HOW BALGRAD.AI WORKS

A good price needs context.

We compare the exact product, real retailer offers, and the price history we actually have.

01

Match the exact product

Search by model, category or a phrase such as ‘cordless drill under €200’. Exact identifiers and variants stay separate so a body-only tool is never silently mixed with a kit.

02

Compare retailer prices

We keep prices in the retailer’s original currency and distinguish item price from shipping. Unknown shipping is never treated as free.

03

Put the price in context

Verified price checks build history over time. When the evidence is strong enough, the Deal Score adds context to the current offer.

What a Deal Score means

The current version uses a transparent rule-based calculation. It measures the current delivered price against the average of observed daily lowest delivered prices for the same product, market and new condition.

BUY · 80–100

A strong price in our available history.

WAIT · 45–79

No strong price reason to act immediately.

PASS · 0–44

Priced significantly above its observed average.

We start at 65 points, add 150 times the relative discount to the observed 30-day average, and add 15 points when the current total is within 2% of the observed 30-day low and at least 3% below the observed average. The result is rounded and limited to 0–100. The thresholds are rules, not a probability or a prediction of future prices.

When we withhold a rating

A rating needs at least 14 observed days spanning 14 days in the recent 30-day window, plus a verified, in-stock offer checked within 72 hours with known shipping. Source snapshots, unknown shipping, other currencies, used products and unavailable offers do not establish a score.

If there is too little evidence, you will see “Too early to call it a deal.” We do not manufacture a score, a discount, or missing history.

What the Public Beta covers

The beta currently focuses on power tools and selected related products across multiple retailer markets. We are expanding product, retailer and market coverage continuously. Some offers are tracked from retailer sources while others may still be source snapshots, so the final price, stock and delivery terms should always be confirmed with the retailer.

Search has limits

Budget and category parsing work without an external AI model. Use-case phrases currently match keywords and listed specifications. They are not expert suitability assessments, and we do not convert currencies when comparing global coverage.

Independent recommendations

Retailer relationships do not affect the Deal Score. Paid placements do not buy a better score, ranking or recommendation. BALGRAD.AI does not sell these products or guarantee the lowest price across the entire market.

Explore the catalog