# Tickets for Papermark (https://www.papermark.com/)

## 1. Repair or remove internal links to 1 error page

- Effort: S · impact 4/5 · confidence 5/5 · score 20.0
- Evidence:
  - §4: https://www.papermark.com/sharepoint-alternatives-for-data-rooms returns HTTP 404; linked from https://www.papermark.com/, https://www.papermark.com/pricing
- Change: For each URL: if the page should exist, restore it (or 301-redirect it to the closest live page); if it is retired, remove the internal links to it. An error page should not be made indexable.
- Done when: Every internally linked URL in a re-crawl returns 200 directly or after one redirect; retired URLs keep returning 404/410 and are no longer linked.

## 2. Make the existing DocSend comparison answer the prompts DocSend wins

- Effort: M · impact 4/5 · confidence 3/5 · score 6.0
- Evidence:
  - §2: DocSend is named in 58% of answers vs Papermark 45% (mean over engines)
  - §3: 39 of 225 answers name DocSend but not Papermark; most on P09 (6), P23 (5), P11 (4)
  - §5: the site already has https://www.papermark.com/alternatives/docsend, https://www.papermark.com/blog/docsend-api-alternative
- Change: Rework the existing page(s) so their static text answers those buyer prompts directly: who should pick Papermark over DocSend and why (price, limits, features), with a dated pricing comparison. Link them from the pricing page and from the pages for those use cases.
- Done when: The pages' static HTML covers the use cases in the listed prompts, they are in the sitemap and linked from pricing, and a new measurement (--new-measurement) reports the DocSend vs Papermark gap on those prompts.

## 3. Add explicit AI-bot groups to robots.txt

- Effort: S · impact 2/5 · confidence 3/5 · score 6.0
- Evidence:
  - §4: no AI bot has its own robots.txt group; access relies on the '*' rules
- Change: Add explicit groups for OAI-SearchBot, ChatGPT-User and PerplexityBot so a future '*' change cannot silently block answer engines.
- Done when: robots.txt has named groups for the three bots.

## 4. Fill in missing title, meta description or H1

- Effort: S · impact 1/5 · confidence 5/5 · score 5.0
- Evidence:
  - §5: 1 of 21 pages miss a title, description or H1, e.g. https://www.papermark.com/blog
- Change: Give each page a unique title, a description that states the category and buyer, and one H1.
- Done when: Every crawled page has a non-empty title, meta description and H1.

## 5. Re-measure AI answers 30 days after the fixes ship

- Effort: S · impact 1/5 · confidence 5/5 · score 5.0
- Evidence:
  - §9: engine answers vary between runs; one measurement is a baseline, not a trend
- Change: After the other tickets in this list are live, take a new measurement with the same prompt set and samples (`./audit run order.json --new-measurement`; a plain re-run only re-renders this one).
- Done when: A second report with the same prompts shows a new measurement ID and answer capture dates after the fixes; share-of-answer deltas are reported with min-max.