I’ll be upfront: when someone first suggested I look into AI tools for the shop, I rolled my eyes so hard I nearly pulled a muscle. I sell vinyl records. I’m an analogue person running an analogue business. What would I possibly need AI for?

Turns out, quite a lot. Not the flashy, headline-grabbing AI stuff. The boring, operational stuff that actually makes a difference to whether a small business survives.

Inventory Prediction

The most immediately useful AI application I’ve encountered is demand prediction for ordering. Every week, I need to decide how many copies of each new release to order. Order too few and I miss sales. Order too many and I’m stuck with dead stock that ties up cash I can’t afford to lose.

Traditionally, this was pure gut instinct informed by experience. And honestly, after twenty years, my gut is pretty good. But it’s not perfect, and the mistakes are expensive.

Some distributors are now offering data-driven ordering tools that analyse pre-order velocity, social media mentions, streaming numbers, and historical sales patterns for comparable releases. The recommendations aren’t perfect, but they’ve helped me avoid several costly over-orders and catch a few releases I would have under-ordered.

Working with a firm offering AI strategy support helped me set up a system that pulls data from my POS, Discogs market activity, and social media engagement to generate ordering suggestions. It’s not magic — I still make the final calls — but having data-informed starting points rather than pure guesswork has improved my hit rate noticeably.

Customer Recommendations

The irony of running a record shop is that you’re essentially a human recommendation engine. People come in, tell you what they like, and you suggest something they’ll love. I’ve been doing this for two decades, and I’m good at it.

But I can’t be in the shop every hour it’s open, and my staff, while knowledgeable, have different strengths and blind spots. An AI-assisted recommendation system that can suggest related artists and records based on a customer’s purchase history isn’t replacing the human conversation — it’s supplementing it.

We’ve started using a simple system where the POS records customer purchases and suggests related titles for staff to recommend during checkout. “You bought Surprise Chef last time — we just got in the new Putbacks, which has a similar vibe.” It’s not revolutionary, but it’s helpful.

Pricing and Market Analysis

The secondhand vinyl market moves constantly. Prices on Discogs fluctuate based on supply, demand, and trends. Keeping my secondhand pricing current used to involve manually checking comparable sales every few weeks — a time-consuming process that I often skipped when the shop was busy.

Automated pricing tools that monitor Discogs market data and flag records that are significantly under or over-priced relative to current market conditions save genuine time. I don’t let any system set prices automatically — that’s my job — but having a weekly report that says “these ten records are priced 30% below recent sold prices” is useful.

Social Media Insights

Understanding which social media posts actually drive store visits and sales has been one of the more valuable applications of data analysis in the shop. Before, I’d post content and hope for the best. Now I can see that a Reel showing new arrivals on a Thursday afternoon consistently outperforms a static photo posted on a Monday morning.

These aren’t groundbreaking insights individually, but collectively they help me use my limited marketing time more effectively. For a one-person operation where every hour matters, that efficiency is valuable.

Getting help from specialists in this space to set up these analytics dashboards was money well spent. The setup was straightforward, and the ongoing insights require minimal technical knowledge to interpret.

What AI Can’t Do

Let me be clear about the limits. AI can’t replace the human elements that make a record store what it is. It can’t read the room when a customer walks in and needs something specific. It can’t tell a story about why a record matters. It can’t create the atmosphere that makes people want to spend an afternoon browsing.

It can’t curate with taste. Recommendation algorithms are based on patterns, but the best recommendations come from understanding not just what someone has bought, but who they are and what they’re feeling today. That’s a human skill, and it’s the most important thing a record shop offers.

The Practical Takeaway

If you run a small music retail business and you’re sceptical about AI, I understand. I was too. But the operational tools available now aren’t trying to replace your expertise — they’re trying to free up your time so you can spend more of it doing the things only you can do.

Use AI for the boring stuff: inventory analysis, pricing research, social media scheduling, basic customer analytics. Save your human skills for the good stuff: recommending records, building relationships, curating events, and running a shop that people genuinely love visiting.

That’s the balance I’ve found, and it’s working.