Price intelligence is the process of turning observed prices into a usable dataset by resolving offer identity, comparison status, quality, and history.
A feed of competitor prices is source data. It becomes price intelligence when the team can tell which offer each number represents, whether that row can be compared, how fresh and complete it is, and what happened to uncertain records.
The distinction matters when evaluating a process or software category. Monitoring, matching, validation, analysis, optimization, and repricing are separate jobs. A product may support one or several of them. Calling every price feed “price intelligence” hides the work that still belongs to people or other systems.
What price intelligence means
Price intelligence is a controlled dataset about market prices and the offers behind them. It connects observations to identified products or packages, records comparison status, retains time and coverage, and routes exceptions for review before the data enters pricing, packaging, positioning, or promotion analysis.
Practical definition: Price intelligence carries the price, offer identity, comparison status, quality, and observed context together. It prepares evidence for a decision; it does not make the decision.
The process can be small. A team comparing a few public plans can maintain a careful manual record. A retailer with a large changing catalog may need specialist matching, validation, and exception workflows. The required system depends on product scale, comparability, decision speed, and the cost of a wrong conclusion.
Move each observation through four states
| State | Question | Output |
|---|---|---|
| Raw observation | What amount and visible conditions appeared, where, and when? | Unchanged source record |
| Resolved offer | Which exact product, plan, variant, or bundle does the record describe? | Identity with unresolved attributes marked |
| Trusted record | Are currency, unit, terms, availability, and source state acceptable? | Accepted, provisional, rejected, or exception status |
| Usable dataset | Which accepted rows belong in the defined comparison and time window? | Rows with comparison class, history, and coverage attached |
Each state constrains the next. A current price attached to the wrong product is unusable. A correct identity with an unknown quantity may remain a valid observation but cannot enter a per-unit comparison. Keeping these states separate prevents an attractive dashboard from overstating the underlying data.
Separate monitoring, intelligence, analysis, and optimization
| Job | Primary question | What it should produce |
|---|---|---|
| Price monitoring | What visible price changed after the baseline? | Dated observations and events |
| Price intelligence | Can those observations be trusted and compared in context? | Validated evidence with identity, confidence, history, and coverage |
| Pricing analysis | What does the evidence mean for this market or decision? | Scoped market or price-position conclusion |
| Price optimization | Which price is expected to improve a defined outcome under constraints? | Modeled recommendation requiring demand, margin, and response evidence |
| Repricing | Which approved price should be changed, where, and when? | Controlled execution with policy and rollback |
Use the guide to monitoring competitor prices for the collection workflow. Use the competitor pricing analysis method when a validated evidence set needs a market or positioning interpretation. Neither step supplies an automatic optimization or repricing decision.
Assign a comparison status instead of guessing equivalence
Cross-store and cross-plan price intelligence depends on product comparability. Exact identifiers can support high-confidence matching for standardized goods. Many categories require a reviewed comparison class instead: similar size, material, service, capability, contract, or customer outcome. A matching name or image is not enough.
- Same offer: Identifiers and material attributes support the same product, plan, or documented equivalent.
- Ruled comparable: The offers differ, but an explicit class rule makes a bounded comparison valid.
- Needs review: Identity, configuration, or terms remain unresolved, so the row cannot yet enter the comparison.
- Excluded: The offer may be market context, but it fails the current comparison rule.
Matching confidence must travel with the price. The competitor product analysis guide provides the deeper evidence method when capability, workflow, or customer value determines whether offers are comparable.
Judge dataset quality across five dimensions
| Dimension | Question | What a weak result changes |
|---|---|---|
| Identity | Is the exact offer resolved, including variant or package? | The row stays provisional or is rejected |
| Terms | Are currency, unit, commitment, availability, and visible conditions known? | Only a narrower comparison is allowed |
| Freshness | Is the observation recent enough for the intended use? | The row can support history but not a current snapshot |
| Coverage | Which competitors, products, regions, and scheduled checks are represented or missing? | The dataset cannot support a claim about the omitted scope |
| History | Are raw observations, corrections, and match changes retained? | Persistence and change claims become difficult to audit |
Preserve the raw observation when a reviewer corrects a currency, rejects a match, or revises a unit. Store the disposition and reason beside it. History can then show whether a value persisted, whether a match rule changed, and whether a gap came from the market or the collection process. It still cannot reveal why a competitor chose a price.
Use exception review instead of hiding uncertain data
Repeated collection will produce ambiguous records. Product identities change, pages move, variants appear, and sources return partial data. Send cases that could alter the comparison to a review queue with a reason and a clear disposition.
- Unresolved product or package identity.
- Currency, quantity, billing unit, or material terms are missing or inconsistent.
- The offer appears promotional, restricted, unavailable, replaced, or duplicated.
- A material attribute changed after the previous match.
- The source is partial, failed, or structurally different from its prior state.
Accept, correct, narrow, exclude, or keep the record provisional, and retain the reason. Quietly filling a gap creates a clean dashboard and a weak dataset.
Evaluate price intelligence software by the evidence workflow
A long source list or frequent refresh promise does not establish decision quality. Evaluate the system against the products, channels, markets, and actions the team actually needs. Ask to inspect how it represents uncertainty and exceptions, not only the finished chart.
- Coverage: which stores, regions, product types, and price conditions are actually supported?
- Matching: which identifiers and attributes drive matches, and how can a reviewer correct them?
- Quality: how are units, terms, unavailable products, stale rows, and failed sources represented?
- History and workflow: are raw records and corrections retained, and can teams resolve exceptions without overwriting them?
- Action boundary: does the product collect, prepare, analyze, recommend, execute, or combine those jobs?
For a small catalog, a carefully governed manual process may be sufficient. For a large assortment, manual matching and validation can become the dominant cost. Scale changes the tooling requirement, but it does not remove the need for explicit definitions and review.
Worked example: an apparent category undercut
Harbor Supply is a hypothetical home-goods retailer. A competitor feed shows six storage products priced below Harbor's visible amounts. The initial chart labels the competitor 18 percent cheaper across the category.
Review resolves two rows as smaller packs, one as member-only, one as out of stock, and one as a different material and warranty class. Those five rows are excluded from the category comparison. The remaining row satisfies the class rule and remains lower across two healthy observations, although shipping is still unknown.
The price-intelligence result is not “the category is 18 percent cheaper.” It is one accepted row, five documented exclusions, and an open shipping field. A separate pricing analysis can decide whether that accepted row supports a wider market conclusion.
Where Content Radar fits in price intelligence
Content Radar Product Monitoring can add compatible public Shopify, WooCommerce, and structured custom stores after a bounded compatibility scan. A baseline and later manual or daily scheduled checks can produce supported product, same-known-currency price, and availability events. Recent Changes and in-app alerts make those events available for review.
That is one public-evidence input into a broader price-intelligence process. Content Radar does not cover every ecommerce store or marketplace, match products across stores, normalize currencies, identify promotions, enforce MAP policies, estimate demand, calculate optimal prices, recommend actions, or reprice a catalog.
The retail competitor analysis guide shows how bounded product observations can be combined with publishing and internal evidence. Keep unsupported channels and missing commercial context visible in the final conclusion.
A price-intelligence dataset is ready for use when another reviewer can reconstruct each accepted row and understand every rejection or unresolved exception. More rows do not compensate for uncertain identity or hidden coverage gaps.
Understand the bounded price-change input
Review the supported public-store events Content Radar can contribute to a wider human-governed price-intelligence process.