General AI assistant
Germany 41%, Sweden 44%, UK 39%.
Books returns when they land and leaves out shipping and payment fees, so the latest month looks better than it is.
The AI that already knows your numbers.
Ask Dema anything, like you would any AI. The difference is that it already has all your commerce data, connected and checked.
€412K in September, up from €351K in August. Every number comes from a connected source:
| September | August | Change | |
|---|---|---|---|
| Net sales | €1.48MShopify orders | €1.39M | +6% |
| Product cost | −€596KCost per SKU, NetSuite | −€570K | +5% |
| Returns | −€188KShopify returns | −€201K | −6% |
| Marketing | −€284KMeta, Google, TikTok | −€268K | +6% |
Used by the teams at
Side by side
AI is only as good as the data under it. Ours is all of it.
The same question to a general AI assistant and to Dema. The model is rarely the difference. What it knows about your business is.
What was contribution margin by market last month? Dema: Germany 31%, Sweden 35%, UK 28%.
What was contribution margin by market last month?
Dema
Germany 31%, Sweden 35%, UK 28%.Germany 0%, Sweden 0%, UK 0%.
Contribution margin after expected returns, fulfilment, payment fees and marketing.
- Shopify
- Returns
- Fulfilment costs
- Google Ads
- Meta Ads
CM3 = net sales − COGS − fulfilment − fees − expected returns − marketing
Illustrative answers on sample data.
What's underneath
What Dema already knows before you ask.
The chat is only the part you see. Behind it is everything your team would otherwise have to build and keep correct themselves.
- 01
All your sources are connected
Orders, ads, returns, stock, costs and customers stay in sync, so nobody has to export or upload files.
View integrations - 02
Every metric is defined once
Net sales, ad spend and contribution margin mean the same thing in every answer, app and agent.
Explore the data layer - 03
Margin includes every cost
Contribution margin goes all the way to CM3, and expected returns are counted before they arrive.
Ask about margin - 04
Marketing measurement is built in
Marketing mix modeling, incrementality tests and causal attribution all run on the same numbers.
Explore measurement - 05
Answers can turn into actions
Turn an answer into a shared app or a scheduled agent. Nothing changes until someone approves it.
Explore agents
From months to one command
“What took two people three to four months now runs in a single command. We built 20 workflows in three weeks without writing a line of code.”
Tom Everson
Digital Director, Tatti Lashes
34 hours
Saved per person, every month
People using Dema save around 34 hours a month, about 8 hours every week, from finding the answer to acting on it.
Self-reported survey of Dema users, 2026. Averages use the midpoint of each answer range.
+40%
Profit growth
Brands using Dema agents grew profit by 40% year over year, on average.
The honest call
When a general AI assistant is enough.
Not every question needs a data platform behind it. Here is how to tell which one you need.
Choose Dema if…
- You need answers on profit, not just revenue, across markets and channels.
- Several teams need to work from the same numbers.
- You want recurring work, like the weekly report or budget moves, to run on its own.
- Actions should happen inside approval limits you set.
A general AI assistant is enough if…
- You have a one-off question about a small export.
- You need help writing, summarising or brainstorming.
- Nobody else needs to rely on the answer next week.
Questions
What teams ask about AI and their data.
Partly, yes. You can chat, prompt and ask follow-ups like with any AI assistant. What's different is that Dema already has your connected commerce data and your metric definitions, and it can turn an answer into an app or a scheduled agent.
Yes. They are excellent for writing, summarising and one-off questions. Use Dema when the answer has to be right on your commercial data, shared across teams, or acted on.
Keep it for finance, BI and reconciliation. Dema is the layer for deciding and acting: measurement, profitability, forecasting and agents on top of your commercial data.
You can get a first answer quickly. Getting the same right answer every week takes a semantic layer, profit logic that survives returns, measurement, permissions and approval gates. That is the part Dema already runs.
Missing sources and unclear definitions are flagged instead of filled in. Dema tells you what it would need rather than inventing a confident number.
Inference and pipelines run in the region you choose, EU or US, and Dema is GDPR compliant.
See it on your data