Commercial exposure$27,222estimated set sales at risk
Prepared action64 vests → 9 storestransfer and approval email ready
Named ownerAlex · Allocationawaiting human approval
One operating layer across your retail company
Retail problems hide in fragments.
Outturn connects the evidence and gets action moving.
Outturn continuously connects fragmented retail data, reports, messages and operating knowledge. It detects problems in combinations no team can monitor at scale, consolidates cited evidence, applies company rules, prepares executable actions for human review and approval, and measures what happened.
- Across your existing systems
- Human-approved actions
- Company memory that compounds
Built for the retailer, not one vendor
One company decision. Many systems behind it.
Planning can recommend. BI can report. ERP can record execution. Outturn connects the issue, cited context, options, owner, approval, action and measured result across them.
| Existing system | What it already does | Outturn · the decision layer |
|---|---|---|
| Forecasting / planning | Recommends what should happen. | Records what the team decides when reality changes the plan. |
| BI / reporting | Shows what happened. | Carries the issue into an accountable decision and action. |
| ERP / operations | Records the approved transaction. | Connects execution back to its rationale and commercial result. |
| Workflow tools | Track whether a task was completed. | Track whether the commercial decision worked. |
Before the meeting starts · Sample workflow run
The issue arrives with its context, options, owner and deadline prepared.
The trigger can come from BI, a planning recommendation or a bounded workflow check. Outturn reconciles the evidence across systems, applies approved company policy and prepares the work; a named person still owns the commercial decision.
Store stockout response · campaign window
Monday · 08:40
Overnight workflow run
14issues checked
9within approved policy
3prepared for review
2need a named decision
Illustrative product data. No decision or external action was taken automatically.
Selected work · RD-0051Linen suit set · 9 campaign stores
Decision due today · 16:00
- Issue ingestedPower BI stockout alert preserved as the trigger.
- Retail data reconciledStore stock, online stock and matching-item sales checked.
- Messy context connectedA delayed-PO email, six store messages and Friday's campaign brief matched to the same issue.
- Company policy appliedTwo options calculated under KRL-04 · keep at least 65% online.
- Action package preparedTransfer plan, approval email and ERP upload file ready. Waiting for Alex · Allocation.
Product proof · One Decision Record from trigger to result
One issue.
One accountable decision.
Follow one obvious commercial problem that no system can see alone. Outturn connects structured retail data with a vendor email, store messages, a campaign brief and company policy—then carries one cited record through approval, execution and measured result.
1 · Detect
Several harmless fragments become one urgent problem.
2 · Reconcile
Retail data and messy context arrive together, cited.
3 · Scenario
Clear options. One human decision.
4 · Route
The approved action reaches the right person.
5 · Measure
The commercial result is measured.
RD-0051
Store stockout response
Sample data
Demo-seeded figures, not customer data. Every amount is recomputable from the shown inputs.
The problem
Retail data exists. Decisions still leak between visibility and result.
An issue appears in one system. The context lives in several others. Ownership is settled in messages, execution is tracked elsewhere and the outcome is rarely tied back to the decision. That leakage delays action and erases learning.
Sample stockout window · RD-0051
- Monday · 08:30Fragments connectedStockout + delayed PO + Friday campaign
- Monday · 16:00Approval deadlineNamed decision required
- Before weekendTransfer must leaveCommercial window closes
The cost is measurable
6.5%
Inventory distortion — out-of-stocks plus overstocks — costs retail $1.77T a year, equivalent to 6.5% of global retail sales.
The manual drag is visible
40%
of merchant time goes to low-value work, including consolidating systems, repetitive spreadsheets and reconciling data across silos.
Published evidence sizes the exposure. The four-week pilot measures your share of it.
Company Decision Memory
Every system records its own output. Outturn records what the company decided—and whether it worked.
Each issue becomes a Decision Record. The record keeps what was known, which options were considered, who approved the action, what happened next and what the company should remember when a similar case returns.
Existing systems
PlanningRecommendation BIAlert ERPTransaction MessagesLocal contextRD-0051 · Decision Record
One issue. Every accountable step.
- Trigger
- Cited context
- Options
- Owner
- Approval
- Action
- Outcome
Company Decision Memory
What was known Why this option won Whether it worked What to reuse next timeThe model does not become the company's brain. The company's own decisions become structured, traceable and reusable memory.
How it works
How Outturn becomes your decision system.
Outturn starts from a built-in Retail Workflow Pack, connects the governed sources required by that workflow, runs deterministic lifecycle checks, and stores the measured outcome on the same Decision Record.
Decision memory
Every decision becomes company memory.
DecisionRationaleOwnerActionOutcomeLearning
Decision loop
From signal to accountable commercial action.
DetectReconcileScenarioRouteMeasure
Always-on commercial intelligence
A shared source of decision context.
ReconcileMonitorPrioritiseDeterministic lifecycle checks
Connected systems
Where your work already happens.
SAP
Teams
Outlook
Gmail
PBI
Snowflake
Excel
ERP
Planning
Commercial context
Built around how your business defines success.
Company goalsFull-price sell-through
KPI definitionsMargin and availability
Department prioritiesPlanning and trading
User objectivesMerchant targets
GuardrailsBusiness rules
1
Detect / ingest
A connected source or a bounded deterministic check finds a critical signal and creates the record. Separate monitors keep the problem moving: a named owner accepts it, approval happens before the deadline, the approved action starts and the result is measured.
2
Reconcile
The context that lives in six places lands on one card: stock cover, sell-through against plan, open orders, supplier lead times - and the vendor email that quietly moved the truck. Every signal cited to its source row or document. Nothing asserted without a receipt.
3
Scenario
Options prepared before anyone asks, from your own playbook: open a transfer scenario, request a buyer review, prepare a markdown review, accept the risk and monitor. Each option carries its arithmetic - never a black-box score.
4
Route
Decisions get owners, not threads. The buyer approves reorders; allocation runs the transfer; store ops confirms the floor. Routed with the full context attached, tracked until the action actually happens - in your systems, with a named approval.
5
Measure
What happened next? Observed sales, availability and execution timing are measured on your numbers and written back with the agreed attribution method. Delayed and missed actions surface instead of vanishing.
Inside the product
Three surfaces. One decision lifecycle.
The Fire Board shows every issue through its lifecycle. The Decision Record holds the accountable work. The Copilot explains one record from cited evidence and helps route what a person approves.
01 · Retail Fire Board
See the work already prepared.
Within policy 9
Ready for decision 3
Waiting for owner 2
The queue is an output of the work—not another inbox for the team to populate.
02 · Decision Record
One issue. Every fact and decision.
RD-0051 · Decision Record
TriggerCited contextOptionsOwnerActionOutcome
Outturn creates the traceable record as your team reviews, approves and acts—without a separate documentation ritual.
03 · Decision Copilot
Ask one record. Get a cited answer.
Why this transfer?
It satisfies the approved online floor and covers all 9 stores. Show sources →
It explains and routes. It never detects or invents a number.
Illustrative product surfaces; all figures are demo-seeded.
Start with one governed workflow
A Retail Workflow Pack starts the work. Your existing sources supply the facts.
The stockout-response pack arrives with the decision definition, required signals, common options, owners and outcome method. Reports, exports and messages provide the evidence; your team only confirms what is genuinely company-specific.
BI report
Excel or CSV
Planning export
ERP export
Messages + store feedback
Decision policy
Built-in retail intelligence
It arrives knowing the workflow. You confirm what makes your company different.
Every stockout-response pack starts with cited metric definitions, required signals, common actions, ownership patterns and outcome methods. Outturn calibrates the defaults on your data; your team only corrects the parts that truly differ.
✎
Industry definition
The pack begins with a cited definition of the decision, the minimum context required and the commercial outcome to measure.
✎
Segment default
Apparel, grocery and specialty retail do not use identical signals or deadlines. The pack starts from the relevant segment pattern, not a generic prompt.
✎
Deterministic calibration
Your history calibrates thresholds and deadlines in code. The record shows the definition, formula and evidence used - never an unexplained model score.
✎
Retailer correction
If your sell-through includes outlets, Texas transfers use Dallas first, or reorders over $50K need two approvals, your correction becomes a versioned company policy after review.
proposed policy
Policy proposal: increase minimum donor coverage from 82% to 88% - based on two negatively affected donor stores, with replay, trade-off and affected scope shown before approval.
A policy proposal must explain why: the Decision Records that triggered it, how those cases replay under the new rule, the expected benefit and downside, where it would apply and what the evidence cannot yet prove. Nothing changes until a human approves it.
Honest AI
AI models are not reliable reasoners. They are excellent accelerators and researchers.
In Outturn, AI researches. Code calculates. People decide. Models read messy sources and draft bounded options; deterministic services run the checks and money; a named person owns every decision and approves every write.
Bad or conflicting inputs hold the decision for review. See the architecture and worked examples →
Measured value
Measure the process. Then measure the money.
Estimated exposure, operating improvement and attributed impact stay separate.
$27,222
Estimated set sales at risk, using the retailer's stated 25% set-purchase assumption.
4 measures
Decision lead time, on-time completion, missed deadlines and records measured.
Evidence first
Recovered or protected value is claimed only when an agreed comparison supports it.
The pilot establishes the baseline: how fast decisions move, whether actions complete and what commercial evidence can honestly be attributed.
Our story
We built Outturn
from inside the problem.
We did not choose retail from a market map. We worked inside the decisions Outturn now prepares.
Outturn began with a pattern we saw repeatedly while working directly with retail teams: the data existed, but the decision still broke apart across systems and conversations.
Planning held the recommendation. BI held the alert. Stock, sales, supplier context and company policy lived elsewhere. By the time the facts, owner and action came together, the commercial window had often passed—and the result was rarely tied back to the decision.
Our founding team combines direct experience with retail operating teams and experience building production AI systems for retail. We understand both the workflow and the trust boundary: what can be researched automatically, what must be calculated deterministically and where a person must decide.
That is why we built Outturn: to give the retailer its own active decision layer across the systems it already uses—and a memory of what the company decided and whether it worked.
Co-founder
Elif Erdem
Elif has worked directly with retail teams on inventory, planning and commercial decisions. At Outturn, she brings that operating context into each workflow—from the evidence a merchant needs to the people who must own the action. Her background combines engineering, entrepreneurship and leadership.
Why this team
Retail operating context.
Production AI experience.
We know the customer, the workflow and what it takes to make AI useful inside an accountable commercial process.
Four-week decision-leakage pilot
Prove where commercial decisions stall—and close the loop.
Use one stockout-response workflow to baseline the path from visible issue to approved action and measured outcome. Start from existing reports and exports; no production integration required.
- Week 1 · Baseline the leakage — map the issue, sources, owners, handoffs and commercial deadlines in one workflow
- Week 2 · Run live records — reconcile cited facts, prepare clear options and route named decisions
- Week 3 · Operate independently — your team chooses and runs the work without us in the daily loop
- Week 4 · Measure the result — compare decision lead time, action completion and agreed commercial outcomes
You finish with a measured decision-leakage baseline, live Decision Records and a clear decision on whether to expand—not an invented ROI promise.
Request received.
If it's a fit, you'll hear from a real person within a couple of working days.
FAQ
Questions retail teams ask before getting started.
What does Outturn actually do?
Outturn is the active decision layer across the retailer's existing systems. It turns an issue into one cited Decision Record, prepares options, routes the chosen action after named approval, measures the result and preserves what the company learned.
Is Outturn autonomous?
No. Outturn works continuously inside a bounded workflow, but it does not choose the commercial action. A named person decides, and your team releases any downstream write.
Does Outturn replace our existing systems?
No. Planning, allocation, ERP and BI remain in place. Outturn connects their outputs around the decision: what changed, what the company knew, what it chose, who owned the action and whether it worked.
Do we have to enter everything into Outturn?
No duplicate entry for the investigation. Outturn starts from the reports, exports, messages, documents and approved policy already used by the workflow. As your team reviews and acts, the evidence, decision, owner and outcome are captured automatically as company memory.
Is Outturn a forecasting tool?
No. Forecasts can be an input. Outturn records what the team decides when reality changes the plan.
How can we trust the numbers?
Every commercial figure is calculated in code and opens to its sources, assumptions and formula.
What does Outturn need to get started?
The reports and exports already used for one workflow. A built-in Retail Workflow Pack provides the starting definitions, checks, options and owners.
How does the pilot work?
Four weeks, one workflow and a scope agreed together. We baseline where decisions currently stall, run live records, hand daily operation to your team and measure lead time, completion and agreed commercial outcomes.
Where does our data go?
Before anything is connected, we agree the pilot's access, hosting, retention and deletion controls with your security team. The pilot starts read-only; no external write occurs without named approval. Read our security and data approach →
What does it cost?
Pilot scope and pricing depend on the decision workflow, data, and team involved. Request a conversation and we’ll discuss what a useful four-week pilot would look like for your business, along with what comes next if it proves valuable.