Nicola Sirchia

What I built, and what it did for brands and sales

Eight pieces of work from a career across consumer goods, technology hardware, consumer health and performance advertising, for consumer and industrial customers. One idea sits behind all of them: brand, e-commerce, product content, data and AI work best as one growth engine.

Selling online

Building the channel and making every euro of media work.

E-commerce built from zero, in three steps

An online sales channel built from nothing, in three steps. First the largest marketplace in the main markets. Then online-only retailers and other marketplaces, both generalists and specialists. Then online and offline as one experience at selected retailers.

€0 to €50Monline sales, built from zero

Read the full case

For the brand
An assortment designed for each channel, with multipacks and bundles, so the brand showed up the right way wherever people shop.
For sales
From zero to about €50M in online sales. Pricing governance and separate online and offline terms let the online channel grow without hurting the business in stores.
How technology helped
A central team led and local key account managers executed, all working from the same product data.
  1. 1The largest marketplaceThe main markets first
  2. 2Online retailers and marketplacesGeneralists and specialists
  3. 3Online and offline as oneOne experience at selected retailers

One retail media playbook for every channel

A playbook in five steps that feed each other: content, brand, advertise, promote, connect.

For the brand
Search-ready titles, images, video and keywords, plus enhanced content and brand stores. The product page does the job a good salesperson does in a shop.
For sales
Sponsored product and brand ads, cross-sell and upsell, deal events and hero campaigns, with social traffic sent through a landing page to the retailer.
How technology helped
The centre sets the strategy and the guidance on pricing and trade terms. Local e-commerce managers run it, and everyone is measured on the same numbers: conversion, basket size, repeat purchase, acquisition cost and availability.
1Content2Brand3Advertise4Promote5ConnectTheflywheel
Each step feeds the next, and the fifth feeds the first again.

Content and platforms

One source of product truth, written once and published everywhere.

One product-content ecosystem, from data to shelf

One source of product truth for attributes, technical data and claims, connected to an AI content factory and an asset library, and feeding every channel.

For the brand
Complete, consistent content for the whole range on every brand website, on marketplaces, at retailers and in the AI advisor. This is what makes a brand readable for AI assistants.
For sales
A real saving on content production, and products that reach every channel ready to sell.
How technology helped
One product information system feeds the AI content factory and the asset library, so the same checked data reaches every channel. The AI also finds missing attributes and harmonises the data.

An AI content factory, with people in charge

A content factory that writes product copy, search and AI-ready texts in every language from one source of product data. It follows the rhythm of a campaign process I used for years: brief, plan, draft, create, approve, go live. The AI took over the drafting. People kept the brief and the approval.

For the brand
The same tone and the same checked claims in every language and channel, and product pages that both people and AI assistants can read.
For sales
People spend their time on the brief and the approval instead of routine drafting, and products reach marketplaces and retailers ready to sell.
How technology helped
Connected to the product information system, so every text starts from checked data. Corporate and local marketers approve every text before it goes out, in line with the EU AI Act and the rules on environmental claims.
  1. 1
    Brief

    Sales and marketing agree the goal, the audience and the budget.

    PeopleWeek 1
  2. 2
    Plan

    Channels, journeys and the numbers that will judge it.

    PeopleWeek 1
  3. 3
    Draft

    Copy for every channel and language, written from checked product data.

    AI drafts, people checkWeeks 2 and 3
  4. 4
    Create

    Pages, emails, ads and assets, with the variants each market needs.

    AI and peopleWeek 3
  5. 5
    Approve

    Brand, legal and the local markets sign off every text.

    PeopleWeek 4
  6. 6
    Go live and learn

    Results go back into the next brief.

    AI and peopleNext month

People decideAI drafts, people checkAI and people together

The six steps stayed. What changed is who does the heavy lifting in the middle, and the approval step that no text skips.

31 brand websites on one platform

A shared design system for many brands and languages, accessible to WCAG 2.0 AA, with shared content templates and one approval workflow.

31brand websites on one platform
For the brand
The same standard in four regions: Europe, the Americas, Asia-Pacific and the Middle East.
For sales
All 31 websites moved to one platform, with one standard for every brand and language.
How technology helped
It was built once at the centre and then adopted. Local teams were trained on the content system, paid media and e-commerce, each market has a clear owner, and adoption is tracked.

Customers and data

Knowing the customer, and turning that into sales.

Lead generation in consumer health, built from zero

A complete lead generation process where none existed before. It joined the website, the customer database and marketing automation into one flow that starts with a free online hearing test and ends with a sale at a partner clinic.

€150Mpotential revenue generated in the sales pipeline

Read the full case

For the brand
The brand helps first, with a free test and useful information, before it asks for anything. Every message after that is personal, based on what the person has told us.
For sales
About €150M of potential revenue generated in the sales pipeline, with a strong conversion further down, from marketing qualified to sales qualified leads. Low-interest leads are filtered out early, so clinics receive people who are ready to talk, with all their information in one view. Marketing and sales worked from one shared funnel for the first time: awareness, involvement, trial, sale. For the clinics, the leads became a service on top of the product, and a reason to stay with the brand.
How technology helped
The online test and web forms create the lead, with consent. The customer database assigns it to a campaign, marketing automation runs the journey, and a clinic finder lets the person choose where to go. Dashboards followed every lead from marketing qualified to sales qualified to revenue, for consumers and for business customers. The same design then took on new sources: a contact form, the clinic finder and a quiz.
  1. 1CaptureAn online test or a form creates the lead, with consent
  2. 2QualifyAn email journey and a short quiz filter low interest early
  3. 3ConnectThe person picks the nearest partner clinic
  4. 4ConvertInside sales follows up to the purchase
  5. 5MeasureDashboards from qualified lead to revenue

From customer signal to live campaign

The business requirements and, together with group IT, the architecture for a customer data platform in six steps: sources, profiles, insights, consent, audiences, activation.

For the brand
One preference centre, so every audience can be used legally in every market, and known customers stop seeing ads meant for strangers.
For sales
Email journeys reach the people who asked for them, and paid media is spent on new customers only.
How technology helped
Where-to-buy clicks, portal activity, AI advisor chats and paid search are matched into one profile per customer. Segments keep themselves up to date, with no lists prepared by hand.
  1. 1SourcesClicks, portal, chats, search
  2. 2ProfilesOne per customer
  3. 3InsightsInterests and demand
  4. 4ConsentOne preference centre
  5. 5AudiencesSegments that update themselves
  6. 6ActivationJourneys out, known customers excluded from ads

AI that helps customers

An assistant people can trust with a real job.

An AI product advisor customers can trust

I still remember the moment this started. I was checking a shelf in a shop, and next to me a man kept walking left and right along it, picking up one product, putting it back, picking up another. He clearly had a job to do at home. He just had no idea which product would do it.

The shelf had all the answers, but none of them was talking to him.

A chatbot is quick to build. An advisor people can trust with a real repair is a very different story. Nobody asks for a product by its technical name. People describe a wobbly chair leg, exactly as the man at that shelf would have.

Together with my team, an AI advisor on our brand websites and on the phone line. You describe the job, show it a photo of what is broken or simply talk to it. It tells you which product to use, how to apply it and where to buy it.

Read the full case

For the brand
Advice at the shelf, at home, or on the phone at eleven at night and at weekends, in the local language. Customers stay loyal to a brand for as long as it keeps solving their problems.
For sales
Each person is sent to the right product for the job, and to a place to buy it. It also looks at the background of the photo, to spot other jobs around the house it can help with.
How technology helped
The parts I am proudest of are the ones you do not see in a demo, and the hardest to get right. Every answer passes safety guardrails: when a question involves safety equipment, the advice is to replace the item, never to repair it. When a case needs a person it hands over to the technical advisers, and outside working hours it escalates. Every conversation can be rated, so we see where it helps and where it does not yet. All of it is connected to the rest of the digital ecosystem.
  1. 1AskType, show a photo or talk
  2. 2AdviseWhich product, how to apply it, where to buy it
  3. 3ProtectSafety guardrails on every answer
  4. 4Hand overTo a technical adviser when a person is needed
  5. 5LearnEvery conversation can be rated

The team

The people who made all of it work.

The team behind the work

Around 30 people across 15 markets and four regions. The team was named best performing team in its company three times.

None of the eight pieces above works without them. I set the direction, the team answers with what it would take and what must stop, and we go back and forth until both sides own the same plan.