The State of the Side Hustle in 2028: What's Working, What's Dead, and What's Next

Every six months or so I sit down and force myself to be honest about the side hustle landscape. Not the Instagram version of it — not the "I made $47k in my sleep" screenshots — but the actual ground-level picture of what's working and what isn't. We're halfway through 2028, and the landscape has shifted enough that I think a real accounting is overdue.

Some things I thought would still be viable are clearly dead. Others I almost wrote off a year ago are quietly thriving. And a few emerging models are starting to look genuinely exciting in ways that weren't possible even two years ago.

Here's where I've landed.

What's Dead (Or At Least on Life Support)

Fiverr arbitrage — the model where you buy cheap freelancers offshore and resell their work at a markup through your own Fiverr profile — is effectively done. The platform has tightened its quality detection, buyers are more sophisticated, and AI tools have collapsed the price floor so hard that the margins were already gone before the policy changes. If you're still doing this, I'm genuinely curious how.

Print-on-demand for generic niches is saturated beyond recovery. Selling dog-breed hoodies or motivational quote mugs was never a great business, but it was at least viable in 2021. By 2028, the top spots in every obvious niche are locked up by sellers with thousands of reviews, and the shipping economics have gotten worse. There are still micro-niche POD plays that work — genuinely obscure hobbies, hyper-regional humor — but the generic approach is a treadmill to nowhere.

Certain dropshipping models — specifically, the "find a trending product on Alibaba, run Facebook ads, ship from China" playbook — are functionally dead for new entrants. The ad costs are brutal, the shipping times have become a dealbreaker for buyers who've been trained by Amazon, and the product selection is so easily copied that you can't hold a winning product for more than a few weeks before you're buried in competitors.

These aren't controversial takes. Most honest practitioners have been saying this for a while. I'm including them here because I still see people asking about these models on forums, and I don't want anyone to waste months finding out the hard way.

What's Actually Working Right Now

Digital products are the standout story of mid-2028, and I don't think that's a coincidence. When I started selling through MadeThis a couple years back, digital products felt like a niche play. Now they feel like the main event. The economics are genuinely different from physical-product businesses: no inventory, no shipping, near-zero marginal cost, and the ability to iterate the product without any physical retooling.

What I've watched work specifically: templates (Notion, Figma, spreadsheet-based tools), mini-courses in the 60–120 minute range that solve a single specific problem, curated resource packs for specific professional niches, and prompt libraries for people using AI tools in their workflows. None of these are glamorous. All of them compound over time in a way that a service-based hustle simply cannot.

AI-assisted content businesses are real, but the amateurs are getting filtered out faster than I expected. The people winning here aren't the ones who just run prompts through ChatGPT and post the output. They're the ones who use AI to extend their own genuine expertise — people who know something valuable and use AI to produce it at 3x their previous rate. That's a meaningful moat. Raw AI slop without a human perspective isn't holding up in search or on social.

Micro-SaaS is harder than the hype suggests, but the winners are winning quietly and consistently. The playbook that's working isn't "build the next Notion" — it's "build a very specific tool for a very specific workflow that a specific type of professional does 10 times a week." Monthly recurring revenue from 200 paying users at $29/month is a real, quiet business. It takes more technical lift than a digital product, but the recurring nature is hard to argue with.

Newsletter businesses with paid tiers are having a moment. This was supposed to be over by now — every pundit predicted newsletter fatigue. Instead, the market has bifurcated: free, ad-supported newsletters are getting harder to monetize, while paid newsletters in high-signal niches are doing well. The model works when the writer has a genuine, specific perspective that saves the reader time or money they care about.

What's Emerging

Three things I'm watching closely that aren't fully baked yet:

AI agent storefronts — people packaging and selling pre-built AI agents for specific business workflows. It's early and the tooling is still rough, but the demand signal is real. If someone can buy an agent that handles a specific repetitive task for $49 once instead of paying a freelancer $200/month, they will.

"Unbundled" professional services — taking what used to be a full consulting engagement and breaking it into productized, self-serve components. An accountant's tax prep checklist, a lawyer's contract template library, a designer's brand audit framework. This is digital products, but it's being pioneered by people with professional credentials who are finally figuring out how to package their knowledge.

Local data products — hyper-local market research reports, local business directories with real value-adds, regional trend analyses. National-scale content is getting commoditized fast, but genuinely local knowledge still has real information asymmetry. This one feels underexplored.

The Pattern Underneath All of This

Here's what I keep coming back to: the side hustles that are dying all have one thing in common — they rely on arbitrage that AI or scale has eroded. The ones that are working all have one thing in common — they leverage genuine knowledge or creative judgment that isn't easily copied.

That's the filter I'd apply to any new side hustle idea in the second half of 2028. Not "is this a good idea?" but "does this rely on something I know or can create that isn't just execution of a commodity playbook?"

If you want to go deeper on the digital product angle specifically — what it actually takes to set one up and how to pick a platform — the MadeThis reviews page is a useful starting point. And if you're comparing platform costs before committing, I'd check MadeThis pricing alongside the alternatives.

The landscape is genuinely hard right now. But hard for generic players is not the same as hard for specific, knowledgeable ones. If you have something real to offer, 2028 is still a good time to be building.

If you're ready to stop trading hours for dollars and start putting something out there that works while you're not working, MadeThis is where I'd start. It's the platform I use, and it removes every friction point that used to slow me down in the early days.