Question 1

What is AI dynamic pricing for ecommerce?

Accepted Answer

It recommends or applies a price per product from live signals instead of a fixed list. Sellarix combines demand, stock cover, competitor position and segment signals, clamped to your price band and lifted to clear your margin floor. Each recommendation explains itself, recommend-only or auto-apply.

Question 2

How does competitor price monitoring work?

Accepted Answer

Rival listings come from a price-intelligence feed and match your products by SKU or GTIN, or by similarity above a relevance floor. Each match records its score and method, then timestamped prices place you below, at or above the market median, feeding your pricing. Adapters take Wiser or Intelligence Node feeds.

Question 3

Is AI pricing better than a rule-based repricer?

Accepted Answer

You get both: the control of rules and the reasoning of AI. Pricing, markdown, matching and ranking are plain formulae over your own data, each with an explanation, so every change is traceable. You still set the price band and margin floor as hard guardrails.

Question 4

Will dynamic pricing ever sell below my cost or margin?

Accepted Answer

No. You set a cost and minimum margin floor per rule, and the floor always wins, so a price is never below cost even when demand or competitors push downward. Your price band sets floor and ceiling guardrails, and markdown trajectories obey the same band.

Question 5

What is product catalogue enrichment and auto-attribution?

Accepted Answer

It fills missing product data so search, filtering, recommendations and ad feeds work. Sellarix auto-attribution extracts a missing value from a product's text and image, then writes it through the validation boundary so it is type-correct. It also normalises units and de-duplicates near-identical products.

Question 6

How does AI merchandising decide product listing page order?

Accepted Answer

Sellarix ranks pages by blending a relevance prior with margin, stock and 30-day velocity, under a strategy preset or your own weights. Ranking is inventory-aware, so low-stock and out-of-stock products drop unless pinned. Your pin, boost and bury rules sit on top, and each score shows its breakdown.

Question 7

Does Sellarix support markdown and promotion planning?

Accepted Answer

Yes. The markdown planner builds a dated, stepped discount plan to hit a sell-through target by a deadline, sizing each step from real velocity and your cost. The deepest discount stays inside your price band, and every step is a concrete price and date you can review first.

Question 8

How do I find duplicate products and fix my catalogue taxonomy?

Accepted Answer

Sellarix de-duplicates near-identical products and normalises your taxonomy as part of catalogue clean-up, over your own data. It matches by SKU or GTIN where present, otherwise by similarity fused with attribute overlap above a floor, and records the match score and method. No attribute names are hard-coded.

Question 9

Can I keep a human in the loop or fully automate pricing?

Accepted Answer

Both, per rule. Recommend-only produces a recommendation and explanation for a person to approve, good for high-value lines. Auto-apply writes the price straight to the catalogue while respecting your margin floor and band. Many merchants run hero SKUs on auto-apply and the long tail on recommend-only.

Question 10

How is Sellarix different from a standalone repricer or merchandising tool?

Accepted Answer

Bolt-on tools each keep their own copy of your data behind a separate integration and bill, with limited reasoning. Sellarix runs pricing, markdown, competitor monitoring, merchandising, clean-up and assortment planning on one shared spine, so they agree on cost, stock and demand. One contract, one source of truth.

	Sellarix	Traditional / point tools
Where the data lives	One shared spine: pricing, merchandising, competitor and catalogue jobs read the same catalogue, costs and stock.	Each bolt-on keeps its own copy, so cost, stock and demand can disagree across tools.
Explainability	Every price, rank, markdown and match shows the signals that moved it and by how much.	Many repricers and widgets give a number with little or no reasoning you can audit.
Use of generative AI	Used in one place: catalogue attributes, written through a validating boundary, nothing written when unsure.	Some tools apply LLMs across pricing and ranking, where a confident wrong output slips through.
Control and governance	You set cost, margin floor, band and weights, recommend or auto-apply. The margin floor always wins.	Guardrails vary by tool, and stitching them across products leaves gaps.
Catalogue quality	Cleaning, auto-attribution and dedup are built in, so search, feeds and ads run on enriched data.	Catalogue enrichment is usually a separate project, and merchandising runs on whatever data exists.
Compliance posture	Compliant by default: tenant-isolated data, EU AI Act disclosure where shoppers meet AI, agent-ready feeds.	Compliance is left to the merchant to assemble across each separate tool.

AI merchandising and dynamic pricing, on any platform

Overview

How it works

One data spine underneath

Plain formulae, one AI touch

An explanation on every decision

Recommend or auto-apply

Capabilities

Use cases

End-of-season clear-out, planned not panicked

Holding price position without watching tabs

Collection pages that lead with what sells

How Sellarix compares

Merchandise & Price FAQ

Part of one platform

Bring this to your store