Polar Analytics reviews in 2026: 4.9 stars, and a USD 750 floor

The public data, verified today

On the Shopify App Store the app is called Polar:AI Platform for commerce, and these are the numbers as of October 30, 2026:

• 4.9 stars across 116 reviews.
• 97% five stars — 113 five-star reviews, none at four, none at three, none at two, and three at one star (3%). That distribution is unusual: there is no middle.
• Published October 28, 2020. Five years in market, developer in Paris.
• Languages: English, French and German.
• Price on the listing: "Core Plan, from USD 750/month," with the note that it is billed separately from your Shopify invoice.

Their own site further claims "4,000+ ecommerce brands and agencies" and "400+ pre-built ecommerce metrics." That is what they say, not a third party, and we label it as such. We could not verify the install count: storeleads.app returns a 404 for the app, so we publish no install number at all. If you see one in another comparison, ask where it came from.

What it does better than anything else in this series

There are four things we do not have at any price, and they are not checklist details.

Amazon, properly

Connectors for Amazon Seller Central and Amazon Vendor Central, with orders, fees and payouts. We do not have Amazon. If you are a brand selling on Shopify and Amazon and you want both sets of books in one place, they win that fight without argument. They also have retail channels — Sephora and Target among their connectors — which is a world we are not in.

Your own database

Every plan includes a dedicated Snowflake database with the keys in your hands, a semantic layer with 400+ pre-built metrics, and SQL access as an add-on. We give you a dashboard; they give your data team a warehouse to build on. With Polar MCP that layer becomes available to Claude and ChatGPT, refreshed every 15 minutes.

Incrementality tests with a data scientist

This is not software, it is a service: a dedicated data scientist running lift tests on Meta, Google, TikTok or TV, measuring impact across Shopify, Amazon or retail channels, with reports and recommendations. It costs USD 3,200 for the first test and USD 4,000 for each one after. If you spend USD 500,000 a month on media, a test that finds 10% of waste pays for itself 15 times over. We have nothing like it, and it is the only honest answer to the question "did this ad cause the sale or just claim it?"

An agent that acts, not one that alerts

Polar Operator lives in Slack or Teams, monitors your store and your ads, answers with the definition behind every number, and once you approve, acts: shifts ad budgets, pauses ads for sold-out products, plans markdowns. That is the difference between an alert and an action, and we are on the alert side of it.

On top of all that: 326 connectors across nine categories, unlimited users, unlimited historical data, a dedicated Success Manager and a Slack channel on every plan. And one thing almost nobody in this category does: they calculate real profit and contribution margin, with LTV and cohorts by channel and by product. This is not a pixel dressed up as a dashboard.

So what is the problem?

Three things, and none of them is a flaw in the product.

1. The floor sits above our ceiling

Their cheapest plan is USD 750 a month per their Shopify listing, and USD 720 per their own pricing page's data — both verified today. Our most expensive plan, the 100,000-orders-a-month one, costs USD 549. That comparison needs no average order value and no conversion rate: Polar's entry price sits above Escalafy's ceiling. A store doing USD 1 million a year — a very good store in Argentina — pays the same USD 720 as one doing USD 4.9 million, because the first band covers all of it.

2. You cannot see the price without booking a demo

Their pricing page renders the price fields empty and they only fill in once you pick your annual revenue band. There is no free plan, no self-serve signup and no published trial: every button says "Book a demo." The full 18-band ladder travels embedded in the page itself and we publish all of it in Polar Analytics pricing, but to sign up there is a sales call in the way. For a founder putting out a fire on a Tuesday at eleven at night, that is already an answer.

3. Your half of the business is not on their list

We searched today, one by one, on their own 326-connector page: Tiendanube, Nuvemshop, Mercado Libre, Mercado Pago, VTEX and WooCommerce all return zero results. The only e-commerce platform they name is Shopify, and that page's header says so plainly. The marketplaces they do support are Amazon Seller Central, Amazon Vendor Central, TikTok Shop, Sephora and Target.

And here is the point that matters most, because it is the one a feature table does not show: Polar calculates profit, but with the costs it can see. No Mercado Pago fee by payment method, no settlement timing, no financing cost for offering 3, 6 or 12 interest-free installments, no Mercado Libre category commission or Full fulfillment costs, no IIBB (gross receipts tax) and no VAT. In a store selling in Argentina that is the difference between a 45% gross margin and an 11% net margin. The number will be well calculated and it will be wrong.

Who it is for, who it is not

Buy Polar Analytics if:

• You sell on Shopify and Amazon, or through retail channels, and you want both sets of books in one place.
• You run multiple stores or multiple brands — it is what merchants themselves highlight in their reviews.
• You have a data team that wants SQL, its own warehouse and a semantic layer to build on.
• You spend enough on media that a USD 3,200 incrementality test pays for itself.
• You want an agent that acts in Slack, not one that alerts.
• Your revenue is eight figures and USD 750 a month is noise in your budget.

Do not buy it if:

• Your store runs on Tiendanube. There is no app to install: the product does not exist there.
• Part of your revenue goes through Mercado Libre. That whole channel is out, along with its commissions and Full fulfillment costs.
• You collect through Mercado Pago or offer interest-free installments. It is the cost that eats the most margin here and it is not in the model.
• You need to know when each sale's money lands by payment method and installment count, to plan your next stock order.
• You want to see the price, try the tool and decide without a sales call.
• You do less than USD 5 million a year and USD 750 a month is a decision, not a detail.

The honest conclusion

Polar Analytics is not a product you can criticize. It is a product that was not built for you if you sell in Argentina. It has been in market five years, it holds 4.9 stars with 97% at five, it answers the profit question for real, and on top of that it brings a data warehouse, Amazon, incrementality with a data scientist and an agent that acts. If we had to recommend a tool to an eight-figure brand in the United States running Shopify plus Amazon, it would be on the short list.

What it cannot do is tell you whether your month closed in the black, because half of your costs do not go into its model and half of your sales do not appear on its connector list. And what it also cannot do is cost less than USD 720 a month. Those two things together, for a Tiendanube store selling on Mercado Libre and collecting in installments, settle the decision before the demo.