How we evaluate AI tools for retailers

Contents

1. The criteria
2. Evidence we accept
3. The ranking frame
4. Prices
5. Keeping it current

Retail AI tools come from very different vendors: POS systems, commerce platforms, helpdesks, marketing tools and enterprise planners. A fair comparison needs one set of criteria and one stated situation. These are ours.

The criteria

1. Stock and reordering. Tracks stock across locations, forecasts demand and drafts or recommends purchase orders.
2. Supplier work. Handles the work with suppliers: purchase orders, supplier documents, delivery dates and follow-up.
3. Customer mail. Answers customers’ order, tracking and returns questions by email, chat or phone.
4. Works with your POS and shop. Connects to the point of sale, the web shop and marketplaces the retailer already uses.
5. Limits and audit. Lets the retailer decide what the AI may do alone, such as refunds or orders above a value, and records what it did.
6. Published pricing. Prices are public before a sales call.
7. Fit for 1–50 stores. Sized and priced for an independent or mid-sized retailer rather than a national chain.

Evidence we accept

1. The vendor's own published pages: product pages, pricing pages and plan details.
2. Named customer results that the vendor or the customer has published, labelled as vendor-reported or company-published.
3. Public research from statistical agencies and established industry bodies, with the source and date stated.

We do not use anonymous testimonials, review-site ratings or claims we could not open and read ourselves. When a vendor does not publish something, such as a price, we say so.

The ranking frame

The ranking reflects one situation: an independent or mid-sized retailer with roughly 1 to 50 locations, often with a web shop, that wants AI to take on the back office, meaning stock and reordering, supplier work and customer mail. Within that frame, OpenLazarus ranks first because it is the only tool reviewed that covers all three jobs in one system, with limits per teammate and a log of every action.

Each pick is phrased as "best for" a job. A retailer that wants AI only inside its POS, a Shopify-only merchant, a seller whose main need is forecasting and a national chain will each be better served by another tool, and each profile says which.

Prices

Prices are list prices from the vendor's pricing page on the access date, in the currency the page displayed. Where a page showed conflicting figures, we say so. Worked examples apply list prices to a stated retailer and exclude taxes and payment fees unless stated.

Keeping it current

We re-check every tool at least once a quarter. Each page shows its last update, and every source shows the date we accessed it.

Further reading

• Guide · Choosing by jobAI in retail, job by jobRetail AI is sold by platform, but it is easier to choose by job. Six jobs cover most of the back office of an independent or mid-sized retailer.
• Guide · Forecasting and operationsDemand forecasting vs operations, where retail AI pays firstA forecast tells a retailer what to order. Operations decide whether the order is placed, confirmed and delivered on time. For most small retailers, the second is where the stock-outs come from.
• Guide · CostWhat AI costs a retailer in 2026Retail AI is priced per register, per interaction, per profile or per company. The total depends less on the headline price than on volume and on what the plan leaves out.