Dema

Monday trade.Ready by 9:30.

Represent

Represent's trade report used to take three teams most of Monday. Now an agent builds it from Dema and posts it on Teams at 9:30, so the meeting is about what to do next.

2,080

Hours saved a year, 40 every week

9:30

Monday trade report, down from 4 pm

3 in 1

MMM, customer and attribution platforms

Industry
Streetwear & luxury
Team
Around 170, across three offices
Commerce
Four Shopify stores, NetSuite, Airtable
Uses Dema for
Trade reporting, apps, measurement
“Dema is a tool that's fundamentally changed how we work and operate as a team and as a wider business.”

Monday,before.

Every Monday, merchandising, e-commerce, marketing and retail meet to review how the week traded. For years, getting there took most of the day. Numbers were downloaded from each of Represent's four Shopify stores, rolled up in Excel, typed into PowerPoint and emailed out. The meeting couldn't start until 4 pm.

That left almost no time for the questions that matter: why a market dipped, what a launch really sold, and what to do about it.

Before

Report ready as the meeting starts16:00

Pulling numbers into Excel and PowerPoint

With Dema

Report on Teams09:30

Commentary, deep dives and actions
  • Building the report
  • Time to discuss it and act
“My team of eight used to spend all of Monday pulling the numbers together. Now we spend it on the actions we need to take.”
Janine HamlettMerchandising Lead, Represent

One data model.One agent.

Represent connected its four Shopify stores, NetSuite and the Airtable launch calendar to Dema. Dema models it all into one commerce data model, with every metric defined once, and puts measurement on top: marketing mix modelling, attribution and incrementality testing in one place, where Represent used to run three separate platforms.

On that foundation, Mark Brookes, Head of Business Solutions, turned the trade report format designed by Janine Hamlett's merchandising team into a single agent. The same agent now builds the Monday trade report, Ricky Jennings' daily e-commerce dashboard and a tailored pack for the team in Los Angeles, all from the same numbers.

Represent's team at their desks, with spreadsheets open on their monitors
Represent's team at work.

Trade meetings,about what to do next.

By 9:30 on Monday the report is on Teams. The numbers are done, and the agent drafts the commentary on why things are up or down. The team adds the context only it has, then a section on what it will do about it.

So the meeting skips the recap. It covers the week, the launches and the stores, and leaves with actions, down to a sold-out product that is still getting ad spend.

Trade · Weekly

DemaMon 09:30

Weekly trade report

Net sales
£1.2M
+8% vs LY
Launch
Sold out
Night one
Stores
£184K
−3% vs LW

Draft commentary: online is up on Wednesday's launch, stores are softer after a quiet weekend. Add your context before the meeting.

  • Pull Meta spend from a sold-out styleHigh views, low conversion, no stock
  • Size the sales missed on the launchOpen the launch sell-out app
Illustration. Figures are examples, not Represent's results.

Apps,in a few prompts.

Represent launches most Wednesdays, often with first-night sell-outs. To size the sales missed, Janine asked Dema's agent to compare launch night with the natural drop-off over the following days, by business unit. Three or four prompts later, she asked it to turn the analysis into an app. Now every new launch is added automatically.

Because every app runs on the same data model, Joe Seabridge's paid media team has built its own, for channel performance, customers and incrementality tests. Each is designed for the people who use it, while access to building stays controlled.

app.dema.ai/trade-explorerDemo workspace
Example prompt
Week 41 vs week 40 · Updated Monday 07:00

Weekly trading

All markets
Net sales
€252K
+4% vs W40
CM3 contribution
€63K
25% of net sales
Gross margin
57%
+1.4 pts
Ad spend
€41K
16% of net sales
Stock cover
8.5 wk
3 styles under 2 wk

Executive summary

Trading report agent · Mon 07:00

A strong week led by Sweden and Norway, held back by stock-outs in the UK.

  1. 1
    Knitwear carried the week

    Cable Knit sold 1,240 units, up 22%, and is the best-selling style overall.

  2. 2
    The UK lost sales to stock-outs

    Two outerwear styles sold out on Wednesday. The UK fell 2% while every other market grew.

  3. 3
    Paid spend needs a reset

    Meta reports a 4.8 ROAS but measures 2.4, and €2,400 a week still goes to sold-out styles.

What moved net sales

€242K → €252K
  • Sweden+6K
  • Germany+2K
  • UK−1K
  • Norway+3K

Markets

Net sales · CM3
Sweden€96K+7%28%
Germany€64K+3%25%
UK€52K−2%19%
Norway€40K+8%25%

Risks and opportunities

  • Risk

    Cable Knit is down to 1.8 weeks of cover across markets.

  • Risk

    19% of Leather Jacket units are returned, above the 15% threshold.

  • Opportunity

    Email measures a 4.1 return, the highest of any channel.

  • Watch

    Beanie, Green is down 14% with 18 weeks of cover.

SourcesShopifyNetSuiteMetaGoogle AdsKlaviyoMarketing mix model

0 items selected for the meeting agenda.

Example app built from a prompt. Figures are illustrative, not Represent's results.

How Represent runs on Dema.

Represent connects its commerce data to Dema's measurement and intelligence layer. Everything the teams use, from the Monday report to launch apps, is built on top of it.

Represent's data

  • Four Shopify stores
  • NetSuite, through MCP
  • Airtable launch calendar
  • Ad platforms
DemaMeasurement and intelligence layer
  1. Data model

    Orders, products, customers and costs, with every metric defined once.

  2. Measurement

    Marketing mix modelling, attribution and incrementality tests in one place.

  3. Intelligence

    Agents that read the numbers, explain them and draft the next step.

What the teams use

  • Monday trade report on Teams
  • Daily e-commerce dashboard
  • Trade pack for Los Angeles
  • Launch sell-out app
  • Paid media apps

One source of truth.

Store, analytics and attribution numbers no longer disagree. Every team and market works from the same figures.

Answers in minutes.

Ad hoc questions that took days or weeks now take a prompt, and the answer is the same every time.

A data team's reach.

With no data engineers or data scientists, a small systems team now takes on questions like who Represent's customer really is.

Automate your Monday reporting.

Bring your current trade pack. See how an agent could build it from your live data and post it to the team before the meeting.