The Spring '26 updates have me rethinking our roadmap. Catalog, UCP, agentic storefronts, the deeper Sidekick stuff—it feels like the ground is shifting under how merchants discover and use apps, and I want to make sure we’re building toward what people actually need rather than what looks good in a changelog.
We build and maintain a handful of apps, so I’m coming at this from the developer side, but I’m hoping to hear from both devs and merchants in this thread. Trying to swap real experience, not pitch anything.
For other developers, I’m curious how you’re actually getting apps in front of people right now:
Which channels are genuinely pulling weight for you? App store SEO, content, communities, partnerships, paid ads, cold outreach, agency referrals? I’d love to know what’s working versus what everyone says works.
Has the shift towards AI discovery changed your marketing at all, or is it still mostly the same playbook?
If you had to name one thing that moved installs or retention more than anything else, what was it?
For merchants- this is the part I care about most, because it shapes what we build:
Of the recent updates, which ones do you actually use day-to-day, and which felt like hype?
Where are you still fighting your stack? The unglamorous, repetitive stuff that no app has fixed well yet.
If you could hand a developer one problem to solve, what would it be?
I’m happy to share what’s worked and flopped on our end too— we’ve spent real money learning some of this the hard way, so I’m glad to trade notes rather than just take them.
Mostly I want to build the right things instead of just guessing.
Great discussion! From what I’ve seen, the fundamentals still matter most. App Store SEO, helpful content, and being active in communities like this continue to drive quality installs. AI is changing how people discover apps, but it seems to be rewarding developers who already have clear documentation, useful content, and a solid product rather than replacing those channels.
From the merchant side, I think the biggest opportunity is still reducing repetitive work. Features that save even 10–15 minutes a day - whether it’s content creation, reporting, or product management- tend to deliver the most value. That’s also why I think Sidekick has a lot of potential if it keeps getting smarter.
By the way, I recently shared a post here on the Community about getting started with AI tools for Shopify merchants. It covers a simple framework for choosing AI tools, building daily habits, and a few practical use cases. Feel free to check it out if you’re exploring how AI fits into your roadmap.
From what I’ve seen, solving a real merchant problem consistently has had a much bigger impact than chasing every new platform update.
On the marketing side, the channels that seem to build the most trust over time are app store SEO, educational content, and genuinely participating in communities. They take longer to show results than paid ads, but the users they bring in are often a better fit and tend to retain longer.
As for AI, I think it’s changing how merchants discover solutions rather than what they need. Clear documentation, real use cases, and helpful content are becoming even more important because AI tools often surface those resources when people are researching solutions.
From conversations with merchants, the biggest pain points still aren’t flashy features—they’re things like unreliable tracking, disconnected data between platforms, repetitive manual workflows, and apps that don’t integrate well with the rest of their stack. Those problems may not be exciting, but solving them well creates a lot of long-term value.
My biggest takeaway is that merchants rarely ask for more features—they ask for simpler, more reliable workflows that save them time and reduce uncertainty. Those kinds of improvements tend to have the biggest impact on both adoption and retention.
On the “what to build” side: with agentic storefronts and the deeper Sidekick stuff, I think the deciding factor is shifting from what the AI can do to whether a merchant trusts leaving it on when they’re not watching. The second an app takes an action that touches money or a customer promise, the real question in their head is “what happens when it’s wrong?” The apps that keep installs through this shift will be the ones that make the AI reviewable and reversible — draft-for-approval on anything that moves money, plus a plain log of what it did and why.
The quiet trust-killer is the failure that doesn’t announce itself — the sync that silently stops, the automation that “succeeded” but actually did nothing. Merchants don’t file those as bugs, they just leave. So if I had to pick one thing to build, it’s whatever surfaces those before the merchant hits them the hard way. And +1 to sophia24 on marketing — the only thing that’s genuinely moved retention for us was talking to people with the exact problem one at a time; the rest was noise until that.
I think one thing that hasn’t changed is that merchants tend to stick with apps that solve a real recurring problem rather than those with the longest feature list.
From what I’ve seen retention usually comes from saving merchants time or reducing manual work. Features like automation clear reporting and reliable workflows often have a bigger impact than headline features that look impressive but don’t get used regularly.
The recent AI features are interesting but I think they’ll be most valuable when they simplify existing workflows rather than adding another layer of complexity.
I agree with your point, AI rewarding good fundamentals rather than replacing them. That’s something I’ve been thinking about as well. It feels like clear fundamentals rather than replacing them. That’s something I’ve been thinking about as well. It feels like clear documentation, practical use cases, and content that actually answers merchant questions are becoming more valuable because both merchants and AI tools can understand them more easily.
Thanks for sharing your perspective (and AI framework as well); I’ll definitely take a look.
I really like this perspective, especially the point that AI is changing how merchants discover solutions rather than what they need. That’s an important distinction.
The part about disconnected data and repetitive workflows also resonates. We often spend time thinking about new features, but a lot of merchants’ conversations still come back to making existing processes more reliable and less manual.
Thanks for taking the time to write such a detailed response.
Great point. I hadn’t thought about trust being more important than capability, but it makes a lot of sense. Reviewable AI actions and clear logs will probably become essential as more automations get introduced. Thanks for sharing your perspective.
Glad we’re on the same page. I think that’s exactly where the biggest opportunities are right now building apps that consistently remove friction and save merchants time rather than simply adding more features.
Thanks again for sharing your perspective. And if you found my earlier reply helpful , I’d really appreciate it if you could mark it as the solution.