AI Work Collaboration Platforms in AI Search: The Biggest AI Visibility Problems and How We'd Fix Them

Companies covered: Coda, Miro, Asana, Slack, Wrike, ClickUp, Discord, Airtable, Atlassian, Productboard, Teamwork.com, Cisco Webex App

Answer Capsule

The November 2026 AI Work Collaboration Platforms benchmark compares 12 companies across 580 AI answers that met its inclusion rules. Asana recorded 552 mentions and 226 recommendations (38.97% of the answers). Being mentioned means appearing in an answer; being recommended means receiving an affirmative endorsement.

We examine each company, compare the available monthly results, and set out a five-step plan for improving the information AI systems can draw on. The work starts with measuring visibility and checking conflicting claims, then moves to company pages, relevant publishers and buyer questions. The cited pages need review before any correction request, and the research does not establish what caused recommendation differences or promise future gains.

Executive industry diagnosis

Questions This Section Answers

• What does the AI Work Collaboration Platforms benchmark show about mentions, recommendations and placement?

The peer set splits into two tiers. Asana and ClickUp appear in nearly every answer and are recommended in roughly two of every five, so they function as default options. Wrike and Slack sit in a middle band with strong reach but far fewer endorsements. The rest appear less often and are recommended less often still.

The gap between being mentioned and being recommended is the central pattern. Several companies are named frequently yet rarely endorsed with positive wording, which suggests they are treated as reference points rather than choices. That distinction matters commercially: appearing in an answer is not the same as being put forward as a solution.

Placement reinforces the tiering. The leading pair also dominate shortlist and first-place positions, while lower-ranked companies cluster in neutral mentions. We would build a scorecard that tracks reach, endorsement and position separately, so the client can see which lever each company actually needs.

What changed since last month

Questions This Section Answers

• What changed across the usable monthly AI Work Collaboration Platforms snapshots?

The monthly samples are separate selections, and the answer totals differ between them. November's included answers are slightly fewer than October's, but the recorded recommendation counts moved far more sharply than that difference would suggest. We treat the two months as distinct snapshots rather than a like-for-like trend.

Several companies show the same shape: mentions held or rose while recommendations fell. Others moved in both directions at once. Because the samples are not matched, we cannot say why any count or rate changed, and we will not present the movement as performance.

What the history does support is a comparability record. We would document how each month's sample was assembled so future comparisons are honest about what changed and what did not.

Company-by-company findings

Coda

Questions This Section Answers

• What measured visibility pattern and review priority does Coda have in AI Work Collaboration Platforms?

Coda recorded 27 mentions (4.66%) and 2 recommendations (0.34%) across 580 answers included in the benchmark. Top 3 placements numbered 1; first-place recommendations numbered 1. Framing among its mentions was 3 positive, 24 neutral and 0 negative.

From October 2026 to November 2026, Coda's mention count rose from 23 to 27, while its recommendation count fell from 13 to 2. The recommendation rate was 2.23% and 0.34% respectively.

AI answers contained 2 conflicting claims about Coda: 0 critical, 0 high, 2 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://coda.io/.

Selected third-party review candidate: https://polarishq.co/compare/notion-vs-coda.

Coda is named occasionally but almost never endorsed, and its few mentions are mostly neutral. The recorded ownership conflict is the practical starting point: the two claims cannot both be current, and the answer depends on entity and date. We would verify the company's status against primary sources before any correction is considered, then brief the owned page and the comparison page only if the facts support it.

Miro

Questions This Section Answers

• What measured visibility pattern and review priority does Miro have in AI Work Collaboration Platforms?

Miro recorded 178 mentions (30.69%) and 59 recommendations (10.17%) across 580 answers included in the benchmark. Top 3 placements numbered 18; first-place recommendations numbered 9. Framing among its mentions was 64 positive, 114 neutral and 0 negative.

From October 2026 to November 2026, Miro's mention count fell from 243 to 178, while its recommendation count fell from 172 to 59. The recommendation rate was 29.55% and 10.17% respectively.

AI answers contained 0 conflicting claims about Miro: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://miro.com/diagramming/workflow-design/.

Selected third-party review candidate: https://aipmtools.org/comparisons/miro-vs-mural.

Miro appears often in the benchmark answers and is framed positively far more than negatively, but recommendations form a much smaller share of its mentions than reach alone would suggest. The two selected months show different mention and recommendation totals; treat them as separate snapshots whose comparability is limited, not as a trend.

With no conflicting claims logged, the practical next step is verification rather than correction. We would draft a page brief for the workflow-design page and fact-check the third-party comparison source against current primary evidence before any outreach, keeping any ambiguous matches out of the queue.

Asana

Questions This Section Answers

• What measured visibility pattern and review priority does Asana have in AI Work Collaboration Platforms?

Asana recorded 552 mentions (95.17%) and 226 recommendations (38.97%) across 580 answers included in the benchmark. Top 3 placements numbered 129; first-place recommendations numbered 49. Framing among its mentions was 239 positive, 313 neutral and 0 negative.

From October 2026 to November 2026, Asana's mention count rose from 545 to 552, while its recommendation count fell from 427 to 226. The recommendation rate was 73.37% and 38.97% respectively.

AI answers contained 4 conflicting claims about Asana: 0 critical, 3 high, 1 medium, 0 low. Of these, 3 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://asana.com/resources/best-project-management-software.

Selected third-party review candidate: https://zapier.com/blog/free-project-management-software/.

Asana is the most visible company in the set and is endorsed far more often than most peers, though endorsements remain a minority of its many mentions. Its disputed claims concern free-plan limits, where answers give conflicting figures. We would verify the current plan terms from primary sources, then prepare a fact-check note and a page brief so the owned resource states limits precisely.

Slack

Questions This Section Answers

• What measured visibility pattern and review priority does Slack have in AI Work Collaboration Platforms?

Slack recorded 333 mentions (57.41%) and 71 recommendations (12.24%) across 580 answers included in the benchmark. Top 3 placements numbered 36; first-place recommendations numbered 31. Framing among its mentions was 77 positive, 256 neutral and 0 negative.

From October 2026 to November 2026, Slack's mention count fell from 382 to 333, while its recommendation count fell from 177 to 71. The recommendation rate was 30.41% and 12.24% respectively.

AI answers contained 2 conflicting claims about Slack: 0 critical, 0 high, 1 medium, 1 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://slack.com/blog/collaboration/best-collaboration-tools.

Selected third-party review candidate: https://guideflow.com/blog/team-collaboration-software-tools.

Slack is named often but endorsed in a much smaller share of those mentions, and most of its framing is neutral. Two low-to-medium conflicts concern free-plan history and starting price, both of which are checkable facts. We would confirm current terms from primary sources, then brief the collaboration-tools page and review the third-party guide before deciding whether a request is justified.

Wrike

Questions This Section Answers

• What measured visibility pattern and review priority does Wrike have in AI Work Collaboration Platforms?

Wrike recorded 299 mentions (51.55%) and 122 recommendations (21.03%) across 580 answers included in the benchmark. Top 3 placements numbered 34; first-place recommendations numbered 13. Framing among its mentions was 128 positive, 171 neutral and 0 negative.

From October 2026 to November 2026, Wrike's mention count rose from 286 to 299, while its recommendation count fell from 228 to 122. The recommendation rate was 39.18% and 21.03% respectively.

AI answers contained 0 conflicting claims about Wrike: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://wrike.com/project-management-guide/faq/what-are-project-management-tools/.

Selected third-party review candidate: https://thedigitalprojectmanager.com/tools/best-wrike-alternative/.

Wrike sits in the middle of the peer set: frequent mentions, moderate endorsements, and a modest share of positive framing. Its recorded counts moved in opposite directions between the two selected months, which we treat as a sample comparability limit. No disputed claims were logged, so the work is a page brief for the project-management FAQ and a review of the alternatives source.

ClickUp

Questions This Section Answers

• What measured visibility pattern and review priority does ClickUp have in AI Work Collaboration Platforms?

ClickUp recorded 518 mentions (89.31%) and 223 recommendations (38.45%) across 580 answers included in the benchmark. Top 3 placements numbered 134; first-place recommendations numbered 56. Framing among its mentions was 238 positive, 279 neutral and 1 negative.

From October 2026 to November 2026, ClickUp's mention count rose from 490 to 518, while its recommendation count fell from 419 to 223. The recommendation rate was 71.99% and 38.45% respectively.

AI answers contained 1 conflicting claim about ClickUp: 0 critical, 0 high, 1 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://clickup.com/blog/what-is-clickup-used-for/.

Selected third-party review candidate: https://zapier.com/blog/free-project-management-software/.

ClickUp is one of the two most visible companies and is endorsed in a large share of its mentions, with almost all framing positive or neutral. One medium conflict concerns whether native time tracking is included on the free plan. We would verify that against current plan documentation, then brief the owned explainer page and check the free-software roundup before any request.

Discord

Questions This Section Answers

• What measured visibility pattern and review priority does Discord have in AI Work Collaboration Platforms?

Discord recorded 42 mentions (7.24%) and 4 recommendations (0.69%) across 580 answers included in the benchmark. Top 3 placements numbered 2; first-place recommendations numbered 0. Framing among its mentions was 4 positive, 38 neutral and 0 negative.

From October 2026 to November 2026, Discord's mention count rose from 36 to 42, while its recommendation count fell from 20 to 4. The recommendation rate was 3.44% and 0.69% respectively.

AI answers contained 0 conflicting claims about Discord: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://discord.com/press-releases/discord-launches-orbs-globally.

Selected third-party review candidate: https://microsoft.com/en-us/microsoft-teams/collaboration.

Discord is mentioned regularly but endorsed rarely, and nearly all of its framing is neutral, so it reads as a reference point rather than a recommended choice. Its recorded counts moved in opposite directions across the two selected months, which we treat as a comparability limit. With no disputed claims, we would brief the owned press page and review the Teams collaboration source before proposing anything.

Airtable

Questions This Section Answers

• What measured visibility pattern and review priority does Airtable have in AI Work Collaboration Platforms?

Airtable recorded 180 mentions (31.03%) and 54 recommendations (9.31%) across 580 answers included in the benchmark. Top 3 placements numbered 15; first-place recommendations numbered 6. Framing among its mentions was 56 positive, 124 neutral and 0 negative.

From October 2026 to November 2026, Airtable's mention count rose from 123 to 180, while its recommendation count fell from 89 to 54. The recommendation rate was 15.29% and 9.31% respectively.

AI answers contained 0 conflicting claims about Airtable: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://airtable.com/articles/project-collaboration-tools.

Selected third-party review candidate: https://reddit.com/r/Airtable/comments/17lto9a/is_airtable_worth_it_any_alternatives/.

Airtable's mentions rose while its endorsements fell between the two selected months, and most of its framing is neutral. Because the samples differ, we read that as an observed movement rather than a trend. No disputed claims were logged. The useful work is a page brief for the project-collaboration article and a careful review of the community discussion source, which is user commentary rather than an editorial claim.

Atlassian

Questions This Section Answers

• What measured visibility pattern and review priority does Atlassian have in AI Work Collaboration Platforms?

Atlassian recorded 172 mentions (29.66%) and 27 recommendations (4.66%) across 580 answers included in the benchmark. Top 3 placements numbered 10; first-place recommendations numbered 3. Framing among its mentions was 29 positive, 143 neutral and 0 negative.

From October 2026 to November 2026, Atlassian's mention count rose from 171 to 172, while its recommendation count fell from 98 to 27. The recommendation rate was 16.84% and 4.66% respectively.

AI answers contained 0 conflicting claims about Atlassian: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://atlassian.com/project-management-software.

Selected third-party review candidate: https://bstrategyhub.com/atlassian-competitors-and-alternatives/.

Atlassian is named about as often as several mid-tier peers but endorsed in a small fraction of those mentions, with framing overwhelmingly neutral. Its recorded counts show mentions holding steady while endorsements fell, which we treat as a comparability limit rather than a cause. No conflicts were logged. We would brief the project-management page and review the competitors source before any request.

Productboard

Questions This Section Answers

• What measured visibility pattern and review priority does Productboard have in AI Work Collaboration Platforms?

Productboard recorded 21 mentions (3.62%) and 7 recommendations (1.21%) across 580 answers included in the benchmark. Top 3 placements numbered 5; first-place recommendations numbered 3. Framing among its mentions was 8 positive, 13 neutral and 0 negative.

From October 2026 to November 2026, Productboard's mention count fell from 32 to 21, while its recommendation count fell from 23 to 7. The recommendation rate was 3.95% and 1.21% respectively.

AI answers contained 0 conflicting claims about Productboard: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://productboard.com/.

Selected third-party review candidate: https://productplan.com/blog/best-product-roadmap-software.

Productboard appears in a small number of answers but is endorsed in a reasonable share of them, and its framing is mostly positive or neutral. Its recorded counts fell between the two selected months, which we treat as a sample comparability limit. No disputed claims were logged. We would brief the owned homepage and review the roadmap-software roundup, focusing on how product-management use cases are described.

Teamwork.com

Questions This Section Answers

• What measured visibility pattern and review priority does Teamwork.com have in AI Work Collaboration Platforms?

Teamwork.com recorded 128 mentions (22.07%) and 46 recommendations (7.93%) across 580 answers included in the benchmark. Top 3 placements numbered 10; first-place recommendations numbered 4. Framing among its mentions was 50 positive, 78 neutral and 0 negative.

From October 2026 to November 2026, Teamwork.com's mention count rose from 84 to 128, while its recommendation count fell from 52 to 46. The recommendation rate was 8.93% and 7.93% respectively.

AI answers contained 0 conflicting claims about Teamwork.com: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected owned-page review candidate: https://teamwork.com/blog/best-project-management-tools/.

Selected third-party review candidate: https://gartner.com/reviews/market/collaborative-work-management/vendor/teamwork/alternatives.

Teamwork.com has moderate reach and a steady endorsement share, with endorsements spread across many of its mentions rather than concentrated at the top. Its recorded counts moved in opposite directions between the two selected months, which we treat as a comparability limit. No conflicts were logged. We would brief the project-management roundup page and review the Gartner alternatives listing before any request.

Cisco Webex App

Questions This Section Answers

• What measured visibility pattern and review priority does Cisco Webex App have in AI Work Collaboration Platforms?

Cisco Webex App recorded 12 mentions (2.07%) and 1 recommendation (0.17%) across 580 answers included in the benchmark. Top 3 placements numbered 0; first-place recommendations numbered 0. Framing among its mentions was 2 positive, 10 neutral and 0 negative.

From October 2026 to November 2026, Cisco Webex App's mention count rose from 7 to 12, while its recommendation count stayed at 1. The recommendation rate was 0.17% and 0.17% respectively.

AI answers contained 0 conflicting claims about Cisco Webex App: 0 critical, 0 high, 0 medium, 0 low. Of these, 0 were rated critical or high priority; the critical claims are included in that total. These ratings help prioritize fact-checking. They do not establish that a company or publisher made an error.

No results are available for this part of the analysis.

Both columns quote AI answers, not verified company facts. Resolve the entity, date and source before deciding whether a correction is warranted; different dates can explain an apparent conflict.

These are recorded answer observations, not search volume, verified demand, or proof of missing website content.

Selected third-party review candidate: https://quire.io/blog/p/collaborative-tools.html.

Cisco Webex App appears rarely and has been endorsed only once, with no shortlist or first-place placements recorded. Nearly all of its framing is neutral, so it functions as a passing reference rather than a recommended option. With no disputed claims, the sensible first step is a review of the collaboration-tools source and a brief for owned team-communication content, not a correction request.

The CiteWorks five-step improvement plan

Step 1. Establish the baseline

Questions This Section Answers

• What should the agency deliver when it works to establish the baseline for AI Work Collaboration Platforms?

We would build a cohort scorecard that records reach, endorsement and position for each company separately, with the sample definition attached to every row. Each entry gets an owner, an evidence requirement and a review status, so later months can be compared honestly. This is the foundation for every later decision and keeps the baseline traceable.

Step 2. Verify disputed facts

Questions This Section Answers

• What should the agency deliver when it works to verify disputed facts for AI Work Collaboration Platforms?

AI answers contained 9 conflicting claims across 4 companies. Another 0 possible conflicts were excluded because their company matches were uncertain. The conflicting claims need fact-checking before any corrections are proposed.

Nine conflicting claims across four companies need checking before anything is proposed. We would resolve entity, date and source for each, using current primary documentation, and record a confirmed, no-change or unresolved outcome. Claims with uncertain company matches stay outside the correction queue. Only verified findings would justify a correction request.

Step 3. Review owned pages

Questions This Section Answers

• What should the agency deliver when it works to review owned pages for AI Work Collaboration Platforms?

Each supplied owned-page candidate gets a page brief rather than an edit. The brief states the question the page should answer, the facts that must be verified first, and the evidence needed before publication. We would not assume a page is wrong because an answer disagrees with it; current contents must be checked directly.

Step 4. Review third-party sources

Questions This Section Answers

• What should the agency deliver when it works to review third-party sources for AI Work Collaboration Platforms?

For each third-party source, we would first confirm what the page currently says and whether it addresses the relevant company. Only then would we prepare a request package with verified evidence. A citation in an answer does not show that the source influenced it, so requests should be framed as accuracy checks, not as leverage.

Step 5. Address buyer questions and measure again

Questions This Section Answers

• What should the agency deliver when it works to address buyer questions and measure again for AI Work Collaboration Platforms?

The observed buyer questions give us content briefs grounded in real answer patterns, such as comparisons and product-manager use cases. We would draft those briefs, assign them, and track delivered work separately from answer outcomes. A later scorecard would show what changed, with the same comparability caveats as this report.

Third-party authority and citation opportunities

Questions This Section Answers

• Which cited publisher pages support targeted source review in AI Work Collaboration Platforms?

These pages are candidates for review, not confirmed errors. Citation appearances are counted across the captured answers and are not added again for each company.

The publisher list should be prioritized by relevance and by whether the page actually covers the company in question, not by citation volume alone. A page cited often across many answers is a reasonable first review candidate, but its current contents and any claims about a company must be verified before a request is drafted.

We would group the candidates into owned pages, editorial roundups and community or video sources, since each calls for a different approach. Community threads and videos are poor targets for correction requests. Editorial roundups and comparison pages are better candidates, provided the evidence is current and specific.

What we'd prioritize first

The first priority is the fact-check register, because several companies have disputed claims that block any credible page or source work. Asana's plan limits and ClickUp's time-tracking question are the clearest examples, and both need primary-source confirmation before anything is said publicly.

The second priority is the endorsement gap for companies with strong reach but weak recommendation shares. That is a content and positioning question, not a correction question, and it should be handled through page briefs and buyer-question content rather than publisher requests.

The third is the baseline scorecard, which makes later months comparable. We would sequence the work so that verified facts, approved briefs and justified requests are each recorded with an owner and a status, and we would report completed work separately from measured answer outcomes.

Methodology and conclusion

Questions This Section Answers

• What do the AI Work Collaboration Platforms sample and source-review limits allow us to conclude?

This report covers 2026-11. The exact dates when the original AI answers were collected have not been verified. The visibility rates use 580 answers that met the benchmark's inclusion rules. A company is recommended only when the answer both endorses it and uses positive wording. Top 3 and first-place counts are part of those recommendations, not extra recommendations.

An answer can mention or recommend several companies, so company counts cannot be added to find the number of distinct answers. The net sentiment score is positive minus negative mentions, divided by all company mentions; it describes answer wording, not customer satisfaction.

The wider citation analysis covers 800 captured answers and 5168 appearances of cited URLs. Citation appearances across all captured answers differ from the number of benchmark answers citing a page and the number that also recommend a company. The current contents of cited pages and disputed company claims still need independent verification.

Monthly sample coverage:

The answer samples differ between months. These comparisons show how the recorded results changed; they do not establish why they changed.

This report describes recorded answer patterns for one month, using answers that met the benchmark's inclusion rules. The exact collection dates have not been verified, and the monthly samples differ, so comparisons show what changed in the record, not why. A mention, a recommendation and a placement are distinct measures and should not be combined.

The proposed work is a sequence of verification and content steps, not a promise of improved visibility. We would start with the scorecard and fact register, move to page and source briefs once facts are confirmed, and close with buyer-question content and a fresh scorecard that keeps the same comparability limits.