Product manager: building the range on reliable figures, without spending the day gathering them.
The product manager decides what the brand or the company puts on sale: which products, at what price, for which customers, on what date. Yet a large share of the job goes into gathering sales, costs and feedback scattered across the ERP, the website, the PLM and reports from the sales force. Automation gives that time back for the choices that matter.
The role
Between the market, development, sales and finance.
The product manager turns market expectations into a coherent offer: a range, prices, margins, a launch calendar. They write the brief, follow development through to launch, equip the sales teams and then measure what works. In a mid-sized company, they often carry several families at once and link departments that share neither the same tools nor the same figures.
Day to day
A typical day.
- MorningReview yesterday’s sales by style and by channel, and spot deviations from forecast.
- MorningCheck in with development on the progress of prototypes and samples.
- MiddayCheck a cost price received and its effect on margin and selling price.
- AfternoonWrite or update a product brief for the next collection or range.
- AfternoonPrepare selling arguments and sheets for the sales force and e-commerce.
- End of dayAnswer requests from markets: prices, availability, competing products, new variants.
The objectives
What the role is measured on.
- Revenue by family
- Margin by style
- Launch dates met
- Sell-through rate
- Share of new products in sales
- Range rationalisation
Indicators are named, never quantified: levels depend on each company, and the diagnostic measures them.
Day to day
The problems that come up most.
Sales scattered across channels
Stores, website, marketplaces, resellers: each channel has its own extract, and the view by style is rebuilt by hand.
Costs that arrive late
The selling price is set before the cost price is stable, and the real margin is discovered after launch.
Briefs that get lost along the way
The brief goes out as a document, then the trade-offs are made by e-mail and the final version is written down nowhere.
A range that keeps growing
Styles are added every season and few are removed, for lack of a clear view of what each one really brings in.
Customer reviews left unused
Ratings, comments and returns are read case by case, with no summary by product or by reason.
Requests from the field piling up
Sales reps and markets send their requests by e-mail or phone, with no priority and no shared record.
By sector
One role, several realities.
- Fashion and luxury
Seasonal collections and carry-overs, variants in materials and colours, prices harmonised between countries and boutiques, wholesale and retail: the product manager balances desirability, margin and workshop capacity.
- Retail
An own-brand offer built with suppliers, negotiated purchase prices, store-by-store layouts and tight promotional calendars: they steer the range as close to sales as possible.
- Cosmetics and beauty
Launches driven by formula, packaging and claims, seasonal gift sets, limited editions and selective channels: the launch calendar depends as much on regulatory affairs as on marketing.
- B2B distribution and wholesale
Thousands of items from supplier catalogues, ranges to rationalise, own brands to develop and purchasing terms that keep moving: they decide what comes in and what goes out.
- Industry
Technical ranges, options and variants, catalogue and made-to-order products, competitors watched closely: they link customer needs, the engineering office and the sales force.
- Pharma and healthcare
Health products, supplements and devices with authorisation and labelling constraints, long life cycles, country-by-country launches: every change to the range goes through regulatory affairs.
What changes
Automate the task, keep the judgement.
What gets automated
- Sales, margins and stock by style consolidated every day from the ERP, the website and sales channels, in a single table.
- The selling price recalculated from the cost price and the target margin, by country and by channel, as soon as a cost changes.
- The first draft of the product brief, the sales pitch and the launch sheets, written by AI from the style’s data.
- Customer reviews and return reasons classified by AI, by product and by theme, with a regular summary.
- Product committee preparation: progress, costs, margins and blocking points gathered automatically.
What stays human
- The choice of products to launch, evolve or discontinue.
- Positioning and the final price.
- Trade-offs between margin, lead time and the ambition of the range.
Use cases
How to automate, case by case.
Each case starts from a question leadership teams ask. The starting point, what gets automated, the tools involved, and what stays in the teams’ hands.
How do you automate the end-of-season review by style?
- Today
- At the end of the season, the product manager extracts each channel’s sales, remaining stock, markdowns and returns, then gathers them in a spreadsheet to prepare the next collection. The review arrives late, after the first decisions for the next season have already been made.
- What gets automated
- The review is calculated continuously by style, colour and size: sales, sell-through rate, markdown, returns, achieved margin against planned margin. The tool compares with previous seasons and an AI agent writes a summary of successes and failures by family. The review is ready the day the season ends.
- What stays human
- Reading the review and what to take from it for the next collection remain the product manager’s work with merchandising.
- ERP
- Sales platforms
- Reporting tool (BI)
- AI writing agent
How do you automate calculating the selling price from the target margin?
- Today
- The product manager receives a cost price, applies channel coefficients in a spreadsheet, rounds, then converts for each country. When the cost changes, everything has to be redone, and published prices do not always match the latest calculation.
- What gets automated
- The pricing rules are written once: target margin by channel, coefficients, price-point rounding, tolerated gaps between countries, taxes. The tool reads the cost price in the PLM or the ERP and proposes the full price grid, recalculated at each cost change, with the styles whose margin falls below the threshold.
- What stays human
- The final price, and the decision to accept a lower margin on a strategic style, remain the product manager’s.
- PLM
- ERP
- Pricing tool
How do you automate competitive monitoring of prices and new products?
- Today
- The product manager regularly visits competitor sites, notes a few prices and screenshots, and shares it all by e-mail. Monitoring depends on the time available and never covers the whole range.
- What gets automated
- A list of sites and families to follow is set. The tool automatically picks up prices, new products, stock-outs and descriptions, and an AI agent matches each competing product to the styles in the range. The product manager receives a summary of notable changes, with price history.
- What stays human
- Interpreting the monitoring and the response to give (price, range, calendar) remain decided by the product manager.
- Monitoring tool
- AI matching agent
- Reporting tool (BI)
How do you automate writing product briefs?
- Today
- Each brief is written in a document starting from an old brief, completed by hand with the sales of similar products, requests from the field and the price target. Headings vary from one product manager to another, and development often has to ask for clarification.
- What gets automated
- A common brief template sets the expected headings: target, use, target price and margin, desired materials or ingredients, constraints, calendar. An AI agent prepares a first draft from the available data and the product manager’s notes, then the approved brief is saved in the PLM as the starting point for development.
- What stays human
- The product’s intent, its positioning and its approval remain written and owned by the product manager.
- PLM
- AI writing agent
- Reporting tool (BI)
How do you automate defining the assortment by channel and by country?
- Today
- The assortment is built in a large spreadsheet: which styles, which colours, which sizes for each channel, each country, each store type. The rules exist in the team’s heads, and every season everything has to be redone.
- What gets automated
- Assortment rules are formalised: depth by type of point of sale, channel exclusives, country constraints (labelling, regulation, seasonality). The tool proposes a starting assortment from these rules and past sales, and flags inconsistencies: a style with no channel, a country with no traffic driver, an exclusive not respected.
- What stays human
- Assortment choices and exceptions for a customer or a market remain settled by the product manager with sales.
- Collection plan
- ERP
- Reporting tool (BI)
How do you automate the launch kit for sales teams?
- Today
- Before a launch, the product manager gathers the sales pitch, tech sheet, visuals, prices and terms, then sends them to sales reps, boutiques and resellers, each in a different format. Questions then arrive one by one.
- What gets automated
- The kit is generated from the style’s data: sales pitch and FAQs written by an AI agent from the brief and the tech sheet, visuals taken from the PIM, prices and availability read in the ERP. It is published in the sales area and updated automatically if a piece of data changes.
- What stays human
- The key launch message and proofreading the pitch remain with the product manager and marketing.
- PLM
- PIM
- ERP
- AI writing agent
How do you automate analysing customer reviews by product?
- Today
- Reviews published on the website and marketplaces, customer service comments and return reasons are read as they come, by different people. Nobody can say which defects or expectations keep coming back on a style.
- What gets automated
- Reviews and comments are collected from each source and linked to the style. An AI agent classifies them by theme (size, comfort, material, colour, scent, packaging, price) and by sentiment, then the tool tracks how each theme evolves over time. Styles that slip are flagged to the product manager.
- What stays human
- The product changes requested from development remain decided by the product manager.
- E-commerce site
- Marketplaces
- Customer service software
- AI classification agent
How do you automate identifying styles to rationalise?
- Today
- The range grows every season. To decide what to remove, the product manager crosses sales, margins, stock and returns by hand, style by style, and the review is often postponed for lack of time.
- What gets automated
- The tool calculates each style’s real contribution: sales, margin, turnover, complexity cost (variants, short runs, single-supplier items), and spots styles that cannibalise each other. It proposes a list to review, with the figures for each and its role in the range.
- What stays human
- The decision to keep a style for image, a customer or the coherence of the range remains the product manager’s.
- ERP
- PLM
- Reporting tool (BI)
How do you automate preparing product committees?
- Today
- Before each committee, the product manager gathers the status of developments, costs, planned margins, delays and samples to present. Days go into formatting a pack that will be out of date by the next meeting.
- What gets automated
- The committee pack is assembled automatically from the PLM, the collection plan and sales: styles to approve, cost and margin gaps, late milestones, decisions expected. An AI agent writes the summary note, and the decisions taken are recorded on each style.
- What stays human
- The committee’s decisions and their rationale remain those of management and the product manager.
- PLM
- Collection plan
- Reporting tool (BI)
- AI writing agent
How do you automate collecting product requests from sales reps and markets?
- Today
- Sales reps, subsidiaries and key accounts ask for new variants, adaptations or competing products, by e-mail, phone or in meetings. Requests get lost, are repeated, and nobody knows which ones have been studied.
- What gets automated
- A single form collects requests, and an AI agent also reads those that arrive by e-mail to record them. Similar requests are grouped, linked to the family or style concerned and enriched with the customer’s or market’s sales. The requester is informed automatically of the follow-up.
- What stays human
- Choosing which requests to keep and their priority remain decided by the product manager.
- Online form
- Messaging
- CRM
- AI reading and classification agent
Questions from the teams
What teams look to improve.
- How do I automate tracking sales by style?
- First set a reference common to all channels, since that is what allows stores, website, marketplaces and resellers to be reconciled. The tool then retrieves sales, returns and stock every day from the ERP and the platforms, and presents them by style, family and channel. The product manager looks at the deviations rather than building the table.
- How do I calculate the selling price automatically from the cost price?
- By formalising the rules that already exist: coefficient or target margin by channel, price rounding, permitted gaps between countries. The tool reads the cost price in the PLM, applies these rules and proposes a price grid. It flags styles whose margin falls below the set threshold. The final price is still approved by the product manager.
- Can AI write a product brief?
- It can prepare the first draft from a brief template, the sales of similar products, requests from the field and the product manager’s notes. The benefit lies in structure and completeness: nothing is forgotten. The intent, the positioning and the trade-offs are still written by the product manager.
- How do I know which styles to discontinue?
- By automatically crossing, for each style, sales, margin, stock, turnover, returns and its role in the range (traffic driver, complement, novelty). The tool proposes a list of styles to review, with their figures. The decision takes account of what the figures do not say: image, key accounts, coherence of the collection.
- How do I use customer reviews to improve a product?
- Reviews from the website, marketplaces and customer service are collected, then an AI agent classifies them by product and by theme: size, material, durability, fragrance, packaging, price. The summary shows recurring themes and how they evolve. The product manager draws from it the changes to request from development.
- How do I prepare a product committee without spending the week on it?
- By defining the content of the pack once: development progress, costs, planned margins, delays, decisions expected. The tool assembles it from the PLM, the collection plan and sales, and an AI agent writes the summary note. Decisions taken in committee are recorded and carried over to the styles.
- How do I monitor competing products?
- Automated monitoring picks up prices, new products and product pages published on the sites the team has chosen to follow, and an AI agent matches them to the styles in the range. The product manager receives notable changes rather than an exhaustive report. Monitoring respects the terms of use of the sites followed.
- Is a PLM needed to automate the product manager’s work?
- Not necessarily. Much of the work concerns sales, prices and margins, which live in the ERP and the sales channels. The PLM becomes useful as soon as tracking of developments and costs has to be shared. Without a PLM, a custom tool can hold the collection plan and connect it to the rest.
02 — Getting started
Describe what’s holding your teams back.
A first conversation to qualify your need. Then a scoped diagnostic: the scope, the solution, the estimate.
Discuss your projectReply within 48 business hours.