Product feed management
Turn metrics into action with rule-based segments
Create product and customer sets from any combination of metrics and dimensions. Dema groups, labels, and syncs them to your ad platforms and feeds automatically.
Wool Overshirt
Margin 52% · Sell-through 24%
Canvas Tote
Margin 38% · Sell-through 61%
Rib Knit Beanie
Margin 58% · Sell-through 19%
Linen Shirt
Margin 44% · Sell-through 47%
Suede Loafer
Margin 49% · Sell-through 27%
Google Shopping
custom_label_2
Meta Catalog
custom_label_2
TikTok
custom_label_2
Shopify
Product tag
Trusted by commercial teams at leading brands
The problem
Ad platforms optimize on your feed. Your feed doesn't know your numbers.
Meta and Google scale whatever the feed tells them to. When the feed carries no margin, stock, or sell-through data, they scale the wrong products — and nobody notices until the P&L does.
In the ad platform
ROAS 4.6
In your P&L
CM3 −8%
Revenue looks great. Profit isn't.
40 products · set manually
14 of 40 are now out of stock
Labels go stale in a week
Slow movers up +18% this week
feed_export_v3_final.csv · edited by hand
Insight and feed live in different tools
Product sets
Group products by any metric, automatically
Combine margin, impressions, sell-through, stock levels, and 100+ other metrics into rules. Products matching all rules are grouped into a set, updated in real time as your data changes. Push sets to feeds or use them to filter reports.
Feed-based profitability
Segment by actual margin, CM3, or GP, not just revenue. Rules built on real P&L data from connected feeds.
Forecasted sell-through
Use predicted sell-through and days of supply to catch slow movers before markdowns. Segments adapt as forecasts update.
Segment from reports
Spot a pattern in any Dema report, turn it into a segment in one click. No switching tools, no rebuilding logic.
01
Segment products by margin, sell-through, and stock
Build rules on the numbers the feed never carries — contribution margin, predicted sell-through, days in stock. High-margin slow movers, full-price heroes, styles about to break cover: each becomes a living set that updates as the data does.
Product segment rules
Define conditions to group products. Matching products update as data changes.
Conditions
3.7% of catalog
02
Every platform gets its own field
Google Shopping and Meta Catalog receive product sets as custom_label_0–4, Shopify gets product tags, and CSV over SFTP covers everything else. Labels refresh automatically whenever a rule's matches change — no exports, no stale feeds.
03
Segment customers the same way
Use LTV, order frequency, return rate, days since last purchase, and acquisition source to build customer segments. Export lists for targeting, trigger win-back campaigns, or filter dashboards by cohort.
Customer segment rules
Define conditions to build a customer segment. Rules are combined with AND/OR logic.
Conditions
14.0% of total base
Build segments in conversation
Ask the agent which products have high margin but low sell-through. It builds the segment rules, previews the match count, and saves the set, all from a single prompt.
You asked
Which products have margin above 45% but sell-through below 30%?
Agent created segment
312 products matched out of 8,430
Say 'narrow it to products with more than 60 days in stock' and the agent adds the rule, updates the preview, and shows how the match count changes.
You asked
Which products have margin above 45% but sell-through below 30%?
Agent created segment
312 products matched out of 8,430
Clarity and control
“Dema has been a game-changer for our brand. It's given us the clarity and control we needed to scale profitably, ensuring we make the right decisions across inventory, marketing, and product strategy.”
Trusted by category leaders
Real teams, real numbers — from cutting acquisition costs to doubling profit to automating whole planning cycles.
“Dema has been a game-changer for our brand. It's given us the clarity and control we needed to scale profitably, ensuring we make the right decisions across inventory, marketing, and product strategy.”
Doubled sales and gross profit in the same period year over year after moving decision logic onto Dema.
Jorge de Castro
COO
2x Sales growth
Enterprise-ready
Built for production.
Where your data is processed, how it is protected, and what an agent is allowed to change are controls you set — not defaults you inherit.
EU & US inference
Choose where your data is processed. Model inference and data pipelines run on European or US infrastructure with full residency control.
Never used to train models
Your data is never used to train models. Processing stays in your chosen region and runs in isolated, encrypted sessions.
Always the latest models
New reasoning models roll out as they ship, so your agents keep improving automatically without migration work.
Permissions and approvals
Each agent has its own scope and integrations. Actions can require a named approval before anything changes.
What you can create
Real setups, linked to how they run.
Every example below is a real prompt from the library. Open the use case behind it to see the full run — what it checks, what comes back, and where it lands.
Questions
Frequently asked questions
Any metric or dimension available in Dema, including margin, impressions, clicks, sell-through, stock levels, order count, LTV, and more. Rules combine metrics with operators (greater than, less than, between, top N) to match products or customers.
Each product set can be mapped to a custom_label field (0–4) in your Google Shopping or Meta Catalog feed, or applied as a product tag directly in Shopify. When a product matches your rules, the label is applied automatically. Changes sync daily or on-demand.
Yes. Customer sets use the same rule-based approach but with customer-level metrics: purchase frequency, LTV, return rate, acquisition source, and cohort data. Export customer lists for targeting or use them to filter dashboards.
Yes. As new data flows in, products and customers are re-evaluated against your rules. A product that drops below a margin threshold exits the set automatically. No manual cleanup.
Bring one commercial job