Qarbotech · channel mockups · Direction 1 — Field Instrument

Four surfaces, executed in the recommended direction — and one of them is the one that actually wins the sale.

The recommendation is Direction 1, Field Instrument, because the decision is made in a room, from paper, by three officers who read instrument output all day. These are the four surfaces that direction has to survive: the one channel where the buyer personally has an account, the one it is weakest on, the one the enquiry actually arrives through, and the dossier — the artefact the other three exist to put in a committee's hands. Two admissions from the losing directions are carried, exactly as the recommendation names them: Direction 3’s dark plate as one bounded image class, and Direction 2’s ledger notation designed now and held.

QB-DOC-DD03 · REV A · 2026-07-31 · governed by CLAIMS.md > canon-locked.md > DESIGN.md · every line of copy re-checked against CLAIMS.md

How to read these mockups

⟨ … ⟩
A value only the client can supply. Pack size, price, concentration, spray volume, pre-harvest interval, response times and staff names are all unpublished. They are drawn as holes, at the size of the thing they hold — never invented, never quietly omitted.
[ PENDING — … ]
An artefact that does not exist yet. It is sized and positioned so its absence is visible on the page rather than hidden by a reflowed layout.
Grey blocks
Unshot photographs, described. There is not one generated, rendered or CGI image in this file — no fabricated micrograph, no scale bar on anything an instrument did not produce, no invented instrument reading. Each block states what would be shot, where, and what stamp it would carry.
Everything else
Is a fact with a source, and the source is on the surface with it.

One stated extension. Direction 1’s notation carries a single marked-placeholder mark, [ PENDING — … ]. These surfaces need two, because an artefact that does not exist and a value only the client holds are answered by different people on different timescales — so the second mark, ⟨ ⟩, is borrowed from Direction 3’s notation set. It is the only thing borrowed from that direction apart from the dark plate, and it is named here rather than smuggled in.

SURFACE 01

LinkedIn company page

THE ONLY CHANNEL WHERE
THE BUYER HAS AN ACCOUNT

5,244 followers, already active, and 100% investor-facing. The officers who decide — the state programme officer, the MySTI focal point, the state agro corporation, the consulting agronomist — are personally here. Nothing addressed to them is.

Company page · professional network mockup — no platform mark reproduced
Cover · dark plate — the one admitted image class

One vial of QarboGrow concentrate on matte black card, lamp off at left, lamp on at right. Same frame, same crop, same exposure. Nothing else in shot — no props, no hands, no smoke, no gradient. Shot in half a day with equipment the company already owns.

PHOTOGRAPH · 365 nm · 12 cm · f/8 · 1/60 s · ISO 200 · ⟨date⟩  —  A PHOTOGRAPH, NOT A MEASUREMENT. NO SCALE BAR, NO n =, NO PANEL LETTER.

Qarbotech Sdn Bhd

Penggalak fotosintesis · Photosynthesis enhancer

SELANGOR, MALAYSIA · AGRITECH · DEVELOPED AT UNIVERSITI PUTRA MALAYSIA · 5,244 FOLLOWERS

Muat turun dosier · Download dossier Follow Hubungi · Contact
About · Mengenai kami

Qarbotech menghasilkan QarboGrow, penggalak fotosintesis semburan daun yang dibina daripada titik kuantum karbon bersaiz bawah 10 nm, dibuat daripada biochar Tandan Kosong Kelapa Sawit — hasil sampingan industri minyak sawit Malaysia. Ia tidak mengandungi nitrogen, fosforus atau kalium. Ia bukan pengganti baja — ia digunakan bersama program pembajaan yang tuan jalankan sekarang.

Qarbotech makes QarboGrow, a foliar photosynthesis enhancer built on carbon quantum dots under 10 nm, made from Empty Fruit Bunch biochar — a by-product of Malaysia’s palm oil industry. It contains no nitrogen, phosphorus or potassium. It is not a replacement for fertiliser — it is used alongside the programme you already run.

Dibangunkan di Universiti Putra Malaysia. Dibina atas kaedah penghasilan titik kuantum karbon yang dipatenkan oleh Universiti Putra Malaysia — MY195075A, diberikan 9 Januari 2023. Paten itu meliputi cara ia dibuat, bukan kesannya pada pokok. MySTI Certified STI202300224. Rakan penyelidikan: Agensi Nuklear Malaysia.

Developed at Universiti Putra Malaysia. Built on a carbon-quantum-dot production method patented by Universiti Putra Malaysia — MY195075A, granted 9 January 2023. The patent covers how it is made, not what it does to a plant. MySTI Certified STI202300224. Named research partner: Agensi Nuklear Malaysia.

What we have not published, and are not going to invent: pack size, price, concentration, spray volume and pre-harvest interval. Until they exist they appear on our surfaces as marked gaps. Ask and we will tell you what we have and what we do not.

QARBOTECH SDN BHD · 1291227-W
MySTI STI202300224
DEVELOPED AT UNIVERSITI PUTRA MALAYSIA

QARBOTECH · 5,244 followers · 1d P1 · THE HONEST NUMBER

No control plot was stated in this pilot. So read it as our pilot data and not as a trial — and here is the number anyway.

Average +11.57% yield increase across 300+ farmers on 500 hectares in Malaysia and Indonesia, 2024. The best plot did 45%. We are not going to pretend that is typical.

We reported this to a funder in 2025. We have not been saying it to farmers, and that is the thing we are changing first.

FIG. 1   QarboGrow rice pilot, 2024 — reported yield increase, mean and peak

MEAN 11.57% PEAK 45%
0 10 20 30 40 50

Reported yield increase (%)

The mean sits hard left because that is where it sits. The line to the cap is the distance between our average and our best plot, and it is the part of this figure most companies delete. Comparator not stated in the source; no control plot. We call it pilot data, not a trial.

SOURCE: QARBOTECH MIT SOLVE APPLICATION, 2025 · n = 300+ FARMERS · 500 ha · MALAYSIA + INDONESIA · 2024 · ASSERTED

From now on: every performance number we publish carries its mean, its range, its n, its crop and its date. Where an older number of ours does not, we are re-cutting it or taking it down. If you find one, tell us and we will deal with it.

QARBOTECH SDN BHD · 1291227-W  ·  MySTI STI202300224  ·  DEVELOPED AT UNIVERSITI PUTRA MALAYSIA

#PadiMalaysia#Pertanian#Biostimulant n AND DATE IN THE POST BODY, NEVER IN A COMMENT
QARBOTECH · 5,244 followers · 4d P3 · THE PROGRAMME RECORD

Fakta penggunaan — bukan pengesahan, bukan sokongan.

Program 2,000 hektar di Seberang Perai Utara, Tengah dan Selatan. 1,319 pesawah. Harga produk RM320 sehektar. Nilai projek RM1.622 juta. Dilancarkan 1 Oktober 2025, dijalankan oleh Jabatan Pertanian Pulau Pinang dalam kerjasama strategik dengan Qarbotech. Dilaporkan oleh BERNAMA, dalam Bahasa Malaysia dan Bahasa Inggeris.

Dua perkara yang kami tidak akan buat dengan angka ini. Kami tidak berkata program itu mengesahkan produk kami — itu keputusan Jabatan, bukan kami. Dan kami tidak mendarabkan RM320 dengan 2,000 lalu memanggilnya nilai kontrak: produk itu RM640,000, dan bakinya bukan milik kami untuk dijelaskan. Tanya Jabatan, bukan kami.


Fact of use — not endorsement, not approval.

A 2,000-hectare programme across Seberang Perai Utara, Tengah and Selatan. 1,319 farmers. Product price RM320/ha. Project value RM1.622 million. Launched 1 October 2025, run by Jabatan Pertanian Pulau Pinang in strategic cooperation with Qarbotech. Reported by BERNAMA.

Two things we will not do with those numbers. We will not say the programme validates our product — that is the Department’s call, not ours. And we will not multiply RM320 by 2,000 and call it a contract value: the product line is RM640,000, and the rest is not ours to explain. Ask the Department, not us.

SOURCE: BERNAMA, 1 OCT 2025 · MEASURED   RM640,000 = 2,000 × RM320 · DERIVED — OUR ARITHMETIC

QARBOTECH SDN BHD · 1291227-W  ·  MySTI STI202300224  ·  DEVELOPED AT UNIVERSITI PUTRA MALAYSIA

NO CRESTNO SEALNO PERCENTAGE ADJACENT TO A DEPARTMENT’S NAME
QARBOTECH · 5,244 followers · 1w P4 · THE HECTARE · DOCUMENT

Enam muka surat, boleh dimuat turun, tiada satu pun dakwaan prestasi di dalamnya.

Kalau tuan yang kena tulis kertas kerja, ini bahagian yang tiada siapa terbitkan: band perolehan mana yang jatuh pada saiz plot tuan, band mana yang memang tidak boleh dicapai, dan laluan sah yang ada — kajian pasaran.

Six pages, downloadable, and there is not one performance claim in it. If you are the officer who has to write the kertas kerja, this is the part nobody publishes: which procurement band your own plot size lands in, which band is unreachable, and the one lawful route in — the kajian pasaran.

01 · KULIT / COVER

Bagaimana sebuah jabatan pertanian negeri membeli input
How a state agriculture department buys an input

Nota untuk pegawai program.
A note for programme officers.

QB-DOC-0044 · REV A

02 · EMPAT BAND / FOUR BANDS

At RM320/ha, your hectares decide your route

≤ 156 ha · pembelian terus
156–313 ha · sebut harga
313–1,562 ha · sebut harga terbuka
> 1,563 ha · tender terbuka

DERIVED · OUR ARITHMETIC

03 · BAND MATI / THE DEAD BAND

156–313 ha is not a choice

The RM50,000–100,000 sebut harga band is Bumiputera-reserved. For some suppliers a first order sized into it is legally unreachable. Aim below it or above it — never into it.

DERIVED

04 · LALUAN SAH / THE LAWFUL ROUTE

We are not allowed to help you write your specification

PK 2.1 §5.2(b): agencies are forbidden from taking specification advice from suppliers. The sanctioned substitute is the kajian pasaran under §5.4. Published material is how a supplier lawfully gets in front of one.

MEASURED

05 · RM320 / WHAT IT SITS NEXT TO

One insentif racun, near enough

RM320/ha product, Penang rate
7.4% of the RM4,300/ha/season the state already gives each paddy farmer
≈ the RM300/ha insentif racun line
RM811/ha all-in on the Penang programme

MEASURED + DERIVED

06 · BELUM ADA / NOT YET

What we cannot answer yet

[ ] dose table
[ ] SDS
[ ] pack size & price
[ ] pre-harvest interval
[ ] control-plot trial, named agronomist
patent that resolves — MY195075A

PUBLISHED AS A GAP
PDF · 6 pagesBM / EN QB-DOC-0044 · REV A · 2026-08-04 NO CLAIM IN THIS DOCUMENT
DOCUMENT POST FORWARDABLE INSIDE A MINISTRY WITHOUT VOUCHING FOR IT
QARBOTECH · 5,244 followers · 2w P2 · WHAT IT DOES NOT DO

Empat perkara yang QarboGrow tidak buat.

1. Ia bukan baja. Ia tiada N, P atau K — itu ayat dari FAQ kami sendiri.
2. Ia bukan racun. Ia tidak mengawal penyakit atau perosak, dan kami tidak akan membuat dakwaan itu.
3. Ia bukan pengganti program pembajaan tuan. Baja bagi makan pada akar; benda ini tolong daun guna cahaya. Tuan tetap membaja macam biasa.
4. Ia tidak sesuai untuk semua tanaman. Senarai penuh belum kami terbitkan — kami sedang membetulkan percanggahan dalam bahan kami sendiri dahulu. Bila ia siap, ia keluar dengan sel yang kosong pun ditunjukkan.

Dan satu amaran dari FAQ kami sendiri: jangan lebihkan kadar. Berlebihan boleh menekan pokok dan menyebabkan perencatan cahaya. Ada nombor di belakang ayat itu, dan kami akan terbitkan nombor itu bersama jadual dos.


Four things QarboGrow does not do.

1. It is not a fertiliser. It contains no N, P or K — our own FAQ says so.
2. It is not a pesticide. It does not control disease or pests, and we will not make that claim.
3. It is not a replacement for your fertiliser programme. Fertiliser feeds the root; this helps the leaf use light. You keep fertilising exactly as you do now.
4. It does not suit every crop. We have not published the full list yet — we are reconciling contradictions in our own material first. When it ships, the empty cells will be shown as empty.

And one warning from our own FAQ: do not over-apply. Excess may stress the plant and cause photo-inhibition. There are numbers behind that sentence and we will publish them with the dose schedule.

CATEGORY, HANDLING AND DOSING FACTS · SOURCE: QARBOTECH FAQ · MEASURED  ·  NO PERFORMANCE CLAIM IN THIS POST

BM LEADSSAME SIZE, SAME WEIGHT THE CROP TABLE IS DESIGNED AND HELD — SEE SURFACE 04

What this does — and what the current 5,244-follower page does not

The page is already there and it is already the right channel: measured at 5,244 followers, 100% B2B and investor-facing, and it is where the state programme officer, the MySTI focal point, the state agricultural development corporation, the corridor authority and the consulting agronomist personally have accounts. What it publishes is announcements. Five things change.

  1. Every claim-bearing post carries its n and its date in the post body, never in a comment. A LinkedIn post is the format most likely to be screenshotted into a tender objection, so the screenshot has to contain its own defence. Today a rival bidder can screenshot a number and a caveat separately.
  2. The lead number becomes the company’s own honest one. +11.57% mean, 45% peak, n = 300+, 500 ha, 2024. It was disclosed to a funder in 2025 and to nobody the market can see. Publishing it is not a concession: the nearest carbon-dot rival publishes “9% average across 30+ independent trials” and out-evidences “up to 60%” with a smaller number. The rival with the smaller number has the better proof — and that is the position being taken back here.
  3. Bahasa Malaysia leads on anything addressed to a state department. There is currently zero BM across 36 captured files on either corporate site. It is the single largest content gap in the file, it is free to fix, and it is the one kill row the district extension officer holds. BM here is set at the same size and the same weight as the English — where a line must hold both, the condensed cut carries it, never a smaller size.
  4. One post a month is a document, not an image. An officer forwards a PDF into a ministry; he cannot forward a link he would have to vouch for. The six-page procurement note contains no performance claim at all, which is exactly why it can travel through an inbox that a claim could not.
  5. The provenance stamp sits on every image and every figure — instrument, place, date, operator. A photograph that cannot carry a complete stamp does not publish, which retires the terraced-hillside stock, the archive rescue and the three empty counters without anyone having to argue about taste.
Three things it deliberately does not do

It does not reuse 10.5 t/ha or the Terengganu 300 ha figure. Both are live on the current page and appear in no row of the trial register; the first sits against Penang’s publicly reported 6–7 t/ha. They need a plot, a date and a farmer before they appear anywhere.

It does not print “our headline is 5.19× our own average”. That arithmetic is ours to act on, not the client’s to publish next to a live claim. The correction ships as a standing rule — from now on every number carries its mean, range, n, crop and date — not as a confession.

It does not mention C4 crops, in any direction. The single most tempting LinkedIn post available is “we are not claiming corn”, and no competitor could write it. It is still off, because saying nothing about C4 means saying nothing, including saying nothing about not claiming it. It ships the day the client answers.

SURFACE 02

Instagram

WEAK FOR THIS BUYER
— AND IT EARNS ITS PLACE ANYWAY

A ministry procurement officer does not buy from Instagram. So the feed is built as three narrow jobs it is genuinely good at — the field rig, Bahasa Malaysia dosing answers, and proof a state officer can show his own committee — and it is not allowed to do anything else.

Profile · photo feed mockup — no platform mark reproduced
QARBOTECH

qarbotech

46 posts2,495 followers1 following

Qarbotech Sdn Bhd
Penggalak fotosintesis · Photosynthesis enhancer
Petak ujian kami, langsung · our trial plot, live
Soalan dos dalam BM · dosing questions in BM
Tuai diisytihar dahulu · harvest date declared first

QB-DOC-0021 · dossier in bio

Column 1 · the rig, weekly
Column 2 · BM dosing & handling
Column 3 · committee-showable proof

RIG 01 · 2026-08-04

The plot, marked out

Two strips, same block, same water. One sprayed, one not. Both flagged before anything goes on.

[ PHOTOGRAPH — flat-lit,
mid-morning, both strips and
the boundary between them ]

⟨plot⟩ · 2026-08-04

DOS · PADI

4 pusingan.
Hari 40 · 50 · 60 · 70.

Itu jadual yang kami terbitkan untuk padi. Semburan awal pagi atau lewat petang.

⟨kadar campuran —
belum diterbitkan⟩

SUMBER: FAQ QARBOTECH

REKOD PROGRAM

2,000 ha
1,319 pesawah

Pulau Pinang, dilancar 1 Okt 2025. Dijalankan oleh Jabatan Pertanian Pulau Pinang. Fakta penggunaan, bukan pengesahan.

SUMBER: BERNAMA

MEASURED · 2025-10-01

RIG 02 · 2026-08-11

The control strip stays in every photograph

If we ever show you a treated strip without the untreated one beside it, you should stop believing us.

[ PHOTOGRAPH — the flag,
the untreated strip, the boundary ]

⟨plot⟩ · 2026-08-11

SIMPANAN · STORAGE

Simpan sejuk dan gelap. Jangan bawah matahari.

Jangan lebihkan kadar — berlebihan boleh menekan pokok dan sebabkan perencatan cahaya. Itu ayat kami sendiri.

Hayat simpanan 2 tahun.

SUMBER: FAQ QARBOTECH

ANGKA JUJUR · THE HONEST NUMBER

+11.57% mean
45% peak

n = 300+ · 500 ha · MY + ID · 2024. Pilot data, not a trial.

ASSERTED · NO CONTROL STATED

RIG 03 · 2026-08-18

The instrument is named before the first reading

LI-COR LI-6800. Declared harvest date 2026-11-18 — set before anyone knows the answer.

[ PHOTOGRAPH — the head
clamped on a living leaf, in field ]

LI-6800 · ⟨plot⟩ · OP:⟨name⟩

KATEGORI · CATEGORY

Ini bukan pengganti baja.

Baja bagi makan pada akar; benda ini tolong daun guna cahaya. Tuan tetap membaja macam biasa. Tiada N, P atau K di dalamnya.

Bukan racun. Kami tidak buat dakwaan penyakit.

SUMBER: FAQ QARBOTECH

BELUM ADA · NOT YET

Six things a committee asks for. We have two.

[ ] dose table  [ ] SDS  [ ] control-plot trial  [ ] named owner of the yield claim   regulatory class stated   a patent number that resolves.

PUBLISHED AS A GAP

Two posts, opened

KATEGORI · CATEGORY · 4:5

Ini bukan pengganti baja.

Baja bagi makan pada akar.
Benda ini tolong daun guna cahaya.
Tuan tetap membaja macam biasa.

Tiada N, P atau K di dalamnya. Untuk padi: 4 pusingan, hari 40 · 50 · 60 · 70.

⟨kadar campuran — belum diterbitkan⟩

QARBOTECH SUMBER: FAQ QARBOTECH · TIADA DAKWAAN PRESTASI

Caption · BM native, EN mirror

Ini bukan pengganti baja. Baja bagi makan pada akar; benda ini tolong daun guna cahaya. Tuan tetap membaja macam biasa. QarboGrow tiada N, P atau K — itu ayat dari FAQ kami sendiri.

Untuk padi, jadual yang kami terbitkan ialah 4 pusingan: hari 40, 50, 60 dan 70 selepas tanam. Sembur awal pagi atau lewat petang. Simpan di tempat sejuk dan gelap.

Kadar campuran? Kami belum terbitkan satu kadar yang kami sanggup pertahankan, jadi kami tidak akan reka satu di ruangan komen. Tanya kami dan kami beritahu apa yang ada dan apa yang belum ada.

Not a replacement for fertiliser. Fertiliser feeds the root; this helps the leaf use light. You keep fertilising exactly as you do now. Contains no N, P or K — our own FAQ. For paddy the published schedule is 4 passes at days 40, 50, 60 and 70. Dilution rate: we have not published one we can defend, and we are not going to invent one in a comment.

SOURCE: QARBOTECH FAQ · MEASURED  ·  NO PERFORMANCE CLAIM · NO DISEASE OR PEST CLAIM · NO CROP LIST

FIG. 1 · COMMITTEE-SHOWABLE · 4:5

Our own pilot data, stated the way we state it to funders.

Data perintis kami sendiri, seperti kami nyatakan kepada penyandang dana.

MEAN 11.57% PEAK 45%
0 20 40 50

Reported yield increase (%) · n = 300+ farmers · 500 ha · MY + ID · 2024

No control plot was stated in this pilot, so we call it pilot data and not our trial. The line to the cap is the distance between our average and our best plot.

Tiada petak kawalan dinyatakan, jadi kami memanggilnya data perintis, bukan ujian kami.

Peta tapak · site map — [ PENDING: which 500 ha, by state ]

QARBOTECH MIT SOLVE APPLICATION, 2025 · ASSERTED — NO CONTROL STATED

Caption · EN lead, BM mirror — the number is identical in both

Average +11.57% across 300+ farmers on 500 hectares in Malaysia and Indonesia, 2024. Peak plot 45%. No control plot was stated, so we call it our pilot data and not our trial. Results vary by crop, variety and growing conditions.

Purata +11.57% merentasi 300+ petani di 500 hektar di Malaysia dan Indonesia, 2024. Petak terbaik 45%. Tiada petak kawalan dinyatakan, jadi kami memanggilnya data perintis, bukan ujian kami. Keputusan berbeza mengikut tanaman, varieti dan keadaan.

This is the slide a state officer can put in front of his own people. It is deliberately not the biggest number we have ever published — it is the one that survives being asked where it came from.

SOURCE: QARBOTECH MIT SOLVE APPLICATION, 2025 · ASSERTED  ·  SAME NUMBER IN BOTH LANGUAGES, ALWAYS

What this does — and the honest part first

Instagram is weak for this buyer and the content direction says so in those words. The measured position is worse than “low priority”: 2,495 followers on the corporate handle, 618 on the Facebook page, and the consumer-brand handle has been converted into the founder’s personal account — someone searching for the consumer product lands on a professor’s CV. Meanwhile a single earned television feature returned 20,000 views against 593 lifetime views on the owned video channel, and 100% of the last 20 captions on both accounts are in English. A ministry procurement officer does not buy from Instagram.

So these fifteen pieces a month are not allowed to be brand mood. They do three jobs and nothing else, and the grid is built in three columns so the system is legible at a glance — which is also what makes it schedulable by someone who is not a designer.

  1. Column 1 — the rig, weekly. A public trial log is the answer to “show me YOUR plot”, the #1 credibility test across ~700 measured comments, and it is the only pillar that can publish in week one, because the method publishes before the result. The plot marked out, the control strip flagged, the instrument named, and the harvest date declared before anyone knows the answer — which is the single most credible thing this brand can do and it costs nothing.
  2. Column 2 — dosing, handling and category, in Bahasa Malaysia. Three of the nine tiles are BM-native and they are the ones the audience actually asks for: the #2 question in the whole category is “is it compatible with MY plant?” and the #5 is “what KIND of thing is this?” The folk taxonomy — baja subur, baja buah, baja daun, racunonly exists in BM, and a product that cannot be filed in it cannot be bought. Note the reframe in the third tile: bukan baja is published as an instruction, not an identity, because the most-upvoted comment in this category is a grower warning others off a non-baja input.
  3. Column 3 — committee-showable proof. A state officer needs something real to put in front of his own people: the programme record as fact of use, the honest number with its n, and — the tile no competitor will ever post — the list of what we do not have yet.
Why the nine tiles read as one system

Every tile carries the same solid black foot band at 11% of the frame, mark left, stamp right. It is the component that lets a daylight field photograph, a type-only BM dosing card and a paper-figure chart sit in one grid without any of them being graded toward the others. One tile inverts — the dark plate, and only one, because that image class is bounded to covers, dividers and the feed and is banned from the document body. No colour appears anywhere in the grid unless it is a measurement.

The thing to be blunt with the client about

This surface will never sell a Perancangan Perolehan Tahunan, and if it starts producing aspirational agriculture it is fifteen pieces of nothing a month. There is also an unscoped elephant next door: the dormant consumer TikTok account holds 24,700 followers — 67.9% of all owned reach — and its only bio link is a hiring form. That is the largest owned-audience waste in the file and it belongs in a scope conversation, not in a quiet workaround.

SURFACE 03

WhatsApp — two lines

WHERE THE FUNNEL PAGE LANDS
AND WHERE A HUMAN TAKES OVER

Four phone numbers are published across the company’s surfaces today, and one of them is simultaneously the consumer WhatsApp and the CEO’s contact on a university page. A smallholder asking about dosage and a state procurement officer are landing on the same phone, and one of them is being answered by the founder. Two lines is a consolidation, not an addition.

The page that sent them · QarboGrow funnel page · QB-DOC-0031
Dose panel — present, and visibly unfinished

The funnel page carries the four-pass schedule as a real bar with real numbers on it, and the dose table as a marked hole at the size of the table it will hold. The button below pre-fills the message with the page reference, so the enquiry arrives already carrying the document it came from.

4 passes · d40 · d50 · d60 · d70 — published
Regulatory class stated — biostimulant, not a pesticide
Patent that resolves — MY195075A, UPM, 9 Jan 2023
[ ] dose table   [ ] SDS   [ ] pack size & price
[ ] pre-harvest interval   [ ] control-plot trial, named agronomist
[ ] a named owner of the yield claim
Tanya di WhatsApp · Ask on WhatsApp pre-fill: “QarboGrow · jadual semburan padi · QB-DOC-0031”
Business account · messages are answered by a team

QarboGrow · jadual semburan padi · QB-DOC-0031

13:41

Terima kasih kerana menghubungi QarboGrow. Talian ini untuk pembelian runcit dan soalan penggunaan.

Balas 3 baris supaya kami boleh jawab betul:

  1. Tanaman apa?
  2. Berapa luas (hektar/ekar)?
  3. Negeri?

Jawapan manusia: 9 pagi–6 petang, Isnin–Sabtu. Sasaran masa jawapan: ⟨belum ditetapkan⟩

13:41 · TEMPLATE

Padi. 1.2 hektar. Kedah.

13:44

Baik. Untuk padi, jadual yang kami terbitkan ialah 4 pusingan — hari 40, 50, 60 dan 70 selepas tanam.

Sembur awal pagi atau lewat petang. Guna sprayer galas atau bermotor. Simpan di tempat sejuk dan gelap, jangan bawah matahari.

Jangan lebihkan kadar — berlebihan boleh menekan pokok dan sebabkan perencatan cahaya. Itu ayat dari FAQ kami sendiri.

13:44 · TEMPLATE

Campur berapa? 1 tutup botol untuk berapa liter air?

13:45

Ini soalan yang kami tidak akan reka jawapannya.

⟨kadar campuran — belum diterbitkan⟩

Kami belum terbitkan satu kadar yang kami sanggup pertahankan untuk padi, jadi saya tidak akan bagi angka di sini. Saya hantar soalan ini kepada orang kami sekarang, dan kami WhatsApp tuan semula bila ia siap.

13:45 · TEMPLATE

Harga sebotol berapa? Boleh POS ke Kedah?

13:46
Diserahkan kepada ⟨nama⟩ · jualan runcit · 13:47

Salam tuan, saya ⟨nama⟩ dari Qarbotech. Saya ambil alih dari sini.

Untuk 1.2 hektar padi di Kedah: jadual dan cara guna saya boleh beritahu hari ini. Saiz peket dan harga masih belum kami terbitkan⟨saiz peket dan harga⟩ — dan saya tidak akan bagi angka yang saya tidak boleh pertahankan.

Boleh saya simpan nombor tuan dan WhatsApp semula hari ia keluar?

13:48 · HUMAN

Line 1 · retail. The volume line. Click-to-WhatsApp from every ad, every product page and every pack QR. Auto-greeting, crop and area qualifying, dosing and handling answers, then a named human against a stated response time. It carries its own pixel and Conversions API feed.

The page that sent them · Contract & trade · QB-DOC-0031
The hectare worksheet — it computes your hectares, and asserts no yield

Same funnel page, second door. The worksheet takes the buyer’s own hectares at the published Penang rate and sets the result against his own input budget. It does not compute a return, a payback or a multiple, because none of those exist as a defensible number.

RM320/ha product, Penang rate · MEASURED
RM811/ha all-in on the Penang programme · DERIVED
7.4% of the RM4,300/ha/season already given per paddy farmer
≈ one insentif racun at RM300/ha
[ ] ROI, payback, return multiple — not computed, by design
Bincang kontrak · Contract enquiry pre-fill: “Contract · hectare worksheet · QB-DOC-0031”
Business account · routed to a named person

Contract · hectare worksheet · QB-DOC-0031

10:02

Terima kasih. Talian ini untuk ladang, pengedar, pihak berkuasa jelapang, program negeri dan ahli agronomi. Ia bukan talian runcit.

Thank you. This line is for estates, dealers, granary authorities, state programmes and agronomists. It is not the retail line.

Three questions, then you go to a named person:

  1. Your role?
  2. Crop and hectares?
  3. State / district?
10:02 · TEMPLATE

Pegawai program, Jabatan Pertanian ⟨negeri⟩. Padi, lebih kurang 900 hektar. Untuk PPT FY2027.

10:09

900 ha, at the published Penang product rate of RM320/ha, is RM288,000 of product. That is our arithmetic, at a rate that is on the public record — not a quotation.

On the Penang programme the all-in cost was RM811/ha. Product is not the whole line, and we are not going to present it as if it were.

900 ha sits in the open sebut harga band (313–1,562 ha). Below 156 ha would be pembelian terus. Both derived from the same rate.

10:09 · TEMPLATE

Sending you two things now:

• QB-DOC-0021 REV B — the technical evaluation dossier, BM/EN. Page 9 is the list of what we have not published.
A booking link, if you want the 30 minutes.

Routing you to ⟨named person⟩, contract sales.

10:10 · TEMPLATE
Assigned to ⟨named person⟩ · contract sales · 10:11

Salam tuan. ⟨nama⟩ di sini.

Satu perkara dahulu, supaya jelas: kami tidak boleh membantu tuan menulis spesifikasi. PK 2.1 §5.2(b) melarangnya, dan kami tidak akan cuba.

Yang kami boleh buat ialah menyerahkan dokumen supaya ia ada di depan tuan semasa kajian pasaran. Dokumen itu menyatakan apa yang ada pada kami dan apa yang belum — muka surat 9 ialah senarai yang belum ada.

One thing first, so it is clear: we cannot help you write your specification. PK 2.1 §5.2(b) forbids it. What we can do is put the document in front of you during your kajian pasaran. Page 9 is the list of what we do not yet have.

10:13 · HUMAN

Line 2 · contract & trade. The value line. Estates, dealers, distributors, granary authorities, state programmes and agronomists. The qualifying flow captures role, hectares and state, then routes to a named person and offers the dossier and a booking. Never used for retail traffic — which is precisely what takes the founder’s number off the consumer surfaces and out of the smallholder queue.

What this does — and how it connects to the funnel page

The connection is the pre-fill, and it is the accession-number idea applied to an enquiry

Every click-to-WhatsApp button on the funnel page pre-fills the first message with the page’s own document reference. So a conversation opens already carrying a resolvable identifier — the team knows which panel produced the question before anyone types a word, and the buyer’s question is attached to the exact version of the page he read. That directly answers the kill-shot on the record from the agronomist every estate defers to: “it took me four minutes to learn yours doesn’t resolve.” A reference that resolves is a deliverable, not a decoration — and it costs one URL parameter.

Two lines, because the current one phone is doing two incompatible jobs

Four numbers are published across the company’s surfaces right now, and the consumer WhatsApp number is also published as the CEO contact on a university institute page. The split is a consolidation: Line 1 takes the volume (dosing, handling, orders, catalogue) and Line 2 takes the value (hectares, role, dossier, booking, a named person). Each line carries its own pixel and Conversions API feed, so a retail order and a 2,000-hectare enquiry never train the same optimisation.

What the transcripts are actually demonstrating
  1. The marked placeholder works inside a chat. When the farmer asks the dilution question — the single most-asked question in the category — the flow answers “this is a question we will not invent an answer to” and shows the hole. The published dilution is unresolved across roughly 100× and one reading of it implies more spray water per hectare than a knapsack can carry. A wrong number here is a burnt field and a story told to the whole kedai baja.
  2. The human takeover is where the honesty is hardest and it is scripted anyway. “I will not give you a figure I cannot defend” is a sales line that loses the order today and keeps the account for ten years, and it is the only line consistent with everything else on these four surfaces.
  3. Line 2 opens by refusing something. Telling a procurement officer, unprompted, that helping write his specification is illegal and that the lawful route is the kajian pasaran is worth more than any claim in the conversation. It is also, precisely, the reason the dossier exists.
  4. The arithmetic is labelled as arithmetic. RM288,000 is presented as our arithmetic at a public rate, immediately followed by the all-in RM811/ha, so nobody can later say the product price was passed off as the programme cost.
Two production notes

The bubbles are square and there is no platform green anywhere. What is being reviewed here is the script, not the chrome — the platform will apply its own colour and corners, and setting the transcript in the brand’s own achromatic register is the only way to judge whether the words hold. And no BM line here is a translation of an EN line. Line 1 is BM-native throughout; the number in any line that carries one is identical in both languages, always — because every measured claim failure in this brand’s history has been a paraphrase that silently deleted a qualifier.

SURFACE 04

The dossier · tender document

THE ARTEFACT THAT ACTUALLY WINS
A PROCUREMENT REVIEW

Penilaian teknikal secara edaran adalah tidak dibenarkan. The technical evaluation committee must physically meet. The sale is three officers and a stack of paper, and the artefact that survives that meeting is a dossier, not a deck. Every other surface in this document exists to put this one into a committee’s hands.

Cover · A4

QB-DOC-0021 · REV B · 2026-08-04 · SALINAN TERKAWAL / CONTROLLED COPY

Dokumen Penilaian Teknikal

Technical Evaluation Dossier

QarboGrow — penggalak fotosintesis semburan daun. Untuk semakan Jawatankuasa Penilaian Teknikal.

QarboGrow — foliar photosynthesis enhancer. For Technical Evaluation Committee review.

1 Apa produk ini · What this product is
2 Bukti · Evidence
3 Skop · Scope
4 Kadar dan penggunaan · Dose and use
5 Rekod program · Programme record
6 Kos sehektar · Cost per hectare
7 Kedudukan kawal selia · Regulatory position
8 Rujukan · References
9 Apa yang belum kami terbitkan · What we have not published

Setiap dakwaan prestasi dalam dokumen ini membawa min, julat, n, tanaman dan tarikh di ruang bukti. Dakwaan yang ruang buktinya kosong tidak diterbitkan.

Every performance claim in this document carries its mean, range, n, crop and date in the evidence margin. A claim whose evidence margin is empty is not published.

QARBOTECH SDN BHD · 1291227-W
MySTI STI202300224
DEVELOPED AT UNIVERSITI PUTRA MALAYSIA

Why the cover carries no photograph. The dark plate is admitted on covers — but a tender cover is printed one-colour, bound, photocopied by the agency and scanned into a file. Four generations of loss. This cover is set to survive all four unchanged, so the type, the 6 mm band and the stepped fore-edge block do the finding, and the vial stays on the surfaces where it is read once and in colour.

The fore-edge block (right edge, above) is a 4 mm solid bar whose vertical position steps down by section, so the closed document shows a printed index on its edge. It is identical in full colour, in one-colour, in a photocopy and in a scan — which a coloured tab is not.

Interior spread · pages 6–7

3.1 · MEASURED Qarbotech FAQ; MySTI listing STI202300224
3.2 · MEASURED Qarbotech FAQ — 4 passes, d40/50/60/70
3.3 · HELD Gated on an unanswered question. See §9.4
3.4 · MEASURED Qarbotech FAQ — overdose / photo-inhibition
TAGSMEASURED · DERIVED · ESTIMATED · UNVERIFIED
3 — Skop · Scope
3.1 — KELAS PRODUK / PRODUCT CLASS

Biostimulan. Penggalak fotosintesis. Bukan baja — ia tidak mengandungi nitrogen, fosforus atau kalium. Bukan racun perosak — ia tidak mengawal penyakit atau perosak, dan tiada dakwaan sedemikian dibuat di mana-mana permukaan kami.

Biostimulant. Photosynthesis enhancer. Not a fertiliser — contains no nitrogen, phosphorus or potassium. Not a pesticide — it does not control disease or pests, and no such claim is made on any of our surfaces.

3.2 — PENGGUNAAN BERULANG / RECURRING OPERATION

Padi: 4 pusingan pada hari 40, 50, 60 dan 70 selepas tanam. Sprayer galas atau bermotor. Awal pagi atau lewat petang.

Paddy: 4 passes at days 40, 50, 60 and 70 after planting. Knapsack or motorised sprayer. Early morning or late afternoon.

3.3 — JADUAL TANAMAN / CROP TABLE

Belum diterbitkan. Bentuknya sudah direka dan ditahan — lihat di bawah. Ia akan keluar dengan sel yang kosong ditunjukkan sebagai kosong.

Not published. The form is designed and held — below. It ships with the empty cells shown as empty.

Held — not for publication · gated on §9.4
Tanaman C3
C3 crops
kuat / strong
Tanaman C4
C4 crops
NO CLAIM
Pokok berkanopi besar
Large-canopy tree crops
[ ]
Akuatik tenggelam
Submerged aquatic
NO

[ ] not yet · NO CLAIM we are not making a claim · NO it does not work. Three different absences, three different marks. An em dash is never used for any of them.

Caveat

No competitor in this category publishes a negative. Neither do we — yet. This table is the strongest asset in the file and it does not ship until the company’s own two websites agree with each other.

QB-DOC-0021 · REV B · 2026-08-046
4 — Kadar dan penggunaan · Dose and use
4.1 — JADUAL SEMBURAN / SPRAY SCHEDULE

Empat pusingan, sepanjang satu musim padi.

Four passes across one paddy season.

FIG. 4   Jadual semburan padi · paddy spray schedule

d40 d50 d60 d70
0 Hari selepas tanam · days after planting 100

Four passes, each carrying its own day number. Every bar carries a number; a bar without one is a picture.

4.2 — KADAR CAMPURAN / DILUTION RATE

[ PENDING — dose table ]
⟨stock concentration, g/L⟩ · ⟨dilution, per crop and stage⟩
⟨spray volume, L/ha⟩ · ⟨pre-harvest interval⟩

This block is sized to the table it will hold. It is not published because we do not yet have a rate we can defend, and an invented one would be the most expensive sentence in this document.

4.3 — AMARAN DOS / DOSE WARNING

Jangan lebihkan kadar. Kepekatan melebihi optimum boleh menekan pokok dan menyebabkan perencatan cahaya. Ayat ini dari FAQ kami sendiri, dan ada nombor di belakangnya dalam kesusasteraan.

Do not over-apply. Concentration above optimum may stress the plant and cause photo-inhibition. This is our own published sentence, and the literature has numbers behind it.

Caveat · sistem tertutup / closed systems

Jangan gunakan dalam sistem hidroponik kitar semula tanpa kajian pengumpulan takungan.

Do not use in a recirculating hydroponic or CEA system without a reservoir-accumulation and clearance study. In a closed loop, concentration only rises.

4.1 · MEASUREDQarbotech FAQ · FIG. 4
4.2 · UNPUBLISHED Client-ask §9.2. Ratio unresolved across ~100× in our own material
4.3 · MEASUREDQarbotech FAQ; dose-response literature
§9 · INDEX Every [ ] and ⟨ ⟩ on this spread is listed in §9 with the name of who can close it
7QB-DOC-0021 · REV B · 2026-08-04
SpecValueWhy
PageA4 · 210 × 297 mmThe buyer’s own paper. Not a deck ratio.
Columns20 + 34 + 6 + 130 + 20 = 210 mmA 34 mm evidence margin carrying every claim’s tag, source and figure number.
Baseline grid5.5 mmMain column runs to roughly 70 characters.
Masthead6 mm solid inkFindability without a colour. Identical through four reproduction generations.
Fore-edge block4 mm, stepped by sectionThe closed document indexes itself. Survives a scan; a coloured tab does not.
One-colour reductiontype 100% ink · panel 15% · control series 45° hatchNothing is lost, because nothing was carried in colour.
Mark, minimum45 mm — ⟨blocked⟩At 300 dpi the supplied 359 px file reproduces at 30.4 mm. There is no valid printed size for the full cut from the asset we hold.

What this does — and why it is the only one of the four that closes a sale

  1. It is the artefact the committee is allowed to keep. Evaluation by circulation is not permitted — the committee convenes, in a room, with paper. A deck is a performance that ends; a dossier is an object that stays in the file and gets cited in minutes. Numbered clauses exist for exactly that reason: a committee cites 3.2, not “the dosage bit”.
  2. The evidence margin is a claims control disguised as a layout decision. Every performance claim on the page must fill 34 mm beside itself with a tag, a source and a figure number. A claim whose margin is empty does not ship — which removes the argument from the copywriter and gives it to the grid. It is the same reason the deck is set type-identical to the document: a deck that can make a claim the document will not defend is how this company got 19 red claims in the first place.
  3. Section 9 is the section that wins it. Every [ ] and ⟨ ⟩ in the body is indexed at the back with the name of who can close it and by when. Twenty-six of the forty-two buyers in the file terminate at the same six-item artefact — a dose table, an SDS, a stated regulatory class, one control-plot trial with a named agronomist, a patent number that resolves, and a named owner of the yield claim. They are not competing asks; they are one ask with twenty-six customers. Publishing the gap, with a date and an owner against it, is the closest thing to the artefact itself that exists today — and no competitor in this category will do it.
  4. It survives the reproduction path the business actually uses. Printed one-colour, bound, photocopied, scanned. Type to 100% ink, panel fill to 15%, chart control series to a 45° hatch. Run it and check it, do not assume it: magick artefact.pdf -colorspace Gray -threshold 50% test.png — every token in this system sits outside the 20–55% band, so the reduction is lossless by construction rather than by luck.
  5. Bilingual at parity, with BM leading, because the buyer is a state department. Same size, same weight, same face; where one line must hold both, the condensed cut carries the 1.2× BM length instead of a smaller size. Every component on this spread is sized to the BM string, not the English one, so nothing reflows when the languages swap order.
  6. The held crop table is the whole strategy on one page. Three different absences get three different marks, because ambiguity in that table is the entire risk. It is designed now and shipped the day the client answers — and the C4 cell reads NO CLAIM, never “measured but weakest”, which would be publishing the exact number the claims discipline forbids.
The one blocker that outranks everything on this page

We do not hold vector artwork for the mark. At 300 dpi the supplied 359 px file reproduces at 30.4 mm against a 45 mm print minimum for the full cut, so there is no valid printed size for it on a tender cover from the asset we hold. Requesting the .ai / .eps / .svg master is the first action of the engagement, in every direction, and it is why the mark is set as type in every mockup in this file rather than drawn.

NOTES

What is not real here, stated plainly

SO NOTHING IN THIS FILE IS
MISTAKEN FOR AN APPROVED ASSET
ThingStatus in this fileWhat it will be
The typefaces System stacks approximating IBM Plex Sans, IBM Plex Mono and IBM Plex Sans Condensed. No webfont is loaded — this file is self-contained by requirement. The real Plex superfamily, self-hosted, SIL OFL 1.1 — chosen partly because a tender PDF is redistributed by a ministry and an open licence removes a question nobody wants raised in a procurement review. Note the one glyph that depends on it: the ⟨ ⟩ placeholder mark is Plex Mono’s; a system-font approximation may substitute a near glyph and lose a little of its distinction from an ordinary parenthesis.
The mark Set as type with a neutral geometric placeholder. The supplied lockup is not reproduced here, and neither is any platform logo or wordmark on any mock frame. The client’s own single-colour black lockup, unaltered, once the vector master exists. Clear space 1 C, never in a box or badge.
Every photograph A described grey block. Nothing in this file is generated, rendered or CGI — no fabricated micrograph, no invented instrument reading, no scale bar on anything an instrument did not produce. Shot, sourced or commissioned against the honesty line: the vial and the biochar shot in half a day; the proof photograph shot at the rig; any micrograph commissioned at UPM ION2, never generated.
Dates, names, numbers in ⟨ ⟩ Holes. Response times, staff names, WhatsApp numbers, the rig plot and the pack ladder are all client-supplied and none of them is invented. Filled by the client, or the surface does not publish.
The harvest date 2026-11-18 Illustrative, so the mechanic is visible. The real one is declared at planting, before anyone knows the answer. A real declared date on a real plot board, standing in the field.
Document references QB-DOC-0021, -0031, -0044 are illustrative of the numbering system, not a live register. A real controlled-document register with revision letters. It costs nothing and it gives the client the filing system they conspicuously do not have.
The crop table HELD. Drawn, marked not-for-publication, and gated. Ships the day the client answers the C4 question and the company’s own two websites stop disagreeing with each other.
Device chrome The phone bezels carry a corner radius. Nothing inside the brand system does — the register refuses corner radius above 0, and the message bubbles are square for that reason. Unchanged. The device is a drawing of somebody else’s hardware and is deliberately outside the system.

The claims check, run over every line of copy in this file

Nothing here carries a yield percentage other than the company’s own mean and peak — 11.57% and 45%, always with n = 300+, 500 ha, Malaysia + Indonesia, 2024, and always described as pilot data rather than a trial because no control was stated.

No crop is named in a works-on position beyond the sanctioned four-row form, and that form appears only inside a block marked HELD and not for publication. C4 crops are not mentioned in any published-facing line, in either direction — not as a claim, not as a refusal.

There is no disease, pest, virus or fungal claim anywhere, in any language. That is the one exposure here that is a Pesticides Act 1974 problem rather than a Trade Descriptions Act one, and it is the fastest way to lose the argument the entire regulatory position rests on.

Nothing says “safe with all fertilizers and pesticides” — no tank-mix trial exists anywhere, so the compatibility list is empty and the surfaces say nothing rather than something. Nothing says “reduce your fertiliser by X” — the most commercially attractive sentence available to this client and the most dangerous one in its stack; the mechanism runs the other way, and the surfaces say “you keep fertilising exactly as you do now” instead.

The programme record is stated only as fact of use: 2,000 ha, 1,319 farmers, RM320/ha, RM1.622 m, 1 October 2025, BERNAMA, MySTI STI202300224 — with the RM640,000 arithmetic disclosed in the same breath, no crest, no seal shape, and no percentage adjacent to a department’s name, because the endorsement offence can be committed by layout alone.

Pack size, price, concentration, spray volume and pre-harvest interval are unpublished everywhere they appear — shown as marked, sized holes, never invented and never silently dropped. No carbon or tCO₂e figure appears anywhere. No superlative, no “up to”, no “only”, no “world’s first”, no “guaranteed”, no organic or safety claim, no patent claim wider than a production method patented by Universiti Putra Malaysia, MY195075A, granted 9 January 2023. And no empty counter — there are three live on the company’s site right now, and a rendered zero beside a performance label is a false trade description on its face.

Three things a client decision unlocks, in order of cost

1 — The vector master of the lockup, .ai / .eps / .svg. Free, and it is a blocker on the one surface that decides the business. Nothing prints at tender size without it.

2 — Do HRTEM or FESEM micrographs of the actual QarboGrow particle already exist at UPM ION2? One email. If they exist it is a free proof asset from the founder’s own institute; if they do not, it is a service booking. Either way it is commissioned, never generated, and it is what fills the largest empty image slot in the whole system.

3 — The stock concentration and the dilution rate, per crop, at a stated spray volume. One answer closes the dose table, the entire dosing content stream, the drone question and roughly half the marked holes in this file. Today those holes are the honest move. Six months from now they are just holes.