If you sell more than a handful of products online, you’ve probably tried bundling — pairing items together at a slight discount to bump up order size. And you’ve probably picked those bundles based on gut feel.
There are artificial intelligence (AI) tools that can do this with data instead of instinct. But here’s the thing: the right tool depends entirely on how many orders you’re actually processing.
The Free Option Most Small Shops Should Start With
If you’re doing fewer than roughly 100 orders a month, you don’t need a fancy recommendation engine. You need a CSV file and a chatbot.
Export your order history from whatever platform you use. Paste it into ChatGPT or Claude (both offer free tiers that can handle this kind of analysis). Ask it to identify the most common product pairings and suggest bundle pricing.
**Before you upload anything:** Strip out customer names, email addresses, shipping addresses, and any other personal information first. You only need product names, quantities, and order IDs for this analysis. Uploading customer data to third-party AI services raises real privacy concerns — and depending on where your customers are, it could create issues under regulations like the General Data Protection Regulation (GDPR) or the California Consumer Privacy Act (CCPA). Keep it simple: anonymize first, analyze second.
That’s it. You’ll get a list of what your customers are actually buying together — not what you *think* they’re buying together.
This isn’t real-time personalization. It’s periodic analysis you’d run monthly or quarterly. But it gets you much of the practical value of a paid tool for essentially nothing.
**Honest caveat:** This only works if you have enough orders to show patterns. A brand-new store with 15 orders has nothing meaningful to analyze. You need at least a few hundred transactions before patterns emerge.
When It’s Worth Paying for a Tool
Once you’re processing several hundred orders a month and carrying a large catalog, manually running spreadsheet analysis stops scaling. A pet supply store with 500+ products can’t realistically identify every meaningful product combination by hand — or even by periodic chatbot analysis.
That’s where tools like **Rebuy** come in. It plugs into Shopify and uses machine learning (ML) to analyze purchase history in real time, surfacing “customers also bought” bundles dynamically at cart and checkout. Pricing starts in the low hundreds per month — check [rebuyengine.com/pricing](https://rebuyengine.com/pricing) for current rates, since their plans and tiers have shifted over time.
**Barilliance** offers similar AI-driven recommendation engines for stores not on Shopify.
Vendors in this space commonly claim noticeable increases in average order value from AI-powered recommendations. But those are self-reported marketing numbers — we couldn’t find independent studies confirming those figures for small businesses specifically. Take them as directional, not gospel.
The “AI or Nah” Verdict
For most small shops: start with the free method. Seriously. Export your data (anonymized), talk to a chatbot, build your bundles based on what it finds. A paid recommendation engine makes sense when you have the order volume to feed it and the catalog complexity to justify it — not before.
**Your one takeaway:** Before you spend anything on bundling software, open your order history and paste it into ChatGPT or Claude with the prompt: *”What are the most common product pairings in this data, and what bundle pricing would you suggest?”* Remember to strip out any customer personal information first. That ten-minute exercise will tell you whether you need a paid tool or whether you’ve been overthinking this.