A comprehensive synthesis of eleven primary evidence sources — ALS & Cawthron analytical reports, REIMS and hyperspectral screening data, the Polyphenol Quantification Project, and the PolySure™ method validation — read through the lens of the Four Pillar Value Architecture. This report traces the line from individual certificates of analysis to a structured, licensable, defensible data and IP platform, and sets out concrete next steps to compound that value.
Eleven documents, three laboratories (ALS NZ, Cawthron Institute, AgResearch/Bioeconomy Science Institute), and three independent analytical modalities. This is not a single study — it is a coherent, cross-validated body of evidence that, read together, is substantially more valuable than the sum of its parts.
112-sample CoA quantifying the four official MPI authenticity markers (4-HPLA, 2-MBA, 2'-MAP, 3-PLA) across the full sample set, including Katikati range variants and the Manuka/Kānuka/Horopito powder trio.
P4 DATAP1 QA 25-18852_MPI_ManukaMarkers · ALS NZ · 11/07/2025Same 112-sample set tested for DHA, MGO, NPA and HMF (the UMF-adjacent activity markers) plus Leptosperin — the authenticity/quality backbone for every product across P1 and P2.
P4 DATAP1 QAP2 CoA 25-18852_3in1_Lepto · ALS NZ · 11/07/2025Targeted + untargeted LC‑HRMS across the same 112 samples. 37 phenolics quantified; 57,024 molecular features detected; 10 novel "good unknown" compounds putatively identified. The discovery engine behind PolySure™.
P4 COREP3 SCIENCE Analytica Final Report · ALS Global · Job 25-18702MPL's own synthesis of the funded project: literature review, 112-sample profiling, and translation into a 7-compound priority shortlist. The strategic narrative document linking science to commercial positioning.
P4 STRATEGYP3 FUNDING MPI Māori Agribusiness Fund · Prepared by Dr Lillian MortonThe formal ISO/IEC 17025 validation protocol for the 7-polyphenol UPLC-PDA method — selectivity, linearity, accuracy, repeatability and intermediate precision, designed against Eurachem and AOAC guidelines.
P4 IP Cawthron Method Validation Protocol · 13/02/2026The completed validation: accuracy 82.9–101.9%, RSDr ≤7.9% (low) / ≤1.6% (high), RSDR ≤15.5% / ≤2.4%. Method declared "fit for purpose" — this is the PolySure™ IP certificate.
P4 IP CORE Cawthron Report 4235 · Issued 24/04/2026First production application of the validated method to three honey-based gel products (LiquidFuel, LiquidGold variants) — proof the method works on finished P1 product, not just raw honey.
P4 APPLIEDP1 PRODUCT Cawthron FaB 1970 · 31/03/2026AgResearch/Bioeconomy Science Institute pilot: 2–10 second-per-sample mass spectrometry fingerprinting. Distinguishes monofloral vs multifloral, separates Rata/Kānuka/Rewarewa clusters, and shows directional geographic signal across 7 NZ regions.
P4 SCREENING AgResearch / Bioeconomy Science Institute · 17 slidesHyperspectral imaging chemical profile data across 41 samples (7 blanks removed), with PCA plots in 2- and 3-axis views. A second independent rapid-screening modality alongside REIMS — non-destructive and plate-based.
P4 SCREENING AgResearch / Bioeconomy Science InstituteThe internal strategic narrative covering all four pillars, clinical trial programme, university partnerships (Massey, Jinan, BSI/AgResearch), Kete Rāraunga, Toka Tūpari, and the Data Strategy / Bioactive Intelligence chapter (Part VIII).
P4 NARRATIVEP3P1P2 MPL Internal · 24+ pagesThe investor-facing strategic architecture document (previously prepared) — defines the $373M revenue target, the 10–20× multiple thesis, exit pathways, and the explicit P4 platform revenue streams this report builds toward.
FRAMEWORK MPL_Four_Pillar_Waterfall_Strategy.htmlThis is the single most important thread running through the evidence base: a documented, three-stage pipeline from open-ended chemical discovery to a defensible, ISO 17025-aligned analytical IP asset. Read in sequence, the four polyphenol-related documents tell a complete origination story for PolySure™.
| Parameter | Result | AOAC / ISO Threshold | Outcome |
|---|---|---|---|
| Accuracy (recovery) | 82.9–101.9% | 75–120% (low) · 80–115% (mid–high) | PASS |
| Repeatability (RSDr) — low level (0.5 mg/kg) | ≤7.9% | ≤8% | PASS |
| Repeatability (RSDr) — high level (20 mg/kg) | ≤1.6% | ≤6% | PASS |
| Intermediate precision (RSDR) — low level | ≤15.5% | ≤16% | PASS |
| Intermediate precision (RSDR) — high level | ≤2.4% | ≤12% | PASS |
| Linear range (R² ≥ 0.99, all analytes) | 0.5–25 mg/kg | Acceptance: R² ≥ 0.99 | PASS |
| Matrices validated | 2 | Honey + honey-based gel | PASS |
Source: Cawthron Report 4235 (April 2026), validated per ISO 5725-2 and AOAC Appendix F guidelines. Every parameter met or beat its acceptance threshold across all seven polyphenol analytes — this is a clean pass, not a marginal one.
Alongside the LC-HRMS/UPLC "gold standard" chemistry, MPL has piloted two independent rapid-screening technologies. This is the part of the evidence base most directly relevant to AI/predictive capability — both produce dense, machine-learning-ready signal in seconds rather than hours.
A sample is vaporised by a laser burn and analysed directly — 2–10 seconds per sample, no preparation. Each burn yields a "fingerprint" of 500–3,000 detectable metabolite features. AgResearch's pilot (Bioeconomy Science Institute) ran MPL's sample set across 24-well plates in positive and negative ionisation modes.
A second, independent rapid-screening modality from the same AgResearch/BSI collaboration. Plate-based, non-destructive, optical — chemically distinct mechanism from REIMS, which strengthens any future ensemble/ML approach combining both.
Your Pillar 4 vision explicitly targets terroir mapping — linking geographic origin, season, and floral source to phytochemical expression. The REIMS geographic PCA plots are an early, real demonstration of this, run across seven NZ regions.
The raw material for terroir mapping is substantially further along than the strategy document's framing as a future ambition suggests. The gap is structuring and database integration, not data generation.
Consistent with the Waterfall Strategy's flywheel logic, none of this evidence sits in a P4 silo. Each document does double or triple duty across the four pillars.
| Evidence | Feeds P1 (Consumer) | Feeds P2 (B2B) | Feeds P3 (Clinical/R&D) | Feeds P4 (Platform) |
|---|---|---|---|---|
| ALS MPI Markers + 3in1/Leptosperin | Label substantiation for UMF/activity claims on every batch | CoA backbone for every ingredient supply contract | Baseline chemistry for trial product characterisation | 112-sample structured dataset — the founding rows of a bioactive database |
| Polyphenol Profiling (Analytica/ALS) | Non-MGO health story for marketing beyond wound-care framing | Differentiates non-Mānuka honeys for B2B diversification pitches | Candidate compound list for future mechanistic/clinical work | Discovery layer — 10 putative novel compounds as future IP |
| PolySure™ Validation (Cawthron) | Defensible polyphenol claims on LiquidFuel/LiquidGold labels | The licensable method itself — direct B2B revenue line | Audit-ready method for any future trial requiring polyphenol dosing data | Core licensable IP asset — SaaS/licensing revenue stream |
| REIMS & HSI Pilots | Future rapid authenticity check — anti-counterfeit consumer trust signal | Fast, low-cost QC screening for B2B partners at scale | Methodology precedent for any future "rapid biomarker" trial design | The throughput layer that makes predictive-formulation-as-a-service viable |
| REIMS Geographic/Provenance Data | "Verified origin" storytelling for premium positioning | Provenance substantiation for export market compliance | Environmental covariate data for future biological-outcome studies | Seed data for terroir mapping & Data Ecosystem Mapping |
Not every output of this work is equally defensible. Distinguishing genuine IP from useful-but-replicable lab work is essential for an honest investor conversation.
The validated PolySure™ method itself (Cawthron Report 4235) — a named protocol with documented accuracy/precision data and an explicit IANZ accreditation pathway is a genuine, licensable analytical IP asset. The 112-sample structured CoA dataset, once formalised into a database with consistent schema, becomes a data asset that is expensive and slow for a competitor to replicate (each sample represents real lab spend and multi-week turnaround). The 10 putative novel compounds are first-mover discovery candidates — whoever characterises and names them first has a real priority claim.
REIMS/HSI classification models are currently pilot-stage with explicit "more samples needed" caveats — the underlying instruments and base technique are not unique to MPL, but a trained classifier on MPL's specific NZ-honey dataset, accumulated over time, becomes harder to replicate. Geographic/terroir signal is real but not yet statistically robust at the regional level — defensibility here comes from accumulated sample volume over multiple seasons, not from a one-off dataset.
Kete Rāraunga and the Cultural Life Cycle Assessment (C-LCA) framework, co-developed with AgResearch/BSI, are governance and methodology IP rather than chemistry IP — their defensibility rests on being first to formalise and publish a globally novel indigenous data-sovereignty model, not on a patent-style technical moat. This is real and valuable, but it is a different kind of asset that needs different protection (trademark, published methodology, formal partnership agreements) than the analytical IP above.
— MP Innovation Overview, Part VIII: Data Strategy and Bioactive Intelligence
Eight concrete, evidence-grounded opportunities, each tied directly to a document in the evidence base — followed by a build-out strategy for the three with the fastest, lowest-capital path to revenue, and a maturity map showing where every opportunity currently sits.
Three opportunities are already pilot-ready or in active build — these are the priority build-out targets below. The remaining five are concept-stage: real, evidence-backed, but requiring a defined first step before they become fundable workstreams.
Sell the validated 7-polyphenol UPLC-PDA method (Cawthron Report 4235) as a standalone, paid testing service to other NZ honey producers, ingredient brands, and contract manufacturers — independent of MPL's own product lines. The lab capability and validation data already exist; what's missing is a commercial front-end.
Not a product in itself — the precondition for nearly everything else in this report. Consolidate the 112-sample CoA data, REIMS metadata, and HSI profiles (currently scattered across separate PDFs with inconsistent sample ID formats) into one structured, queryable database. This is unglamorous but is what makes every AI/predictive claim in the Pillar 4 thesis real rather than aspirational.
FaB Results Report 1970 already proves the PolySure™ method works cleanly on finished P1 gel products, not just raw honey. Formalising this into a standard QA step for every new product launch is near-zero incremental cost — the method, lab relationship, and validation data already exist — and it strengthens label-claim defensibility across the entire consumer portfolio.
| Opportunity | What It Is | Evidence Basis |
|---|---|---|
| PolySure™ CoA-as-a-Service | Offer the validated 7-polyphenol method as a paid testing service to other NZ honey producers and ingredient brands, independent of MPL's own product lines — the most immediately executable revenue line, since the method is already validated and production-tested. | Docs 05, 06, 07 |
| Rapid Authenticity Screening (REIMS-lite) | Once REIMS sample volume scales, a fast (seconds-per-sample) screening service for honey authenticity/mislabelling detection — relevant to MPI compliance, export certification, and B2B partner trust, at a fraction of LC-HRMS turnaround and cost. | Doc 08 |
| Non-Mānuka Diversification Line | The polyphenol report's clearest commercial finding — Kānuka and Rewarewa carry distinct, under-recognised bioactive profiles. A direct opportunity to launch differentiated, evidence-backed non-Mānuka products at lower input cost, reducing exposure to Mānuka oversupply and price volatility. | Docs 03, 04 |
| Provenance/Terroir Certification Mark | Use the REIMS geographic metadata as the seed for a structured provenance database, eventually supporting a "verified origin" mark on premium SKUs — directly extending the Data Sovereignty/C-LCA narrative into a tangible consumer-facing asset. | Doc 08, geographic PCA |
| Novel Compound Characterisation Pipeline | Formal NMR/structural confirmation of the 10 putative novel compounds, in partnership with a university (Massey, or the emerging Jinan University relationship) — converts discovery into either a publishable scientific asset or pharma-licensing optionality. See the dedicated Ten Compounds IP Strategy → report. | Doc 03 |
| Bee-Dross / By-Product Bioactive Recovery | The HSI sample set already includes a "Dross" and "Hive dross" class — early raw material for the bio-waste predictive recovery model named in the Waterfall Strategy. Worth a dedicated mini-study to quantify recoverable bioactive value in current waste streams. | Doc 09 sample list |
| Structured Bioactive Database (MVP) | Consolidate the 112-sample CoA data, REIMS metadata, and HSI profiles into a single queryable database (even a well-structured spreadsheet-to-SQL step) — the unglamorous but highest-leverage action, since it's the precondition for every AI/predictive claim in the Waterfall Strategy. | Docs 01, 02, 08, 09 |
| Gel/Finished-Product Polyphenol QA Line | FaB Results Report 1970 proves the method works on finished gels. Formalise this into a standard QA step for every new P1 product launch — strengthens label-claim defensibility company-wide at near-zero incremental cost, since the method already exists. | Doc 07 |
| Opportunity | Revenue Mechanism | Pricing Lever | Horizon |
|---|---|---|---|
| PolySure™ CoA-as-a-Service | Per-sample / per-batch testing fee, paid by external honey producers and ingredient brands | ISO 17025 validation data justifies premium pricing vs. uncertified labs; IANZ accreditation raises ceiling further | Y1 |
| Gel/Finished-Product QA Line | Internal cost-avoidance + indirect revenue via defensible premium label claims on existing P1 SKUs | Claims defensibility supports 30–50% price premium positioning, per existing P1 benchmark | Y1 |
| Non-Mānuka Diversification Line | New SKU revenue (P1) and/or B2B ingredient supply (P2) using lower-cost, currently underutilised honey stock | Evidence-backed differentiation claim; reduces COGS exposure to Mānuka price volatility | Y1–Y2 |
| Rapid Authenticity Screening | Per-sample screening fee, positioned as a lower-cost triage layer ahead of full LC-HRMS confirmation | Speed and cost advantage vs. existing lab turnaround; MPI/export compliance value-add | Y2 |
| Provenance/Terroir Certification Mark | Consumer-facing premium positioning (P1) and a licensable certification mark for third-party use (P2) | "Verified origin" trust signal, tied to the Data Sovereignty/C-LCA narrative | Y2 |
| Bee-Dross / By-Product Recovery | Converts a current waste-management cost into a forecastable, monetisable input stream | Cost-avoidance plus potential new low-cost bioactive input source | Y2–Y3 |
| Novel Compound Characterisation | Licensing/royalty optionality once compounds are confirmed and bioactivity-screened (see Ten Compounds report) | First-mover scientific publication + potential use/process patent protection | Y2–Y3 |
| Structured Bioactive Database | No direct revenue — enabling infrastructure for SaaS/licensing revenue described in the Waterfall Strategy's P4 platform thesis | N/A — foundational asset, not a priced product | Y1, ongoing |
Horizons are indicative, sequenced relative to each other rather than fixed calendar commitments — the database MVP and the two near-zero-cost extensions (CoA service, gel QA line) are deliberately front-loaded since they require the least new capital or evidence to begin.
Sequenced so each step compounds the next, consistent with the Waterfall Strategy's flywheel logic: data mass before SaaS, validation before licensing, structure before AI.
| Horizon | Action | Why / What It Unlocks |
|---|---|---|
| Now | Consolidate all 112-sample CoA data (MPI markers, 3in1, Leptosperin, polyphenols) into one structured database with consistent sample IDs across documents — currently the same lab IDs (e.g. 25-18852-66) appear across three separate PDFs with no linking system. | The single highest-leverage, lowest-cost action available. Without this, "AI-driven predictive formulation" remains a slide, not a capability. |
| Now | Pursue IANZ accreditation for the PolySure™ method, building directly on the completed ISO 17025 validation (Cawthron Report 4235). | Converts an internal method into an externally trusted, licensable standard — the precondition for B2B PolySure™ licensing revenue. |
| Now | Formalise the REIMS and HSI sample metadata (species, geography, season) into the same structured database — both datasets already contain this, it just needs to be merged. | Creates the terroir-mapping seed table the Waterfall Strategy describes as a future ambition — much of the raw material already exists. |
| Near-term | Commission confirmatory characterisation (reference standards + NMR) for the 10 putative novel compounds identified in the Analytica untargeted screen. | Converts "putative" discovery into defensible, potentially patentable or pharma-licensable IP. |
| Near-term | Expand REIMS and HSI sampling across at least one further season, targeting underrepresented regions (Waikato, Canterbury) and a larger Mānuka monofloral cohort. | Addresses the explicit "more samples needed" finding and starts building the multi-season dataset that is genuinely hard for competitors to replicate. |
| Near-term | Run a dedicated bee-dross / by-product composition study, building on the existing HSI "Dross" sample class. | First real data point toward the bio-waste predictive recovery revenue stream named in the Waterfall Strategy. |
| Near-term | Package PolySure™ as an external testing service offer (pricing, turnaround, sample submission process) for non-MPL honey producers. | First externally-facing P4 revenue line — proof point for the "first paying PolySure™ licensing subscribers" milestone in the Series A narrative. |
| Far-term | Train a supervised classification model (species / authenticity / provenance) on the combined REIMS + HSI + LC-HRMS dataset, once sample volume supports it. | The actual AI layer the Waterfall Strategy describes — only viable once the "Now" data-structuring work is done. |
| Far-term | Formalise the Kete Rāraunga / C-LCA framework into a licensable governance product, separate from the analytical IP, with its own documentation and partnership terms. | Realises the "globally unique indigenous IP category" thesis as an actual licensable asset rather than a narrative. |
For an investor conversation, the honest framing is more persuasive than an inflated one. This evidence base supports a specific, credible claim — not the full $3.7B–$7.4B platform thesis on its own, but a concrete proof point that the thesis is being executed, not just described.
This framing keeps the long-range $3.7B–$7.4B platform value thesis from the Waterfall Strategy intact as the destination, while giving near-term investors something more concrete to underwrite: a working method, a real dataset, and a credible 12–18 month path (database consolidation → IANZ accreditation → first external PolySure™ licensee) to the next valuation re-rating trigger.