Ten registered experiments do not mean ten managed publishing sites

Direct answer and observation scope

A directory entry identifies an asset worth observing. It does not define who may publish to that asset, which outcomes belong to it, or which denominator belongs in a weekly report. Keep an inventory count, a domain grouping and an operational scope as separate records. This method is intended for experiment operators who maintain a network directory while the listed properties retain independent publishing owners.

On 2 November 2026, our authoritative inventory contained ten experiment properties. Eight used DatumHarbor subdomains and two used external domains. Our direct publication responsibility covered one site: the English DatumHarbor main site. The ten experiments were observation subjects, rather than ten additional sites under this publisher's writing authority. These are current inventory and authorization facts, not measured evidence of customer demand.

The case: three correct counts with different meanings

These counts need not agree. They answer different questions. Adding the main site to the ten experiments would create a different measure, while counting only the eight subdomains would omit two registered assets. Neither transformation should happen silently. A label such as network size is too ambiguous unless its population is stated beside it.

The two externally hosted assets were GSBI at case.shengyibiaoben.cn and China Business Demand Radar at signals.shengyibiaoben.cn. Their external hostnames did not remove them from the inventory. Equally, the eight properties on our subdomains did not become writable simply because their names shared a suffix. Our public allowlisted case dataset lists the ten public names, domains and classifications used in this count.

Define the record before changing the page

Start with the asset identity and its public destination. Preserve the actual URL supplied for the project. Next record why it belongs to the experiment network and what stage is supported by evidence. Then separately record the current operator's permitted actions: observe, update the main directory, or publish to a specific destination. These fields serve different decisions and should not be inferred from one another.

For this case, a directory addition belongs to the main site's content workflow. A change to an independent project's article belongs to that project's workflow. Keeping those actions separate prevents a directory maintenance request from turning into a cross-project deployment. It also keeps the network visible without requiring a migration or a common backend.

This is an operational method derived from our recorded boundary. It is not a claim that every organization must use the same ownership model, and it does not establish legal ownership of a domain. Teams with a different mandate should write their own explicit publication scope before using a shared reporting table.

A reconciliation procedure operators can reuse

1. List the public assets from the existing authoritative inventory, using stable asset identifiers rather than a fresh manually maintained list.
2. Check each listed URL against the intended public destination. Classify the hostname separately from the asset's membership.
3. Write the managed publication-site list from current authorization. Do not derive it by filtering for a shared suffix.
4. Declare the unit of every summary: registered assets, managed sites, available pages or accepted new articles.
5. Compare each dashboard number with the matching population. Investigate a mismatch before changing either the inventory or the counter.
6. Leave unsupported performance fields UNKNOWN. A complete directory supplies neither a traffic measurement nor a sales register.

Run the reconciliation before a reporting period starts and after an authorized inventory change. Preserve a dated snapshot so that a later reviewer can explain which assets were included. A membership change should update the relevant directory count; it should not rewrite an earlier period's population without an explicit restatement.

Applying a publication target without inflating the network

In our current workflow the weekly target applies to the one directly managed main site. The ten observed experiment properties are not ten extra publishing obligations for this operator. Their own owners may have separate targets, but this case contains no verification of those targets or their completion.

The main-site count also excludes posts that happened to appear on observed properties. Linking to a new child article can be useful directory maintenance, but it is not a new original main-site article. The distinction protects both sides: observed activity remains visible, while the publisher does not take credit for work performed elsewhere.

For a hypothetical network that explicitly delegates publishing on several sites, maintain one progress row per authorized site and count accepted URLs against that row. This hypothetical extension is not a description of our current permissions. If the managed set changes, record the effective date and preserve the earlier boundary.

When to stop and request a narrower decision

Stop a proposed cross-site write if the only justification is domain similarity or network membership. Confirm the responsible project and the exact destination before changing source or deploying. Continue independent main-site work when it remains authorized; one uncertain child boundary need not prevent an accurate directory description.

Do not resolve an inconvenient external URL by moving the project. Hosting changes introduce a different task, with their own consequences and authorization. Here, preserving the two external destinations was part of accurate registration. No URL migration was needed to explain the experiment network.

Frequently asked questions

Does an external domain make an experiment less central to the network?

Not in this inventory. Membership follows the registered asset record. Hosting is a separate attribute. The dataset shows two external properties included in the same ten-asset population, without asserting that they share a publishing backend or measurement system.

Can an availability check update an experiment's stage?

An HTTP response supports a technical availability observation at a stated time. A stage representing demand or commercial progress needs evidence relevant to that decision. The hostname classification and directory membership cannot fill that gap.

Should all ten sites appear in the main site's article quota table?

They can appear in an observation appendix if clearly labeled, but our publication acceptance table should use the directly managed site population. Mixing observation subjects into that table would imply a delivery responsibility the current mandate does not provide.

What happens when a listed site closes?

Record the evidence and the decision using the existing lifecycle process. An unreachable request alone may be transient and should not silently delete an asset. The historical inventory remains useful for understanding earlier reports even after a supported closure.

Is this case evidence that external hosting improves results?

No. We did not compare acquisition, conversion, cost or revenue between hosting groups. The case verifies classification and operational scope only. Any performance comparison would need its own matched measures and observation period.

Conclusion

Ten registered experiments, eight subdomain properties and one directly managed publication site are compatible facts when their units are explicit. Preserve the real destinations, keep authority separate from membership, and use the correct population for each decision. The resulting directory is an honest map of the experiments rather than an accidental expansion of the publisher's mandate.

References and observation scope

Inventory scope case, 2 November 2026: ten public inventory entries and the current main-site publication boundary. See the experiment directory and DatumHarbor's role. Search and commercial effects are UNKNOWN.