---
title: "I Read the Domains Before I Size the Budget"
description: "I size an AI visibility budget from the domains in the top ten, because in every measured category the largest block of those slots has no source class."
canonical: "https://jaxonparrott.com/blog/founder-decision-ledger-unclassified-top-ten-slots-2026"
date: "2026-09-27"
last-updated: "2026-11-04"
---

# I Read the Domains Before I Size the Budget

I stop sizing an AI visibility budget by source class. In all 14 categories the Machine Relations Index measures across every buyer question, the largest single block of the top ten belongs to no class at all. I read the domains in the six leaderboards before I fund a program built around a class.

The class-capture study those leaderboards belong to is [Four Categories One Source Class Wins; Nine Where None Does](https://machinerelations.ai/research/source-class-capture-by-category-2026). That page is the deeper dataset. This essay is how I would use it as a founder: the order I read a category, the spend I would still make, and the classifier change that would put the class table back in front.

## What I measured

Release `mri_score_v2.0+2026-09-26+2b779408cfda`, window 2026-05-10 to 2026-09-26. 16,925 answer runs across six engines, 1,010 monitored prompts, 23,978 cited domains, 132,514 citation events.

The Index splits buyer demand into six question shapes and scores every category-shape pair separately, publishing a leaderboard only once that pair clears 10 observed runs across 7 distinct run dates. Fourteen categories currently publish all six. That is 84 leaderboards. I took the top ten of each, 840 slots, and counted the source role assigned to the domain holding each slot.

I used the top ten rather than the whole leaderboard on purpose. A buyer reads an answer, not a distribution. If a class matters to a buying decision, it has to be in the visible part. HR and Talent publishes only four of its six shapes and is excluded. Every number below is read from the segment's own public page.

| Category | Unclassified | Editorial | Vendor-owned | Community |
|---|---:|---:|---:|---:|
| AI Infrastructure | 19 | 9 | 8 | 12 |
| Cybersecurity | 23 | 4 | 19 | 8 |
| Consumer Finance | 27 | 16 | 5 | 6 |
| AI Security and Privacy | 29 | 7 | 6 | 8 |
| AI Visibility and GEO | 31 | 4 | 8 | 10 |
| Enterprise Software | 31 | 4 | 11 | 4 |
| Deep Tech and Hardware | 34 | 4 | 1 | 10 |
| Consumer Health | 35 | 7 | 0 | 5 |
| Consumer Products | 38 | 10 | 0 | 5 |
| Family Software | 38 | 15 | 0 | 6 |
| Fintech | 39 | 3 | 12 | 3 |
| Emergent Prosumer | 40 | 10 | 0 | 7 |
| Education and Training | 45 | 2 | 1 | 8 |
| Healthcare Services | 46 | 8 | 2 | 0 |
| All 840 slots | 475 | 103 | 73 | 92 |

56.5% of the visible board carries the label the Index calls "Other observed source," which means the classifier has not filed it under any of the eight named types. Editorial publications hold 12.3%. Vendor-owned sources hold 8.7%. Community and social platforms hold 11.0%. The remaining 11.5% is academic and government, market databases, search platforms, and analyst research combined.

Unclassified is the plurality in 14 of 14 categories. Only two categories, AI Infrastructure (41 of 60) and Cybersecurity (37 of 60), have named classes that add up to a comfortable majority of the top ten, and in ten of the fourteen the named classes together are a minority. Editorial, vendor-owned, and community slots, the three a founder can pitch, publish, or show up in, hold 268 of the 840 between them. That is the addressable fraction of the visible board, and it is not the board.

## What the 475 is, and what it is not

It is not news that a catch-all is large in aggregate. Across the whole index, 20,323 of 23,978 cited domains carry no source role and together hold 72,589 domain-level cited runs, more than the entire editorial class's 13,993. Most of those domains are long tail, cited a handful of times, below the evidence floor. No taxonomy will reach them all. The tail being unlabeled costs a founder nothing.

This cut is the top. Every slot in the table is above the evidence floor, inside a published leaderboard, in the ten positions a buyer sees. "The tail is unclassified" and "the top is unclassified" are different facts. I ran the second because the first was being used as a comfort.

Unclassified is a state of the classifier, not a property of the domain. Some of that 56.5% is a competitor's own marketing pages, which are reachable by writing a better page and not by buying a placement. Some of it is a trade blog with a small audience and a sitemap. Telling those apart is manual work. It is the work.

Class share still decides whether a class program is worth anything. It cannot tell me the ceiling. Funding a press, research, or community program sized to a board it reaches a fifth of is the cost of skipping the domain list.

## The category that forced the second read

Education and Training holds 45 of its 60 top-ten slots in the catch-all, and gives editorial publications 2.

The visible answer to the shopping question in that category, the segment called Best tools, was measured on 106 runs across 7 dates out of 263 ranked domains:

| Rank | Domain | Source role | Citation rate |
|---:|---|---|---:|
| 1 | Reddit | Community and social platform | 33.96% |
| 2 | D2L | Other observed source | 18.87% |
| 3 | YouTube | Search or media platform | 16.04% |
| 4 | CYPHER Learning | Other observed source | 15.09% |
| 5 | Valamis | Other observed source | 15.09% |
| 6 | BestColleges.com | Other observed source | 14.15% |
| 7 | Coggno | Other observed source | 14.15% |
| 8 | TalentLMS | Other observed source | 14.15% |
| 9 | Docebo | Other observed source | 13.21% |
| 10 | PrepScholar | Other observed source | 12.26% |

Six of those ten are learning-platform companies writing on their own domains, filed under no class. The first domain the Index recognizes as an editorial publication is at rank 12. Forbes is at 29 with 6.60%. Gartner is at 58 with 3.77%. G2 is at 23.

A founder who reads a class table sees a market with no press and concludes there is nothing to buy. A founder who reads the domain list sees six competitors' websites in the top ten of the question buyers ask first. Those are different briefings. Only the domain list tells me what to do on Monday.

One more detail a class table erases. In Education and Training head-to-head comparisons, the most-cited source is University of the People at 28.00%, filed as an academic and government source, in a category where it competes commercially. The label is right by the taxonomy and useless as a buying signal.

## Where a class budget still earns its line

The spread in the table runs from 19 to 46 unclassified slots out of 60.

Cybersecurity gives vendor-owned sources 19 of 60 top-ten slots, the highest own-domain payoff of any category measured, and its 23 unclassified slots are the second-smallest catch-all, behind AI Infrastructure's 19. If I sold security software, I would treat documentation as a citation asset with this evidence behind it.

Consumer Finance gives editorial publications 16 of 60, and Family Software gives them 15. In those two categories a press program is the highest-leverage line in the visibility budget, because a quarter of the visible answer is held by publications that can be pitched.

AI Infrastructure is the most balanced board in the index: 19 unclassified, 12 community, 9 editorial, 8 vendor-owned. Every class pays something. No class pays enough alone.

Fintech gives editorial 3 of 60 and vendor-owned 12. The correct budget is close to the inverse of Consumer Finance, even though the two markets sit next to each other. Sizing one from the other is the mistake this count exists to stop.

## Three budgets I would refuse to copy across

The spread from 19 to 46 unclassified slots out of 60 is the whole point of reading category by category. Three of them are enough to stop a single company-wide line item.

Healthcare Services holds 46 unclassified, 8 editorial, 2 vendor-owned, and 0 community. Almost the entire visible board is the catch-all, and the community column is empty. I would not arrive with a forum plan or a large press plan. I would open the sixty domains and say what they are. If they are hospital systems, insurers, and government pages the classifier has not filed, the move is a page that answers the buyer's question better than those institutions do, plus the 8 editorial slots that can actually be pitched. A press budget sized to Consumer Finance's 16 editorial slots would be buying inventory this category does not show.

Consumer Finance is the other end of the press question: 27 unclassified, 16 editorial, 5 vendor-owned, 6 community. Sixteen of 60 is the strongest editorial payoff in the table besides Family Software's 15. I would fund a serious pitch list there, and I would still read the domains, because 27 unclassified slots remain the plurality. Family Software, 38 unclassified, 15 editorial, 0 vendor-owned, 6 community, says the company's own site is not what this release is citing into the top ten, and publications are. Copying a cybersecurity "publish the docs" plan into family software would aim at a column that is zero.

Cybersecurity is that docs plan, and it is not a general rule. Vendor-owned sources hold 19 of 60, editorial only 4, community 8, unclassified 23, the second-smallest catch-all in the fourteen. If I sold security software I would treat our own documentation as a citation asset and I would keep the press line small until the 4 editorial slots told me otherwise. Fintech, next door in the way founders talk about it and not next door in this table, is 39 unclassified, 3 editorial, 12 vendor-owned, 3 community. Vendor-owned matters. Editorial barely does. A fintech team that copies Consumer Finance's press budget is funding the line this category does not pay, and skipping the 12 vendor-owned slots it does.

AI Visibility and GEO, 31 unclassified, 4 editorial, 8 vendor-owned, 10 community, is a third shape. Community is larger than editorial. A program that only pitches publications is working 4 of 60 slots. Enterprise Software is similar on editorial, 4, with vendor-owned at 11 and community at 4. Deep Tech and Hardware is the extreme own-domain miss: vendor-owned holds 1 of 60, community 10, editorial 4, unclassified 34. The papers and the trade infrastructure live in that unclassified and academic remainder, which is why the class table alone would tell a hardware company to do almost nothing.

I want the sixty names. The columns tell me which program is even eligible. They do not tell me the Monday list.

## The decision

I open the category's six leaderboards and read the domains before the class column. I name every domain in the top ten out loud: a competitor, a trade publication, a forum, a database listing, a university. That list is the board.

Then I ask what each one costs to reach. A publication has a price and a pitch. A competitor's own page has no price and no pitch, and the move against it is a better page that answers the same question. A forum cannot be bought. That is a fact about the forum. It is not a reason to pretend the forum is absent when it is rank 1.

Then I size class programs against the slots they can reach in that category. In Consumer Finance the number justifies a serious press budget. In Education and Training it justifies almost none, and the same money goes to displacing six vendor pages.

Christian Lehman, Chief Growth Officer, has the publication half of this. On the September 22 release, of 1,238 publications AI engines cite, 47 hold a top-ten position on any buyer question and 35 of those hold exactly one. His bound says a pitch list has to be built per question. Mine says how much of the question a pitch list can reach at all. Both have to be true before a budget is real.

The scope is the 840 top-ten slots in the 14 categories that publish all six shapes on this release. It does not say AI visibility cannot be bought. The purchasable fraction is a category-specific number between roughly a fifth and two thirds of the visible board, and I can read it before I commit. It does not say unclassified means low quality. Healthline is the 15th most-cited domain in the index and carries no source role. The label records how far classification has reached, not how much an engine trusts the source.

I would put the class table back in front if the classifier's coverage of the top ten improved enough that the catch-all fell below a third of the visible board. That is an engineering problem and I expect the number to move. The Index republishes on a rolling window, so individual slots move. The category spread has been stable across the releases I have read, and the spread is what a budget should be sized against. Until the catch-all drops below that third, the domain list is the primary read and the class table is the second.

The discipline under the read is [Machine Relations](/blog/machine-relations): which sources a machine reaches for when a buyer asks, not how a page ranks in a list built for people. The class labels I am refusing to budget from are defined in the [glossary](https://machinerelations.ai/glossary). AuthorityTech is where I do this work, and a practice that cannot say when its own line is the wrong first move is not worth hiring.

## FAQ

### Does this mean AI visibility cannot be bought?

It means the purchasable fraction is a category-specific number. Editorial, vendor-owned, and community slots together are 268 of 840 in this cut, 103 plus 73 plus 92, and the share inside one category runs from a market where press holds 16 of 60 to a market where press holds 2. I read the 60 before I commit the budget.

### Why the top ten instead of the whole leaderboard?

A buyer reads an answer, and an answer is short. A domain at rank 58 of 263 with a 3.77% citation rate, Gartner's place in the Education and Training best-tools list, is not in the room where that decision happens. The tail is real. It is the wrong place to size a program.

### Is "unclassified" the same as low quality?

No. The label is the classifier's reach. Healthline carries no source role and is the 15th most-cited domain in the index. Six learning platforms in an Education and Training top ten are unclassified because they are company sites the taxonomy has not filed, and they are the most actionable rows on the board.

### How often should I resize the budget off this?

The Index republishes on a rolling window, so slots move. I would resize when the category spread moves, not when one domain swaps places inside a top ten. The reversal I am watching is the catch-all falling below a third of the visible board. Until that release, a class-first budget is sized to a minority of the answer.

### What is the single number I want from my own team?

In our category, how many of the 60 visible slots are held by something we can actually buy or displace. If nobody can answer that from the six leaderboards, no visibility budget in the company is sized correctly.

---

*Published September 27, 2026*
By [Jaxon Parrott](https://jaxonparrott.com) — Founder & CEO, [AuthorityTech](https://authoritytech.io)
Updated 2026-11-04

Part of [Machine Relations](https://jaxonparrott.com/machine-relations) · [Glossary](https://jaxonparrott.com/glossary) · [Founder judgment](https://jaxonparrott.com/topics/founder-judgment)


Tags: AI Citations, Founder Judgment, Measurement

## Continue reading

- [The Category Ownership Trap](https://jaxonparrott.com/blog/category-ownership-trap): Owning a category definition does not mean owning the buyer journey. How I connect Machine Relations to the questions people ask before they know its name.
- [I Count the Research Record Before I Fund It](https://jaxonparrott.com/blog/founder-decision-ledger-academic-sources-category-2026): Before I fund a research program for AI visibility, I count whether academic and government sources show up in that category's answers at all.
- [Do Not Hire an AI Visibility Vendor on Rank Alone](https://jaxonparrott.com/blog/do-not-hire-ai-visibility-vendor-rank-alone): I would not hire an AI visibility vendor on a rank screenshot. In the buyer-choice segment, 399 of 699 cited domains show up in a single run.

---

*Machine-readable version of [https://jaxonparrott.com/blog/founder-decision-ledger-unclassified-top-ten-slots-2026](https://jaxonparrott.com/blog/founder-decision-ledger-unclassified-top-ten-slots-2026)*