Has AI Killed Software Moats? The Team Behind Bartender Doesn’t Think So

From startup podcasts to X threads, one idea has become almost conventional wisdom in the AI era:

Software is no longer the moat. Distribution is.

AI has made it easier than ever to build software. As more products compete for attention, many founders argue that audience, trust, and distribution now matter more than code itself.

But not everyone agrees.

When we spoke with Stephen Shan, Senior Product Manager at Bartender, he largely agreed that distribution is becoming more important. He just doesn’t think great software has become irrelevant.

Why this debate exists in the first place

The idea that distribution is becoming the primary moat didn’t appear out of nowhere.

As AI-assisted development becomes mainstream, developers can launch products faster than ever. Across the software industry, founders increasingly argue that building software is becoming commoditized while attention, audience, trust, and distribution channels are becoming harder to replicate.

We heard a similar theme in our recent conversations with indie Mac developers.

Several developers told us that building the app is often no longer the hardest part.

Getting people to discover it is.

Stephen agrees with much of that assessment.

 “AI has definitely made it easier than ever to build software so the ability to create a basic app is becoming more commoditized. That means distribution, growth, brand, and trust matter more than ever.”

That’s a perspective many independent developers are now experiencing firsthand.

The barrier to creating software is falling.

The barrier to earning attention is rising.

But software still isn’t a commodity

Where Stephen diverges from the more extreme version of the argument is in what happens after launch.

A growing number of founders suggest that software quality matters less in an AI world because competitors can quickly replicate features.

Stephen doesn’t buy that.

“I do not think software itself has stopped being a moat entirely.”

Building version one may be easier.

Building version fifty is not.

“The first version of an app may be easier to create now, but maintaining it, supporting real users, making good product decisions, and continuing to innovate are still very human tasks.”

That’s an important distinction.

AI can generate code.

It cannot automatically decide which features should exist, which user requests should be prioritized, or how a product should evolve over years of real-world usage.

Bartender app is a useful example

Bartender is exactly the type of software that challenges the “software no longer matters” narrative.

On the surface, the idea sounds simple: organize and manage menu bar icons.

But long-time Mac users know the challenge isn’t building the first version.

It’s keeping the software reliable through years of macOS updates, API changes, UI redesigns, compatibility issues, and evolving user expectations.

As Stephen explained:

“In our case, continuing to evolve Bartender and create new product lines like Bartender Pro requires taste, judgment, and a strong understanding of what users actually need.”

Those are qualities that don’t easily translate into a prompt.

If you’re unfamiliar with Bartender, we’ve also published an in-depth review covering how the app helps organize and manage the macOS menu bar.

The invisible work behind great software

One theme that consistently appears when talking to indie developers is how much work users never see.

People see the app.

Developers see everything surrounding the app.

Stephen described the reality this way:

“Users only see the app, but we are responsible for the entire business around the app too.”

That includes:

• Payments
• Customer support
• Documentation
• Refunds
• Marketing
• Onboarding
• Compatibility testing
• New macOS releases

And for utility software, much of that effort remains invisible when everything works correctly.

“Users expect it to ‘just work,’ but macOS changes constantly, and maintaining that level of reliability takes a lot of ongoing work.”

Ironically, that’s exactly the type of work many discussions about AI overlook.

Shipping an MVP is easier than ever.

Running a software business is not.

The real moat may be the combination

After speaking with multiple Mac developers over the past few months, one pattern keeps emerging.

The debate between software and distribution is probably the wrong debate.

The best products need both.

Stephen summarized it perfectly:

“The real advantage is the combination: a product people genuinely love, plus the ability to reach and retain the right users. “

Distribution helps people discover a product.

Great software gives them a reason to stay.

One without the other rarely creates a sustainable business.

Why this matters beyond Bartender

Interestingly, Stephen isn’t alone in this view.

In our earlier conversations with other Mac developers, distribution emerged as a recurring theme. Yet none of them argued that software quality had become irrelevant. Across multiple indie Mac developers, there appears to be growing agreement that while AI makes building software easier, creating products users trust and continue using remains a very different challenge.

The timing of this discussion is particularly interesting.

Apple recently updated its App Store guidelines with stricter language around low-value and low-effort apps, signaling a growing concern about software quality as app creation becomes easier.

Whether those changes ultimately reshape the App Store remains to be seen.

But the broader industry trend is clear.

Building software is becoming easier.

Standing out is becoming harder.

And according to the developers we’ve interviewed, success increasingly depends on both sides of the equation: creating software people genuinely love and building a way for people to discover it.

The takeaway

Distribution may be becoming a bigger moat. But great software still matters.

In fact, the easier it becomes to build software, the more important qualities like product judgment, long-term support, trust, and customer understanding may become.

AI can help someone build an app.

It still takes people to build a product worth recommending, supporting, and using for years.

That may prove to be the moat that’s hardest to automate.

Read also:

• Find the Bartender on Setapp
• Inside Bartender 6: Trust, Control, and the Future of the macOS Menu Bar
• Vibe Coding: Can AI Help Mac Users Escape Bloated Software?
• Top 30 Mac Apps for Students – Productive Study Guide

Loved the article, share!