How Much Should You Actually Spend on AI? A Revenue-Stage Framework For Shopify Brands

Steve Hutt:
Hey everyone, welcome back to eCommerce Fastlane. I’m Steve Hutt.

Today it’s just me, no guest this week, which honestly I’m excited about, because there’s something that’s been on my mind and I really want to walk and talk through it directly with you.

If you’ve been paying attention to what’s happening in AI right now – and I know many of you have – you’ve probably felt this too. It feels like the pace of change is almost hard to articulate. Every week there’s some new tool, new capability, a new game changer that everybody’s shouting about on LinkedIn or Twitter. And for a lot of founders and marketers, it’s creating a weird tension.

You know a lot about AI. You know you need to be doing something with it. You can see other brands who appear to be pulling ahead using these systems. But the honest question, and the one no one seems to answer clearly, is: what should I actually be doing and spending on AI right now, given where my business is today?

Everybody has different maturity, complexity, GMV size. That’s what I want to dig into today. No hype, no shiny-object syndrome. I’m not going to talk about specific solutions. I want to give you real feedback and a framework for thinking about your AI investment based on your exact revenue stage.

After working with a lot of different Shopify brands over the last few months, what I’ve seen is that the brands winning with AI are not the ones spending the most. They’re the ones being intentional about what they invest in and when.

Before we dive in, I want to say thank you. I genuinely appreciate that you’re here today. Thank you for pressing play and giving me your time. I don’t take that lightly.

If you enjoy this podcast, I’d love you to also check out my newsletter. I launched it last January. It’s all about building your owned audience, and it’s grown to over 40,000 founders and marketers now. I don’t want to pitch it too much, but a lot of my best writing and thinking goes into that newsletter each week. It goes out every Thursday and I truly believe it gives you a competitive edge in the Shopify ecosystem. I work really hard on it, and I’m sure you’ll dig it. You can find it at ecommercefastlane.com.

Let’s talk about the notes I’ve jotted down in my phone before recording today. I’ve been thinking about the different types of customers listening to this show: merchants, founders, and marketers.

My main observation is this: some enterprise brands – maybe doing 5, 10, 50, even 100 million plus – are not just using AI. They’ve built AI into their operating system. It’s baked into customer service workflows, inventory planning, ad creative, merchandising decisions. And I believe the gap is really widening.

It’s not because these brands are spending a ton on AI. It’s because they started with a clear framework:

• What problem am I solving?
• What data do I need to solve it?
• What’s the simplest AI solution to get me there?

Those are the three questions you need to think about.

Meanwhile, if you contrast that with brands doing, say, 500K to 2 million a year, it gets interesting. These founders have spent years in their business. They know it deeply. But that depth sometimes freezes them when they look at the number of AI options out there.

I’m seeing a lot of mistakes right now when brands think about a social media solution or another problem they want to solve and try to solve everything at once. That’s where what I call the “premature complexity” trap shows up. And it’s happening a lot with AI.

If you’ve listened to this show for any length of time, you know I talk about this a lot. I believe it’s the number one pattern I saw when I was a Shopify Merchant Success Manager. So many brands would keep stacking apps, channels, tactics. Every time I got on a call with them, every couple of weeks, I’d end up doing an app audit: “Why do you have all these things? Why are you trying this?”

Often they were just trying to see what sticks. I was concerned about that then, and I’m seeing the exact same pattern play out now with AI.

Founders are literally signing up for seven different AI tools a month: an AI copywriting tool, an AI analytics platform, an AI chatbot, an ad creative generator, and so on. None of them are properly configured. They’re not really integrating with each other. No one on the team actually knows how to use them all really well.

More AI tools do not automatically give you an AI advantage. They don’t create a competitive moat. One well-implemented AI system that’s trained on your data and integrated into your workflow will outperform ten loosely used tools. Most teams are using maybe 20–30% of a tool’s power. Multiply that by 7–10 tools, and you’re not getting great value.

That’s the core reason I wanted to chat today. I want to set the stage and then talk about: based on where you’re at, what is an appropriate AI investment? Is there a framework you can follow?

I’m going to give you some honest numbers around what brands are doing at each stage and what they should realistically invest in AI – not what some vendor wants you to buy, but what actually makes sense for your current stage.

Steve Hutt:
If you’re under 500K in annual revenue, every dollar matters. You have some product-market fit, but you’re working hard to get to solid profitability.

The good news: you likely already have a lot of AI built into your Shopify store. Many of you are using it, but a lot of you are not.

Shopify Magic, for example, gives you AI-powered product descriptions, email subject lines, and basic content generation at no extra cost. It’s native in the admin. You should use that.

Beyond that, there are free tools like ChatGPT, Claude, Perplexity. They can handle a surprising amount of content needs: product descriptions, email drafts, social media captions, and even some basic customer research.

At this stage, my priority for founders is simple: don’t buy anything yet. You don’t need to. If you’re under 500K, lean on what’s already built into Shopify.

Get comfortable with prompting. Learn to use the tools you already have. That skill alone will save you thousands of dollars before you invest in more robust point solutions.

Steve Hutt:
The next tier is brands doing roughly 500K to up to about 1–2 million a year.

A realistic AI spend here, from my perspective, is in the range of 200–800 dollars a month. But this is where you need to be really intentional, because this is the stage where premature complexity starts to kill brands.

I would pick one, maybe two areas where you believe AI can make a measurable impact.

For most brands at this stage, two high-leverage places are:

First, email and marketing automation. Tools like Klaviyo already have AI features built in: send-time optimization, predictive analytics, subject line testing. If you’re already paying for Klaviyo, you may not need another AI tool. You might just need to turn on the features you’re already paying for.

Second, a basic AI chatbot for customer service. Not a 500-dollar-a-month enterprise solution. Something simple that answers: where’s my order, what’s my tracking number, shipping info, and FAQs. Have the bot handle those routine queries directly.

Things I’d avoid experimenting with at this stage include dynamic pricing, advanced personalization engines, and custom AI models trained on your data. You typically don’t have enough data yet for these to work well. Vendors don’t always highlight how much data you need for real impact.

At 500K–2M annual revenue, pick one or two use cases, understand what data you’re feeding in, and measure the ROI coming out.

Steve Hutt:
If you’re in the 2 to 10 million range, this is where things get more interesting. You have enough traffic, order data, and customer data for AI tools to do things humans simply can’t do at the same speed or accuracy.

Here, dedicated tools for specific problems start to make sense. For example, AI-powered site search and product recommendations. Search that understands intent, not just keywords, can meaningfully improve conversion rate. Tools like Klevu or Searchspring go beyond core Shopify search.

Inventory forecasting is another big one. If you’re managing more than a few hundred SKUs, an AI forecasting tool can help prevent stockouts and overstocks, which eat into margin. These systems can pay for themselves quickly.

The key takeaway: not all brands need to be fluent in every solution. You don’t need this to be a full-time role for someone. But you do need education for the team members who actually use these tools, space for them to think about and optimize them, and a regular cadence to review ROI.

Ask monthly: what are we getting from this tool, is there real ROI, or are we just paying subscriptions? At this revenue band, you want to be intentional about the few tools you use, and you want to review their performance consistently.

Steve Hutt:
In the 10 to 50 million range, ad spend is usually meaningful, and AI starts to shift from “a few tools” to more of an integrated infrastructure.

You might be spending anywhere from 3,000 to 15,000 a month on AI across your stack. For bigger brands, even more.

At this level, examples of where AI fits include an AI-powered customer service stack – not just a chatbot, but a system that handles most routine inquiries, intelligently escalates complex issues to human agents, and ties into your help desk and knowledge base.

Dynamic pricing tools also fit well here. At this scale, you have enough transaction volume for AI to actually learn patterns and optimize pricing.

Predictive analytics becomes another major piece. When you connect marketing, inventory, and customer data, AI-driven predictive models become a real competitive edge.

The big shift at this stage is brands training models on their own proprietary data: customer service transcripts, product return patterns, seasonal trends. This data is unique to you. When you feed it into an AI system, it creates an advantage competitors cannot easily copy.

Steve Hutt:
For brands north of 50 million – you’re definitely on Shopify Plus – I won’t go too deep because you likely have a dedicated team thinking about this already.

At true enterprise scale, AI becomes infrastructure, not just a tool: custom model development, a dedicated AI team, autonomous systems managing pricing and inventory allocation, and campaign optimization with human oversight rather than fully human execution.

The important thing is that these enterprise brands didn’t start here. Many started just a year or so ago at tier 1 or tier 2 and intentionally built up their capabilities over time.

Steve Hutt:
Before we wrap, I want to leave you with one filter I use for any AI investment – including inside my own media company. This applies whether you’re at 100K a year or 100 million.

The question is: will this AI investment compound over time, or is it just a one-time productivity boost?

The best AI investments get smarter as you feed them more data. Email personalization learns from every send – open rates, click-through rates, conversions, language that resonates. Search improves with every query. Inventory forecasting becomes more accurate every season.

These investments compound and get more valuable the longer you use them, because they’re powered by your proprietary data.

Contrast that with a tool that just saves you 30 minutes writing product descriptions. Useful, yes. Shopify Magic does that and I think it’s great. But that’s a linear improvement. It doesn’t get dramatically better over time. It’s worth using, but it’s not where your strategic AI budget should go.

You want to prioritize tools that compound.

That’s it for today. I love doing these solo episodes. They’re fun because I get to sketch out what’s on my mind and talk directly to you.

I hope this gives you a clearer picture of realistic AI investments at each stage. The last thing I want is for you to feel like you’re falling behind because you’re not spending 10,000 a month on AI tools when you’re doing 500K in revenue. That’s not the game.

The game is being intentional. Start where you are and build systems that get smarter over time.

If you found this valuable, I’d really appreciate it if you shared it with another founder or marketer who’s trying to figure out their AI strategy. It’s a tough market out there. I understand the anxiety and analysis paralysis. That’s a real thing.

That’s why I’m here. There’s so much content being produced now on eCommerce Fastlane: the blog, this podcast, solo episodes, and interviews with founders, marketing apps, and agency partners talking about how they’re implementing AI.

You’re going to see a slight pivot in the show because AI is the hottest topic people are asking about. They want to create their competitive moat, and I think we’re onto something here.

Thanks for being here today. I appreciate you, and I’ll see you on the next episode of eCommerce Fastlane.