Try to answer one simple question about a project building on the Canton Network: what have they actually done? Open three tabs and you will get three answers. A directory lists them as Live. A thread calls their newest partnership a milestone. Their own documentation, read closely, says something quieter, with the word targeted doing a surprising amount of work. Now imagine your own money riding on which of those three tabs is telling the truth.
Most of crypto has a brutal but functional answer to this problem, and it is the price. Tokens trade, and however noisy the market gets, a live price at least forces every claim to survive contact with people who are paid to disagree with it. Canton is different in a way that matters. The most serious projects on it have no tokens at all, because they are building settlement rails, lending desks and trading venues for institutions, and many of them may never launch one. Which means there is no price, and nothing pushes back on the talk.
When nothing pushes back on the talk, talk wins. Announcements are free while verification is work, so attention flows toward whoever publishes most, and the teams doing the heaviest institutional building are often the quietest, sometimes because their counterparties sit behind confidentiality agreements and sometimes because they are simply busy. The public picture of the network and the reality of it drift apart, and nobody keeps score.
That was the situation I kept running into with my own capital deployed. I could not answer the most basic due-diligence question, what has this project verifiably done, from any existing source. So I built the scoreboard I needed and made it public. It is called the Canton Quadrant.
The idea, in two questions
Strip everything away and a serious reader wants two things from any pre-token project: what have you actually built that I can check, and how far have you gone toward launching a token? The Quadrant is simply those two questions turned into a chart.
The first question becomes the vertical axis. Every project is scored against seven yes-or-no checks: a live deployment on the real network, a named institutional counterparty, an official ecosystem listing, an infrastructure role such as running one of the validators that secure the network, a legal entity with named leadership, named backers, and an independent audit or a regulatory registration. Each check is worth exactly one point, and each is either publicly verifiable or it scores zero, which keeps my opinion out of the arithmetic entirely.
The second question becomes the horizontal axis, a five-step scale of token progress. Zero means no token statement of any kind. One means structural hints, the kind of architecture choices that usually precede an asset. Two means a rewards programme is live. Three means a token generation event, the industry's term for the moment a token actually launches, sits on a published roadmap. Four means the token is live and trading, at which point the project graduates off this chart entirely, because from that day forward the market can price it directly.
One rule holds the whole picture together, and it is the rule this entire publication runs on: if a team says it but I cannot verify it, it scores zero. The claim does not disappear, it becomes a ghost, a dashed outline showing where the project would sit if its words check out, so that ambition stays visible without ever being mistaken for fact. Claims move the ghost. Evidence moves the chip. Once that distinction is comfortable, you can read the entire instrument.
One issue publishes each month, at month end, and every issue is a permanent page, archived the day it goes out and never edited afterwards, with any mistakes corrected openly in the next one. Every chip links to a project page showing exactly which claims were checked, what passed, what did not, and the sources behind each call, so nothing on the chart asks to be trusted. The full methodology lives at /quadrant.
Where the idea comes from
The shape will look familiar, and it should. Anyone who has worked around enterprise software knows the Gartner Magic Quadrant, the two-axis chart that compresses a crowded vendor market into a single readable picture, and I kept that compression because it genuinely works. What I dropped were the parts that do not survive contact with crypto: axes built on analyst opinion, methodology behind a paywall, and scores that move in ways outsiders cannot audit. In a market this good at storytelling, an instrument you cannot check is just a better-dressed story.
The replacement discipline comes from my day job in data and AI governance for large organisations, where the core habit is sorting every claim into one of three buckets, confirmed, announced, or speculated, and never letting the buckets blur. The Quadrant is that habit turned into a chart: confirmed items score, announced items go dashed and score zero, and speculation does not make the page at all. Underneath it sit two older traditions. Credit analysts have run the right posture for a century, treating a rating as a view of the evidence today, revisited on a schedule, revised when the evidence changes, and never as a recommendation to buy anything. And if your work involves maturity models, the token axis will read naturally as one: it stages what a project has verifiably done about a token, not whether the project is good.
The last piece of the origin story is the uncomfortable one, so it goes on the record plainly. I hold positions in this ecosystem, disclosed on this site, and I built the instrument because I needed it. Then I made it public, because a private scorecard disciplines nobody, including its author.
From issue 01 to issue 02: the instrument improved
Issue 01, published on 14 July, planted the flag: the rubric in full, three projects placed, one reviewed and deliberately left off with the reason stated, and seventeen more named to a watchlist. Its most lasting contribution turned out to be procedural rather than analytical. The first serious correction this record carried was my own, a live-deployment claim I had accepted from a directory label when the project's own documentation was more careful, and it was fixed in public, dated, with the reason attached. Rather than damaging the instrument, that early test defined it, and every rule got stricter as a result.
Issue 02 is the improved instrument that discipline produced, and the improvement runs along three lines. Verification is deeper: every placed project is now re-checked against primary sources each issue, and whether a token already exists anywhere is established before anything else gets scored, which sounds obvious and catches more than you would think. Coverage is wider: eight projects placed against the founding three, drawn from a candidate pool of 47 that the record will keep working through monthly. And the vocabulary is richer: the chart now distinguishes a movement, meaning the world changed and new evidence confirmed it, from a correction, meaning my record was wrong, and gives both the same prominence, because a record that only shows the first kind is marketing.
What that improved instrument found is the reason this piece exists, so let me introduce you to the projects properly.
The quiet giants: Temple and Cashen
The two most substantiated projects on the network are ones you have probably heard the least about, and both are a pleasure to study. Temple is a 24/7 trading platform built for institutions, live on Canton's mainnet, the real network where value moves, as opposed to the test environments used for rehearsal. It runs its own validator, carries an independent security audit and a FinCEN money-services registration in public view, and is backed by twelve named investors behind a Temple Digital Group entity with named leadership. Cashen has built something Canton specifically makes possible: a marketplace where large Canton Coin holders lend out their locked positions, operating live with over 285 million CC in active locks, supported by BitGo, one of the most established custodians in digital assets, and backed by Ergonia, a Cumberland DRW company, which is about as traditional-markets a pedigree as this industry offers.
Each verifies six of the seven checks. And here is the detail that reframes the whole network: neither has published a single word about a token. No roadmap, no points, no hints. They are building revenue businesses on Canton's rails because the rails are the point, and for an investor that is a genuinely unusual signal. It means the strongest verified activity on this network is not currently investable through any ticker, which in turn means the standard crypto reflex, find the token, misses the real story entirely. The practical takeaway is a watch condition rather than a trade: a token statement from this cohort would be a structural event for the whole map, and the record will show it the month it happens.
The steady one: Alpend
Alpend, a money market built around ONE, a dollar instrument collateralized by Canton Coin, led issue 01 on substance and holds its placement in issue 02 at four of seven checks, audited, with a named entity behind it. On a chart built around change, unchanged is easy to underrate, so it deserves saying: a project whose claims survive a full re-verification pass without a single adjustment is giving you information too. Reliability of evidence is its own signal, and Alpend has now provided two issues of it.
The one running toward a token: Canborsa
Canborsa is the chart's outlier in the most literal sense, the only project sitting past step one on the token axis. It is an ambitious retail-facing trading venue, live in public beta, running a points programme that explicitly leads toward an airdrop, the practice of distributing free tokens to early users, with a token generation event on its published roadmap. That places it at step three with two of seven checks verified, and none of this is a criticism of ambition: building a consumer venue loudly and in public, while the institutional cohort builds quietly, is the harder marketing road, and Canborsa walks it openly.
For an investor, though, this profile is a different kind of object than a Temple or a Cashen, and the chart exists precisely to make that difference visible at a glance. A published token path tells you where a future liquidity event may emerge, which is exactly the kind of catalyst this publication tracks. A thin verified base underneath it tells you how much of the story still rests on claims rather than proof. The re-check also caught something no press release will ever announce: the launch timing language on Canborsa's own roadmap softened between issues, from a specific 2026 window to an open-ended 2026 plus. Drift like that is a data point, and it is the kind only a kept record can see. The watch condition here is the inverse of Temple's: claims converting into verified checks would be the signal that the loudest token path on Canton is growing the substance to carry it.
A distinction worth the price of admission: Cantex and the meaning of rewards
Issue 02 also produced my favourite teaching example so far, because it answers the fair question of whether an instrument like this can capture nuance or merely counts press releases. Cantex, a live exchange built by CaviarNine with two decades of investment-banking background between its named founders, pays traders rolling rebates in Canton Coin itself. Canborsa, as above, pays points that path toward its own future token. On the surface these are the same behaviour, rewards for activity. In meaning they are opposites: one is cashback denominated in the network's own asset, a design that aligns users with Canton itself and says nothing about a Cantex token, while the other is an explicit step on the road to a launch. The chart holds them three columns apart, Cantex at zero and Canborsa at three, and the reasoning sits on their project pages where anyone can audit it. That, in miniature, is what the token axis is for: not counting noise, but classifying what the noise actually commits a project to.
The graduate and the door left open
Two projects were reviewed in full and deliberately not charted, and both placements carry insight. Send, the network's dominant wallet and the creator of the CUSD stablecoin, turns out to already have a live token, $SEND, trading on another network and bridgeable to Canton, stated plainly on Send's own site. A live token is step four, and step four graduates a project off this chart, because the market can price it directly from that day forward. The scope lesson matters for readers: this instrument deliberately covers the pre-token frontier, the part of Canton no ticker can see, and hands projects over to the market the moment a ticker exists.
ACME Markets, reviewed and left off in issue 01, has since gained an official canton.network listing, its first verified check, while its live-product claim remains, for now, its own word, a beta behind a sign-in. So the record registers it honestly as closer, still off the chart, reason stated. Between hype and dismissal there is a third option, and it is simply keeping score. The door, as with every project here, stands open.
What the chart is saying
Put the vignettes together and the chart makes its first real statement. Seven of the eight placed projects sit at zero or one on the token axis, while the two strongest evidence profiles on the network have never mentioned a token, and the one project with a published token path carries one of the thinner verified bases, with a launch date that just softened. Proven substance and token noise, on Canton today, live at opposite ends of the chart, and if your picture of this network comes from announcement feeds, you have it roughly backwards.
I will not predict what happens next, because the chart cannot and neither can I. What the record gives you instead is a set of watch conditions with the history attached: whether that inverse pattern holds, tightens, or breaks, and which specific project breaks it. A quiet giant publishing a token roadmap, or the loud token path growing verified substance underneath it, would each be a structural event, and either will be visible here, sourced, the month it occurs. For anyone deciding whether to take Canton seriously, that is a far better question than which project is loudest this week, and the current answer is quietly encouraging: the network's substance is being built by teams that treat the rails as the point.
If you are building on Canton
This section is addressed to you directly, and it comes from genuine admiration for what this ecosystem is attempting, because the chart's neutrality can hide how much I enjoy watching it work. Most of the empty space on the chart is not absence of work. It is absence of verifiable evidence, and almost all of it is inexpensive to fix.
The most commonly missed check across ten assessments is also the simplest: a legal entity with named leadership. Real products, live deployments, and no human being's name anywhere on the public record. Institutional finance treats anonymous operators as a category error, and no amount of shipped code routes around it, so name your entity and your leadership. It costs a webpage. Beyond that, the pattern from the gaps I actually recorded: link your audit rather than claiming it, secure your official ecosystem listing, since a canton.network or Digital Asset directory entry is a primary source this instrument confirms on sight, put dates on claims, and keep live and targeted precise in your own materials, where precision protects you even when the ecosystem's directories run ahead of you.
Most of all, treat this record as the door it is meant to be. Every project page carries a standing offer: send relevant material and it appears there in your own words, clearly labelled as your claim until it verifies, at which point it becomes part of your score. Show me public evidence that any placement is wrong and the correction runs in the next issue with your name on the catch. A surface where verifiable work counts and announcements do not is, for a team that is actually building, the most favourable coverage format that exists, and I would much rather write next month's issue with your primary sources in hand than without them.
What this cannot tell you
Honest limits, stated plainly. The Quadrant measures public verifiability rather than quality, so a brilliant team in stealth scores badly here, and that is the instrument working as designed rather than a verdict on the team. Genuine institutional work behind confidentiality agreements gets undercounted, and I would rather undercount than take anyone's word for it. The chart does not judge whether a tokenless business model is good business, it cannot see private traction, and it will never say a word about price: mechanics and evidence, and you draw your own conclusions. That last part is not a disclaimer so much as the design.
Where it goes from here
The machinery is now built for the long run. Every score lives in one data file that draws the chart, every issue costs one archived page plus one line on the index, and every chip stands on a project page full of receipts. Issue 03 arrives at the end of August with all eight placed projects re-checked, two or three new assessments from the pool, and whatever movements the month produces. One issue is a snapshot; the archive is the asset.
The chart currently says Canton's substance is being built quietly, by teams that are not running token playbooks, and the record will show the moment that changes along with who changed it. So here is the open question this piece leaves you with, and it is one a real person can answer: which project on the watchlist has verifiable evidence I have underweighted? Send the source and it moves up the scoring queue.
Disclosure
I hold $CC and $CNTN. Alpend’s ONE asset is collateralized by CC. Cantex is an access programme partner of this site, disclosed on its project page, and its placement is scored on the same public rubric as every other project. Full positions are listed on this site. Nothing in this piece is a recommendation to buy or sell anything.
References and further reading
The Canton Quadrant methodology and issue index · Issue 01 · Issue 02 · The project registry · Zenith correction log · Send token page · ACME on canton.network · Halborn audit of Alpend