AI, in climate tech, in defence, in infrastructure — are moving fast and don't have time for extensive credentialing processes. They look at GitHub repos, portfolios, live products, published writing. Things that exist in the world and can be evaluated directly.
A certification tells a hiring manager what you studied. A side project tells them what you can do when nobody's watching.
What "Proof of Work" Actually Means
Proof of work is anything you made that someone can look at, use, or evaluate without you being in the room.
It doesn't have to be impressive. It has to be real.
- A data analyst who built a dashboard tracking local housing prices using public datasets
- A marketer who grew a newsletter from 0 to 3,000 subscribers writing about their industry
- A developer who built and shipped a small tool that 200 people use
- A security researcher who published a write-up of a CTF challenge they solved
- A designer who redesigned a broken app interface and documented why
- A product manager who wrote a teardown of a product's onboarding flow with specific recommendations
None of these require a company behind them. All of them say more than a certificate.
Why It Works Psychologically
Hiring is fundamentally about risk reduction. A hiring manager approving a candidate is staking their own reputation on that person performing. Anything that reduces uncertainty is valuable.
Certifications reduce uncertainty about whether you sat through a course. Proof of work reduces uncertainty about whether you can actually do the thing.
When a hiring manager can see your GitHub, read your writing, or use your tool — they've already started evaluating you before the interview begins. By the time you walk in, part of the risk is already gone. You're not a stranger with a credential. You're someone they've already seen work.
That changes the entire dynamic of the conversation.
What to Build by Field
You don't need a startup. You need one concrete thing that demonstrates the core skill of the role you want.
Software Engineering
Build something small and ship it. A CLI tool, a browser extension, a simple web app with a real use case. Open source it. Write a README that explains what it does and why you built it. The project doesn't need to be novel — it needs to exist and work.
If you want to go further: contribute to an open source project that employers in your target field actually use. A merged PR in a relevant repo is worth more than most certifications.
Data / Analytics
Pick a publicly available dataset that intersects with an industry you want to work in. Clean it, analyse it, find something interesting, visualise it, publish it. Write 500 words explaining what you found and why it matters.
Kaggle competitions give you a ranking. Independent projects give you a story.
Marketing / Growth
Start something with a real audience — a newsletter, a content series, a community. Document the growth. Be specific about what you tried, what worked, what didn't, and what the numbers were.
A marketer who grew their own newsletter to 5,000 subscribers has more credibility talking about audience growth than one with a digital marketing certificate.
Cybersecurity
CTF competitions (Hack The Box, TryHackMe, PicoCTF) give you flags and rankings — document your write-ups publicly. Set up a home lab and blog about what you're learning. Find and responsibly disclose a bug through a bug bounty program.
One CVE disclosure or a series of detailed write-ups signals more genuine skill than a stack of CompTIA badges.
Design
Redesign something broken. Pick a product you use that has obvious UX problems, redesign the flow, and document your reasoning — the problem you identified, the constraints you worked within, the tradeoffs you made.
The reasoning matters as much as the output. Designers who can articulate why they made decisions are rare.
Product Management
Write product teardowns. Pick products in your target industry and analyse them deeply — what's working, what's broken, what you'd prioritise and why. Publish them. If you can get feedback from people in the industry, even better.
A PM candidate who has publicly documented how they think about product decisions is easy to evaluate. One with a PSPO certificate is not.
The Compounding Effect
Side projects compound in ways certifications don't.
A certificate is static. It represents what you knew on the day you passed the exam, and it doesn't grow.
A project grows. You ship v1, get feedback, improve it. You write about what you built, which attracts readers, which leads to conversations, which leads to opportunities. You put it on your CV, which prompts questions in interviews, which gives you stories to tell instead of credentials to recite.
The newsletter you started becomes a case study. The open source tool you built gets 400 stars and someone at a company you want to work for notices. The blog post you wrote gets referenced in someone else's article.
None of this is guaranteed. All of it is possible. Certifications offer neither.
The Objection: "I Don't Have Time"
Most side projects that matter took less time than people think.
A useful CLI tool: a weekend.
A blog post that demonstrates genuine expertise: 3–4 hours.
A data analysis with visualisations: a Saturday afternoon.
A redesign case study: two evenings.
The barrier isn't time. It's starting with something too ambitious, getting overwhelmed, and stopping.
The rule: scope it down until it's embarrassingly small. Then build that. Ship the embarrassingly small version. Iterate if it gets traction. Move on if it doesn't.
A shipped embarrassingly small project beats an unshipped ambitious one every time.
Certifications Aren't Worthless
Some credentials still carry weight in specific contexts:
- Regulated industries — healthcare, finance, law have certification requirements that aren't going away
- Government and defence contracting — clearances and specific certs are often legally required
- Early career with no experience — a relevant certification combined with a project is better than the project alone when you have nothing else
- Specific technical depth — AWS Solutions Architect, CPA, CISSP — these signal a level of domain knowledge that's hard to fake with a side project alone
The point isn't to never get certified. It's to not mistake the certificate for the skill, and not to use certification as a substitute for building things.
What Hiring Looks Like in 2026
The hiring managers paying attention are already asking for portfolios, GitHub links, writing samples, and live products. Technical interviews increasingly involve real work — take-home projects, pair programming sessions, case studies evaluated on reasoning not just output.
The candidates winning those processes are the ones who have been building in public — sharing what they're learning, shipping things, documenting their thinking. Not because it's a strategy, but because they're genuinely interested and can't help themselves.
That's the thing about proof of work that no career advice can manufacture: the best version of it isn't strategic. It's what happens when you build things because you find them interesting, and then have the presence of mind to document and share them.
Start building something this week. Scope it down. Ship it. Write one paragraph about what you learned.
That paragraph is worth more than the next certificate you were going to buy.