ASF — Workshop Output

Full summary across all sessions  ·  Story & Strategy Workshop  ·  18 August 2026

01 — Starting Point
The product works.
The commercial model doesn't exist yet.
ASF measures airline food waste. It audits flights, captures what was loaded versus what was wasted, and produces data, insights, and a dashboard that lets airlines and caterers see their waste profile for the first time. The product creates real value. The problem is it captures none of it.
Data
49 audited flights. More than anyone else. Not statistically robust, but enough to start conversations with senior buyers.
App
Built on Microsoft Power Apps. Functional. Never tested in the wild. Going live for the first time at SATS in September.
Framework
Audit methodology, accreditation, insight meetings. Being refined. Clients get a data report + a second meeting with recommendations.
What's Missing
A repeatable commercial equation. Proven revenue model. Validated AI product. Enough data to make intelligence claims.
02 — The Equation
Matt's model. Drawn live.
Works inside-out.
10% fewer meal loadings
× $X cost per sealed meal
= Margin saved on wasted meals
Gate question: Is saving > cost of ASF system?
Yes → caterer commits → passes cost through + ASF dashboard fee → airline sees same 10% proof
Layer 2 — Commercial Sell
Caterers are the gatekeeper. They control the data, run the ops, and hold the margin on wasted meals. Sell to caterers first. Once they commit, the airline layer activates on top.

Airline sees: dashboard + 10% saving proof + sustainability metrics.
Caterer sees: margin recovery. Simple.
Layer 3 — Benefits
Caterer benefitMargin saved on meals
Airline benefitLoad reduction + CSR
ASF revenueDashboard fee on top
Proof caseSATS POC
The clubcard reference: Behaviour change driving measurable economic value. Loyalty → measurement → saving. Same logic: waste audit → measurement → saving.
03 — Honest Inventory
What the data actually is.
Not what we wish it was.
5
Predictive Intelligence
Future projections. "If passengers grow 20%, waste grows by X." Requires large, clean longitudinal dataset. Not possible yet.
Not reached
4
Intelligence
System proactively surfaces trends, anomalies, and recommendations. Requires machine learning on clean, large sample. Not yet.
Not reached
3
Insight with consulting
Data + a human layer of "here are the 3 categories to focus on." ASF does this today. Post-audit insight meeting. Dashboard visualisation. General, not yet actionable at scale.
← ASF today
2
Reporting
Raw data formatted and delivered. Already passed this stage.
Done
1
Data Collection
Audits, spreadsheets, app entry. The foundation.
Done
The app risk: Built on Microsoft Power Apps. Tested at desk. Failed in Canada in a previous test. Fixes applied. Goes live at SATS in September — simultaneously with the most important POC in the company's history. Running parallel spreadsheet as fallback.
September data additions: Sealed beverages being added to the data set for the first time. Loading data being captured. New methodology for economy class waste splits. Second-year audits at same locations allow year-on-year comparison.
04 — The Key Debate
Intelligence vs Confidence.
The sharpest disagreement of the day.
Matt's Position
The market uses the word "intelligence." Four senior caterers at Toulouse said they want AI intelligence. You need a door-opener. David's competitor line: "The operating system for airline catering intelligence."

If that's what buyers are asking for, you need to be in that conversation.
Diana's Position
Don't claim what you can't prove. What the airline buyer needs is 95% accuracy on counts so they can make the management decision to reduce loads. We're doing A minus B. But we do it so accurately and consistently that they can act on it.

"That's not intelligence. That's confidence. Own that word."
Resolution: Use the word if you must — but define it precisely. The first sell is "confidence to act." Intelligence is the aspiration once data volumes justify it. You can't claim intelligence with 49 flights. You can absolutely claim accuracy and confidence.
The unresolved word
Intelligence  ·  Confidence  ·  Accuracy  ·  Standardisation
One of these wraps the product. Decision still outstanding. Must be settled before the narrative can be written.
05 — Binary Gate
SATS is not a client.
It's a hypothesis test.
SATS is Singapore Airlines' catering company. They are pursuing an AI-driven food waste POC, alongside an unnamed AI partner under NDA who holds the technology licence. ASF's inclusion in that POC is not yet confirmed. Matt goes to Singapore in week one of September. Diana rates inclusion at 70% confidence. Everything downstream — commercial model, rebrand, Hamburg launch, roles — waits on a yes from that conversation.
If hypothesis holds
10% saving > ASF cost → caterer commits.

Commercial model locked. Brand ready to scale. Rebrand triggered. Hamburg WTCE April 2027 is the launch moment. Roles defined. Grant funding directed at team.
If hypothesis fails
Pivot before March 2027.

The commercial model equation still gets written by November — but with a different proof case. The brand does not scale until a new proof case is found.
06 — Who We Sell To
Two buyers. One gatekeeper.
Three trade bodies that matter.
Buyer Contact Motivation What They Need Priority
Caterer GM / COO / Procurement Margin recovery on wasted meals Saving > cost of ASF. Simple gate question. Primary buyer
Airline In-Flight / Catering Ops / Procurement / Sustainability Load reduction, cost savings, sustainability metrics 95% data accuracy. Confidence to make the management call. Secondary buyer
Airports Waste / Sustainability Waste management integration Not buyers. Can't handle international catering waste. Understand their agenda only. Network only
Trade Bodies — Priority Order
1 — ACA  Business Critical
Recently joined. The gateway to caterers at scale. Strong sustainability group. Webinar vehicle. ACA is the channel for Stage 3 sales. SATS's role in ACA credibility must be defined in the commercial proposal.
2 — IFSA
Active but lighter touch. Secondary to ACA. Worth maintaining presence. Fans — think ASF is great — but no direct action.
3 — Others (IATA, ACI)
IATA — ignore. ACI — airports, network awareness only. Light touch. Not revenue targets.
07 — Gaps
What was explicitly identified
as missing.
Missing Item Why It Matters Status
AI partner licence The unnamed partner under NDA holds the technology the SATS POC depends on. Without the licence, the POC doesn't run. Biggest assumption. Unconfirmed.
ASF inclusion in the SATS POC Binary. In or out. Decided by September 7. 70% confidence. Matt to confirm.
Commercial / financial model The written equation. "This + this = that." Not yet built. November milestone. In progress
New business narrative No positioning statement that works yet. Intelligence? Confidence? The word isn't settled. Not written
Data volume 49 flights. Not enough for intelligence claims any sophisticated buyer will accept. September adds more
Roles of the three founders If the licence is granted and the POC runs — what does each person do? Not agreed. Not defined
Singapore NFP office Not-for-profit entity in Singapore. Needs to close — not structured for commercial operation. Close within 30 days
08 — What the Plan Rests On
Five assumptions.
All of them matter.
Unknown
AI licence is granted The unnamed AI partner under NDA provides the technology layer for the SATS POC. The entire roadmap assumes this comes through. It's the single biggest unknown and it has no decision date yet.
70%
ASF is included in the SATS POC Diana's confidence level. Decision point: week of 7 September. Matt travels to Singapore. If the answer is no — stop and pivot.
Hypothesis
10% saving > ASF system cost The core commercial equation. Written in words by November. Tested with real numbers at SATS. If it fails, the proposition pivots before March 2027.
Untested
The app works in the wild First live test is September at SATS. The same event as the most important POC in ASF history. Spreadsheet runs in parallel as fallback.
Unknown
Grant funding can be directed at ASF The plan to fund team growth and structure partly depends on directing grant money toward ASF. Not confirmed. Needs to be raised in the Singapore conversation.
09 — What Happens Next
Sequenced actions.
Everything flows from September 7.
September — Week 1  ·  THE GATE
Matt goes to Singapore
Binary decision
One goal: confirm ASF is included in the SATS POC. Decision by 7 September. If yes → everything below runs. If no → stop and pivot.
September
September Audits + App Goes Live
  • App tested in the wild for the first time. Parallel spreadsheet as fallback.
  • Sealed beverages added to data set for the first time
  • Loading data captured (first time ASF has this)
  • New economy class split methodology applied
  • Year-on-year comparison possible for second-year audit sites
    • September — Ongoing
    Commercial Proposal Written
  • ASF licence model — what ASF pays, what it gets, how it flows
  • ASF rationale — why ASF is needed in the POC (not just SATS + AI partner)
  • Product + price proposition — what is being sold and at what cost
  • SATS's role in ACA credibility — what the POC unlocks for trade body strategy
  • Regulatory section — PPWR (now law in Europe, Aug 12) + refresh of 2-year-old review
    • September — Decision
    Settle the word: Intelligence or Confidence?
  • This has to land before the narrative can be written
  • Options: Intelligence · Confidence · Accuracy · Standardisation
  • Must be defensible under regulator and journalist scrutiny
  • Drives all brand, pitch, and ACA messaging downstream
    • September — Within 30 Days
    Close Singapore NFP Office
  • Not-for-profit entity — not structured for commercial operation
  • All commercial activity moves to UK entity
  • Lower priority but time-sensitive for legal and tax reasons
    • October — November
    Matt Travels to Singapore for the POC
  • Licence model goes on the table
  • Roles of all three founders defined — what each person does if ASF is in
  • Grant funding conversation: can any be directed toward ASF team structure?
  • Real flight data collected. Waste counts audited. Real cost numbers applied.
    • November — Milestone
    Commercial Model Written in Words
  • "This + this = that." The written formula — not numbers yet
  • The equation that frames every pitch, every ACA conversation, every investor meeting
  • Hypothesis written clearly enough to test at SATS
    • January — March 2027
    POC Data Validated. Model Locked or Pivoted.
  • Credible, defensible claims on waste reduction established from SATS data
  • Commercial model proven (or pivoted)
  • Rebrand decision: if proven → rename and brand for scale
  • If not → pivot proposition before March 2027
    • April 2027 — Target
    WTCE Hamburg — Launch Moment
  • World Travel Catering & Equipment Expo. The industry event.
  • Planned for April 2026 — didn't happen in time. April 2027 is the reset.
  • If commercial model is proven by March → this is the reveal
  • AI video content as the pitch asset for Stage 3 sales begins here
    • Bottom Line SATS is the hypothesis test. September tells us if ASF is in it. November tells us if the equation works. March 2027 tells us if the business scales. Everything before that is prep. Everything after that is proof.