5. Technical writing and online help improvements

5. Technical writing and online help improvements

This chapter follows a single writing process from research and scoping through release work to improvements a reader can see in the help. The early examples establish what I investigated and what a topic needed to answer; the later comparison shows how guidance, structure, and coverage changed. Where the record shows a team-wide change rather than my individual commit, that attribution remains explicit.

5.1 Establishing the documentation scope

My product-writing work involves determining what has changed, what users need to understand, and which existing topics should be updated. I use Jira and requirements to establish scope, inspect the relevant product or test environment, consult design material, and ask SMEs for details that are missing or uncertain. The resulting content must fit the existing help structure and review process. The next examples show how I investigated a documentation gap and shaped a proposed improvement; where a final topic is unavailable, I do not describe the research alone as a published change.

The DOC-3063 example shows the value of that scoping work. In August 2024, I reviewed a new SOTI MobiControl interface with Bob Kong, identified the affected topics, and described a smaller set of targeted changes instead of the larger reorganization originally anticipated. For a reader unfamiliar with the product, this meant updating the instructions where the interface had changed instead of reorganizing unrelated help.

5.2 Signal properties conditions and events

While working on DOC-5028 and MC-234922, I requested complete inventories of Signal properties, supported device families, system components, configurations, and events. Signal helps administrators monitor device conditions and act on them; incomplete lists would leave readers unsure which inputs and events the feature actually supports. My documentation improvement was to seek the missing scope before writing or revising guidance.

This research relates to the later help-improvement report's account of Signal content moving into a wizard-oriented structure with condition, action, schedule, property/event, example, upgrade, service, troubleshooting, and macro content. The broader restructuring is a team-wide comparison; my individual research and specific commits remain separately identified.

5.3 Device audit logging

For Device Audit Logging associated with MC-104549, I investigated design material and sought clarification on the location and composition of generated device logs. Readers need to know where diagnostic records appear and what each log contains; an inaccurate location would make a troubleshooting procedure unusable. When the available information was uncertain, I pursued another SME rather than relying on an unverified interpretation. The source identifies conversations with Yousef Said and Andrei Vesselkov in February 2026.

This example connects support-style investigation with writing: I identified the questions the guidance needed to answer. The certificate-deployment before-and-after shows a separate, directly inspectable example of turning technical detail into a guided procedure.

5.4 Declarative and reactive application policies

I sought SME clarification on how declarative and reactive app policies differ, including update behaviour, managed application configuration, connectivity conditions, retry behaviour, and activation predicates. I then checked a draft with the SME. The writing analysis identifies a February 2026 conversation with Youssef Mohamed.

These are distinctions that affect how an administrator interprets a policy, not simply alternative labels. In the master, the research belongs with the writing case study, while the help report's platform-specific App Policy structure belongs with the broader information-architecture comparison. The underlying conversations and draft should be attached before making a more detailed product-behaviour claim.

5.5 Apple enrollment documentation

My Apple enrollment review package reworked the overview, added a capability matrix, differentiated ADUE, ADDE, ADE, and related flows, and included a dedicated iOS ADDE topic. The cited review artifact is “links for review.docx.”

The reader-facing Apple help comparison shows the broader reorganization, while the Apple enrollment before-and-after isolates a specific committed overview change. These views answer different questions: how the section evolved, and what one selected revision actually changed.

The improvements report describes a management-type structure with device-based, user-based, and account-driven enrollment, macOS and tvOS organization, Apple Business Manager terminology, and supporting policy-management tasks. It records publication with a caveat that the SME requested time for another pass. That is a historical publication observation in the June report, not confirmation that all later review work was completed.

5.6 Feature research in the AI work history

The work-history report records additional feature investigations and documentation requests. Examples include MC-306121 for Microsoft outbound email/Office 365 SMTP, MC-331878 for feature scope and designs, MC-326798 for Enterprise Apps moving from Policies/App Policies into Device Resources, DOC-5424 for a trunk update with review constraints, MC-260379 for Testing/trunk comparison, and MC-184954 for feature and device-support research. It also records DOC-5601 for a targeted 2026.1.2-and-later runtime note. These are descriptions of historical work, not current product requirements.

The history associates DOC-5327 with eSIM-related research, while the improvements report's commit register also associates that key with Signal work. The ticket's actual scope should be checked before assigning a single project title. More broadly, the register includes initiated and exploratory threads, so a request to update content is not by itself evidence of a completed publication.

5.7 Release procedures and structured delivery

The career report describes the Technical Writing process as a sequence of Jira intake, prioritization, research, drafting, SME and peer review, publication, communication, and closure. Inputs include source materials, QA test plans, SME interviews, and sprint demonstrations. Outputs include updated help, reviews, approvals, and resolved tickets where closure is recorded.

I contributed an earlier version of the release-note procedure with Michael Pham. The described process uses a Jira request, a shared Word draft, Technical Writer review, manager peer review, and the subsequent publishing workflow. The procedure later evolved, so I retain my contribution without claiming authorship of every part of its current version.

The following sections show the reader-facing result of that writing work, with the Writing collection providing exact before-and-after source pairs.

5.8 Comparing the help from the reader perspective

SOTI MobiControl provides the case-study setting for this work. The transferable contribution is clearer procedures, better navigation, and more useful technical context. The audit separates my committed changes from the wider team's changes.

The earlier research and release examples explain how a writing task started. The before-and-after collection shows selected source revisions at the other end of that process. Together with the section-level comparison below, they make both the editorial reasoning and its visible result easier to follow.

The improvements report I prepared compares the committed SOTI MobiControl help available before January 1, 2024 with the committed trunk state in June 2026. Its focus is what changed for readers: findability, task guidance, platform coverage, explanatory completeness, and layout. The baseline is commit 09e2cb9f42ce6797505414c318e8ba9d6315230d from December 20, 2023; the current comparison state is ca9d87f1a5b3c07ff2e0654bb200ac5445ec4322 from June 12, 2026. The review period ends June 24, 2026.

The comparison excludes uncommitted edits. Mapped-topic counts come from ConsoleHelp.ditamap and exclude reltable entries. Device Details and Actions use a path-based scope because the work is spread across several topic areas. These definitions matter when interpreting the numbers. The report represents changes across multiple writers; the aggregate results are not solely my output.

Source: H-T01, independently reproduced from the same Git revisions and counting method. The media total sums <fig>, <image> and <object> elements; a figure wrapper containing one image contributes two. It is not a count of unique screenshots. These are content-structure measurements, not measurements of reader success, support-call reduction, or time saved. See the .

5.9 Profiles and the SSO content cluster

Profiles has the largest change footprint in the comparison. The current structure gives readers more ways to browse by operation, configuration category, platform, and payload/configuration area. Assignment options, access permissions, assignments, queue management, and stopping deployment sit closer to the main profile tasks. Platform coverage includes additional ChromeOS/macOS Wi-Fi, mobile private network, encrypted DNS, local users, Google Account, declarative profiles, and certificate-related content.

The SSO example makes the navigation change concrete. A single mapped Single Sign-on entry became a cluster of 26 topics covering Android, Imprivata, Apple/macOS, Extensible SSO, Kerberos, and Microsoft Authenticator workflows. The detailed table records which topics are new in the mapped structure and which were repositioned from earlier sections. I retain every row in the comparison archive so the cluster is supported by named pages and links.

Profiles remains an ongoing section-level effort in the report. Growth and reorganization should not be rewritten as a declaration that every profile topic is complete.

5.10 Changes verified in my commits

The Git audit compares each selected commit with its own parent, isolating what that commit changed. These examples show the development of my writing and review practice. Command counts include substeps, and image counts refer to image elements in the file.

The rendered comparison collection also includes my profile-assignment peer review and a new Apple dynamic-configuration task in Testing. The former preserves suggestions on a colleague's work; the latter is explicitly AI-labelled in its commit subject and has no before file. These demonstrate editorial judgement and AI-assisted authoring without implying accepted final text or production publication.

The comparison also covered Samsung application-firewall and manually uploaded-certificate examples. Its June endpoint counts are different from a single earlier draft commit. For example, FileVault has 14 commands and 5 image elements at the June endpoint, compared with 15 and 7 in the selected March draft. The certificate-payload overview later became the 11-command manually uploaded-certificate task at the same file path; those later counts must not be attributed to the May conversion alone.

The audit also traced all ten paths into later revisions. Michael Pham's immediate follow-up incorporates several of my profile-assignment review suggestions, including the direct opening, separate navigation step, consolidated filter guidance and related link. My May 2026 post-review revisions improve the Wi-Fi, FileVault and API examples; FileVault's application/profile wording is corrected. The records exact revisions, attribution and remaining issues, and the gallery includes four later-source views. Review activity does not by itself prove publication.

The audit resolved several misleading conversion labels. Three Windows Mobile pages and Android hotspot changed to task roots on 30 July 2025 and returned to reference roots in the next day's transformation-fix commit. They are excluded from lasting-conversion claims. Other rows combine historic detection with a different endpoint form; the row-by-row validation table records the actual roots and counts.

5.11 Global Settings

Global Settings changed mainly through grouping, terminology, and connections to nearby workflows. The report describes a flatter reference catalogue becoming easier to browse by console settings area, including Device Agent and Plugin, Enterprise Migration Certificate, App Store License Management, APNS, Automated Device Enrollment, API clients, Device Maintenance, Terms and Conditions, LDAP, and Google Workspace bindings.

Outbound email is an example of placing related help near a task. Notification content appears under Monitoring while Global Settings references support viewing connections and sending test email. The detailed records also include Microsoft Entra application and SMTP outbound connections, Apple root certificates, Chrome Enterprise bindings, Cloud Link, deployment-server settings, and other additions. The comparison archive retains all 42 topic-level rows.

5.12 Device details actions and updates

The Device Details and Actions area grew through targeted additions rather than a single root-node rebuild. The comparison includes device-action behaviour, details and search, update workflows, eSIM policy management, Windows OS images, the Updates dashboard, and rollback handling. Within the defined path-based scope, task topics increased from fourteen to thirty-two.

These help improvements connect to actual administrator tasks. The commit archive identifies my updates-dashboard and rollback contributions, including an April 2026 entry for DOC-4980.

5.13 Policies

App Policy content expanded from twenty-six mapped topics to fifty-four and was grouped by Android, Apple, Windows, Enterprise App Store, and management operations. Signal grew from eight mapped topics to twenty-two, with wizard steps and nearby property, event, example, upgrade, service, troubleshooting, and macro references. Compliance and Enrollment Policy sections gained permission and label-management support.

The comparison archive retains the forty-eight App/Signal detail rows and the commit records behind the broader work. My recorded contributions include declarative/reactive app management, Signal material, peer-review revisions, and policy-related updates. The content tree and the commit evidence are complementary: one describes the reader's route, and the other records who touched particular source files.

5.14 Apple enrollment

Apple enrollment received a larger revamp around management types and enrollment choices. The source describes a capability matrix, device-based and user-based flows, account-driven enrollment, macOS and tvOS organization, Apple Business Manager, Managed Apple Accounts, supervised enrollment with Apple Configurator, discovery-service support, and policy permissions and labels.

The report also uses a broader folder scan for Apple/enrollment-adjacent topics: ninety-four to 102 topics, 359 to 377 command steps, and fifty-one to 150 media markup elements. That scan has a different denominator from the mapped-policy numbers in the summary table and must remain separately labelled.

5.15 Certificates

Certificate help became more lifecycle-oriented in the comparison. It brings together manually uploaded certificates, dynamic deployment, certificate-authority integration, templates, Subject Alternative Names, dashboard visibility, and device-level management. The mapped scope increased from six topics to thirteen.

The certificate commit records include my work on DOC-3852, DOC-3856, and DOC-4854, with drafting, review requests, and incorporation of feedback recorded in the source. Other writers contributed to the same section, including dashboard and template changes. The master preserves those author distinctions.

5.16 Consolidation and the limits of the comparison

The removals table is limited to cases where the current structure supplies a clear consolidation or replacement explanation. It does not turn deleted paths into current deliverables. The comparison did not apply an NPI-only filter, and it treats appropriate reference topics as legitimate parts of the help.

The report says its generated current and 2024.0 URLs were checked with GET requests on June 24, 2026. That historical check is retained with its date; it is not a guarantee that every link remains available today. The archive preserves the full topic records, seven supplied visuals, and 140 section/commit rows. Repeated commits across sections are not counted as unique projects or unique outputs.