LatentFactors

Solutions

Commerce intelligence, from data to action.

Ready-to-use datasets, specialized models applied to your data, and an emerging execution layer for intelligent agents.

Data as a Service

Commerce intelligence. Ready to use.

Turn market signals into decisions with structured sales and search datasets. Integrate our intelligence through APIs or scheduled feeds, tailored to your data formats, reporting periods and business needs.

Built for intelligence providers, digital commerce platforms, agencies and brands.

Amazon Sales Intelligence

Know what sells. Act on what changes.

High-accuracy Amazon product sales estimates with consistent trends and broad category coverage. Track units ordered and sales value at daily or weekly frequency to monitor market shifts, assess competitive performance and identify growth opportunities.

Connect sales intelligence with pricing, promotions, availability and product-page history to investigate changes in performance. Receive structured outputs directly in your platform or internal analytics environment.

Frequency
Daily and weekly, on reporting weeks you configure.
Freshness
Within one day of supporting Amazon reporting, then scheduled revisions.
Context
Optional pricing, promotion, availability and product-page change feeds.
Traceability
Historical backfills, revision identifiers and confidence indicators.
Delivery
APIs, scheduled feeds or your own formats, into your platform or analytics environment.
Amazon Sales Intelligence – Full CoverageLF / 02
Amazon Sales Intelligence, Full Coverage: the structure of the solution Schematic, no data. Left, the catalogue as a field of signals; one product is followed into the estimate at the centre, daily or weekly, with four fields: units ordered, sales value, revisions and confidence; two outlines behind it are its earlier versions. Above, four optional feeds, pricing, promotions, availability and page changes, join the estimate on the same product and day. Below, the time axis: history, backfilled, then the first estimate one day after Amazon reporting and the scheduled revisions. Right, three outputs: API, scheduled feeds, your formats. Animated: the catalogue appears, every Amazon category, left to right ; one product is followed ; its line enters the estimate ; units ordered ; sales value ; revisions ; confidence ; its earlier versions appear behind the card ; the optional feeds join it ; history, backfilled ; the first estimate, one day after Amazon reporting ; a scheduled revision ; the current revision ; delivery ; API ; scheduled feeds ; your formats. DELIVERY API scheduled feeds your formats OPTIONAL FEEDS pricing promotions availability page changes CATALOGUE every Amazon category DAILY · WEEKLY units ordered sales value revisions confidence history +1 day scheduled revisions Amazon Sales Intelligence, Full Coverage: the structure of the solution Schematic, no data. Left, the catalogue as a field of signals; one product is followed into the estimate at the centre, daily or weekly, with four fields: units ordered, sales value, revisions and confidence; two outlines behind it are its earlier versions. Above, four optional feeds, pricing, promotions, availability and page changes, join the estimate on the same product and day. Below, the time axis: history, backfilled, then the first estimate one day after Amazon reporting and the scheduled revisions. Right, three outputs: API, scheduled feeds, your formats. API scheduled feeds your formats OPTIONAL FEEDS pricing promotions availability page changes CATALOGUE DAILY · WEEKLY units ordered sales value revisions confidence history +1 day revisions
TrackConnectReceive
Discuss Amazon sales coverage

Amazon Search Intelligence

Track demand. Detect change. Capture the opportunity.

Discover what shoppers are searching for, which products capture their attention and where demand becomes sales. Track keyword volumes, trends, product visibility, clicks and conversions for the top three products, alongside keyword-attributed sales and historical performance.

Identify emerging demand, competitor gains and new keyword opportunities. Use the intelligence to prioritize content, guide advertising budgets and find opportunities to improve conversion.

Coverage
Daily, across keywords, products, brands and categories in supported Amazon markets.
Demand
Keyword volumes and trends, linked to product visibility and performance.
Top products
Clicks and conversions for the top products.
Sales
Keyword-attributed sales and historical trends.
Delivery
APIs and scheduled feeds, integrated into your analytical workflows.
Amazon Search IntelligenceLF / 03
Amazon Search Intelligence: the structure of the solution Schematic, no data. Left, a field of small squares: keywords, products, brands and categories in the supported Amazon markets; one keyword is lit and followed into the daily record at the centre, with four fields: search volume and trend, product visibility, clicks and conversions, attributed sales. Above the record, the top products join it. Below, the time axis: history, then one mark per day up to today. Right, two outputs: API, scheduled feeds. Animated: keywords, products, brands and categories appear, left to right ; one keyword is followed ; its line enters the daily record ; search volume, trend ; product visibility ; product #1 ; product #2 ; product #3 ; the top products join the record ; clicks, conversions ; attributed sales ; history ; one mark per day ; today ; every day, today ; delivery ; API ; scheduled feeds. KEYWORDS TOP PRODUCTS DELIVERY API scheduled feeds #1 #2 #3 keywords, products,brands and categories DAILY search volume, trend product visibility clicks, conversions attributed sales history every daytoday Amazon Search Intelligence: the structure of the solution Schematic, no data. Left, a field of small squares: keywords, products, brands and categories in the supported Amazon markets; one keyword is lit and followed into the daily record at the centre, with four fields: search volume and trend, product visibility, clicks and conversions, attributed sales. Above the record, the top three products join it. Below, the time axis: history, then one mark per day up to today. Right, two outputs: API, scheduled feeds. API scheduled feeds TOP PRODUCTS #1 #2 #3 KEYWORDS DAILY search volume, trend product visibility clicks, conversions attributed sales history every daytoday
DemandAttentionSales
Discuss search intelligence

Models & Applied AI as a Service

Your data. More powerful decisions.

Apply our specialized AI/ML models to your own data to estimate sales, forecast demand and harmonise and code data.

We manage the modeling, processing, and delivery, so your teams can focus on insights, customer value, and execution.

Dedicated Sales Estimation

Your data advantage. Our models. Your dedicated estimation service.

Build sales intelligence around your own data assets. A dedicated managed service combines authorized sales contributions and shopper-panel data, with isolated training, calibration and delivery tailored to your organization.

Expand your analytical capabilities while retaining control over your contributions and integrating results into your existing products.

For
Intelligence providers, with a dedicated model.
Isolation
Private training, calibration and delivery, isolated for your organization.
Your data
Provider-specific sales and panel data, under your control.
Outputs
Customized outputs through a managed API or scheduled feeds.
Dedicated Sales EstimationLF / 04
Dedicated Sales Estimation: the structure of the service Schematic, no data. Left, two fields of small squares: your authorized sales contributions and your shopper-panel data; one item of each is lit and followed into a bounded zone at the centre, your dedicated service, labelled isolated and private, where training, calibration and delivery happen in that order and produce your estimates: a dedicated model, calibrated on your data, with customized outputs. Right, three outputs: managed API, scheduled feeds, your products. Below: you keep control of your contributions. Animated: your sales contributions appear ; your shopper-panel data appear ; one contribution is followed ; one panel record is followed ; the contributions enter the service ; the panel data enter the service ; the dedicated service closes around them: isolated, private ; training ; calibration ; delivery ; your estimates: a dedicated model ; calibrated on your data ; customized outputs ; delivery ; managed API ; scheduled feeds ; your products ; you keep control of your contributions. YOUR DATA YOUR DEDICATED SERVICE DELIVERY ISOLATED · PRIVATE training calibration delivery sales contributions shopper-panel data managed API scheduled feeds your products YOUR ESTIMATES dedicated model calibrated on your data customized outputs you keep control of your contributions Dedicated Sales Estimation: the structure of the service Schematic, no data. Left, two fields of small squares: your authorized sales contributions and your shopper-panel data; one item of each is lit and followed into a bounded zone at the centre, your dedicated service, labelled isolated and private, where training, calibration and delivery happen in that order and produce your estimates: a dedicated model, calibrated on your data, with customized outputs. Right, three outputs: managed API, scheduled feeds, your products. Below: you keep control of your contributions. ISOLATED · PRIVATE training calibration delivery YOUR DATA sales contributions,shopper-panel data managed API scheduled feeds YOUR ESTIMATES dedicated model calibrated on your data customized outputs
ContributeCalibrateDeliver
Discuss a dedicated model

Data Harmonisation & Coding

Structure, enrich, and connect product data.

Structure and clean your inconsistent product information, align taxonomies and connect corresponding products across datasets. Our graph-based approach brings together product attributes and category relationships to support coding, classification and matching.

Build consistent product views for reporting, comparisons and downstream analysis, using outputs aligned with your data requirements.

Cleaning
Inconsistent product information, structured and cleaned.
Taxonomies
Aligned across datasets, from product attributes and category relationships.
Matching
Corresponding products connected across datasets.
Outputs
Consistent product views for reporting, comparisons and downstream analysis, aligned with your data requirements.
Data Harmonisation, Coding & MatchingLF / 05
Data Harmonisation, Coding & Matching: the structure of the solution Schematic, no data. Left, two irregular fields of small squares: dataset A and dataset B, your product information as it comes; one item of each is lit and followed into one card at the centre, one product, with four fields: category coded, attributes structured, taxonomy aligned, matched across datasets. Above the card, a small graph: a category, its parent and two attributes, the relationships the approach uses. Right, three outputs: reporting, comparisons, analysis. Below: consistent product views, aligned with your data requirements. Animated: dataset A appears, as it comes ; dataset B appears, as it comes ; one record of A is followed ; one record of B is followed ; A enters the one-product card ; B enters the one-product card ; category, coded ; attributes, structured ; the category and its parent ; its attributes ; the relationships feed the card ; taxonomy, aligned ; matched across datasets ; outputs ; reporting ; comparisons ; analysis ; consistent product views, aligned with your data requirements. YOUR DATASETS ATTRIBUTES · CATEGORIES OUTPUTS dataset A dataset B reporting comparisons analysis ONE PRODUCT category, coded attributes, structured taxonomy, aligned matched across datasets consistent product views, aligned with your data requirements Product Coding & Matching: the structure of the solution Schematic, no data. Left, two irregular fields of small squares: dataset A and dataset B, your product information as it comes; one item of each is lit and followed into one card at the centre, one product, with four fields: category coded, attributes structured, taxonomy aligned, matched across datasets. Above the card, a small graph: a category, its parent and two attributes, the relationships the approach uses. Right, three outputs: reporting, comparisons, analysis. Below: consistent product views, aligned with your data requirements. YOUR DATASETS dataset A dataset B reporting comparisons analysis ONE PRODUCT category, coded attributes, structured taxonomy, aligned matched across datasets
CleanAlignMatch
Discuss data harmonisation

Forecasting & Estimation

Anticipate demand. Prioritize your next move.

Apply forecasting and estimation models to your own data, with the option to incorporate LatentFactors intelligence. Generate predictions tailored to your business context to support planning and investment decisions.

Your data
Forecasting and estimation models, applied to your own data.
Optional
LatentFactors intelligence, incorporated at your option.
Context
Predictions tailored to your business context.
Use
Planning and investment decisions.
Forecasting & EstimationLF / 06
Forecasting & Estimation: the structure of the solution Schematic, no data. Left, a field of small squares: your own data, scattered on the left and settling into a lattice; one item is lit and followed into the card at the centre, forecast and estimate, with three fields: forecast, estimate, your business context. Above the card, a dashed chip, LatentFactors intelligence, optional, joined to it by a dashed line. Below, the time axis: history with its marks, today, then the horizon ahead, dashed, with outlined marks. Right, two uses: planning, investment decisions. Animated: your own data appear, left to right ; one record is followed ; its line enters the card ; history, learned ; today ; LatentFactors intelligence joins, optional ; forecast ; estimate ; your business context ; the horizon ahead ; a first prediction ; a further one ; decisions ; planning ; investment decisions. YOUR DATA OPTIONAL DECISIONS planning investment decisions LatentFactors intelligence your own data FORECAST · ESTIMATE forecast estimate your business context history ahead today Forecasting & Estimation: the structure of the solution Schematic, no data. Left, a field of small squares: your own data, scattered on the left and settling into a lattice; one item is lit and followed into the card at the centre, forecast and estimate, with three fields: forecast, estimate, your business context. Above the card, a dashed chip, LatentFactors intelligence, optional, joined to it by a dashed line. Below, the time axis: history with its marks, today, then the horizon ahead, dashed, with outlined marks. Right, two uses: planning, investment decisions. planning investment decisions OPTIONAL LatentFactorsintelligence YOUR DATA FORECAST · ESTIMATE forecast estimate your business context history today ahead
LearnAnticipateDecide
Discuss your forecasting needs

Agentic AI

More intelligence. Less context overhead.

Planned · Early access

Give your AI agents access to commerce intelligence without sending entire datasets into their context. Our planned query and execution agent will retrieve the relevant information, run the requested analysis and return compact, structured results.

Designed to reduce unnecessary token consumption and help your agents accomplish more with less data handling.

Planned · MCP and A2A

Agentic Commerce Intelligence

Ask for the outcome. Receive the intelligence.

Connect your agents to LatentFactors datasets and models through MCP, with A2A support planned for delegated agent tasks. From finding rising search demand to comparing sales trends or processing product data, the execution layer will perform the work close to the data and return the results your agent needs.

Retrieval
Targeted, by market, product, keyword and period.
Execution
Queries, comparisons and aggregations, executed close to the data.
Results
Compact, with source references, timestamps and relevant quality indicators.
Access
MCP access to tools and intelligence; A2A support planned for agent-to-agent delegation.
Agentic Commerce Intelligence · PlannedLF / 07
Agentic Commerce Intelligence, planned: the structure of the solution Schematic, no data. Left, your agent, lit, with a second agent below it joined by a dashed line: A2A, planned. From your agent a lit line, the ask, enters a bounded zone at the centre, the execution layer, MCP: inside, a field of small squares, datasets and models, feeds three operations, retrieve, compare, aggregate, close to the data. Their output goes to a card on the right, compact result, with four fields: results, source references, timestamps, quality indicators. A line returns from the card, under the zone, back into your agent. Below: retrieval by market, product, keyword and period; results back to your agent. Animated: the ask draws from your agent into the execution layer ; the datasets and models answer, column by column ; the operations are reached ; retrieve ; compare ; aggregate ; the output goes to the result card ; results ; source references ; timestamps ; quality indicators ; the return draws back under the layer ; a dot travels it into your agent ; the agent receives. YOUR AGENT EXECUTION LAYER · MCP RESULT your agent ask A2A, planned datasets and models,close to the data retrieve compare aggregate COMPACT RESULT results source references timestamps quality indicators retrieval by market, product, keyword and period results back to your agent Agentic Commerce Intelligence, planned: the structure of the solution Schematic, no data. Left, your agent, lit, with a second agent below it joined by a dashed line: A2A, planned. From your agent a lit line, the ask, enters a bounded zone at the centre, the execution layer, MCP: inside, a field of small squares, datasets and models, feeds three operations, retrieve, compare, aggregate, close to the data. Their output goes to a card on the right, compact result, with four fields: results, source references, timestamps, quality indicators. A line returns from the card, under the zone, back into your agent. Below: retrieval by market, product, keyword and period; results back to your agent. YOUR AGENT your agent A2A, planned datasets, modelsclose to the data retrieve compare aggregate COMPACT RESULT results source references timestamps quality indicators
AskExecuteReceive
Join the agentic early-access list