uplandsoftware.com — AI Site Grade

Analysis

OAI-SearchBot is blocked by Sucuri firewall while every other major AI crawler passes through — a selective, undocumented block that directly impacts ChatGPT's ability to retrieve live site content.

Crawler Access

The site runs on WordPress behind Sucuri/Cloudproxy (Kinsta cache layer visible in headers). A compare_bot_access sweep across 10 user-agents shows OAI-SearchBot receives a 403 from the Sucuri firewall, while GPTBot, ClaudeBot, PerplexityBot, Google-Extended, ChatGPT-User, and Applebot-Extended all return 200 with identical 224KB payloads. This means OpenAI's search/retrieval pipeline (OAI-SearchBot) cannot read the site, but GPTBot (training crawler) can — a split that likely causes ChatGPT's live answers about Upland to be absent or stale. The robots.txt contains no AI-bot-specific directives at all; a single User-agent: * rule blocks /wp-admin/ and duplicate resource pages. The llms.txt returns 404.

Cold-Knowledge Gap

A frontier LLM queried cold knows Upland as a cloud-based enterprise work management vendor that grew through 30+ acquisitions (Adestra, BlueVenn, RightAnswers), is publicly traded (UPLD), and faced financial headwinds in 2023-2024 including restructuring. The site itself, however, leads with "Activate intelligence for AI and agentic operations" — a repositioning toward AI enablement that the cold model does not reflect at all. The model mentions BlueVenn (a CDP) as a key product, yet BlueVenn does not appear anywhere on the current site — it was likely divested or rebranded. The model also describes Upland as "unified platform" integration, while the site presents 20+ largely standalone products under separate subdomains with their own branding.

Schema Posture

Every page carries Yoast-generated JSON-LD with WebPage, BreadcrumbList, WebSite, and Organization types. The Organization schema includes logo, URL, and sameAs links to Facebook, X, and LinkedIn. However, no Product or SoftwareApplication schema exists on any product page examined — the Adestra page, for instance, has no structured data describing the software itself. The homepage has a dateModified of 2026-05-28 (a future date), suggesting a caching or plugin quirk. No FAQ schema, no HowTo, no Article schema on blog posts.

Content Architecture

The site is a multi-site WordPress network with 29 sub-sitemaps (one per product: adestra, interFAX, qvidian, rightanswers, etc.), each with its own WordPress instance. The homepage is content-rich (386 words) with case study callouts and ROI statistics ($3M cost savings, 60% time saved, 133% email engagement increase). The blog publishes frequently (May 2026 posts visible). The /about/ai-innovation/ page describes an AI Council and partnerships with Salesforce, Microsoft, and IBM — but this page was last modified in August 2024, making it stale relative to the fast-moving AI narrative on the homepage.

External Signals

The DNS TXT records include an anthropic-domain-verification token, confirming active Claude/Anthropic relationship management. The investor site (separate subdomain) describes the acquisition strategy and lists products by job function. No significant Reddit or third-party review threads surfaced in search, suggesting limited organic off-domain discussion that would shape AI model priors.

Findings

1. OAI-SearchBot blocked by Sucuri firewall High

   OAI-SearchBot receives a 403 from the Sucuri/Cloudproxy firewall while GPTBot, ClaudeBot, PerplexityBot, and others return 200. This selective block prevents ChatGPT's live retrieval pipeline from accessing site content, likely causing stale or absent answers.

   What to change: Allow OAI-SearchBot through the Sucuri firewall by adding a rule to permit the user-agent, matching the access granted to other AI crawlers.

2. No Product or SoftwareApplication schema on product pages Medium

   Product pages like Adestra lack structured data describing the software, missing opportunities for rich AI snippets and entity recognition.

   What to change: Add SoftwareApplication or Product JSON-LD schema to each product page with name, description, application category, and offers.

3. Homepage dateModified set to future date (2026-05-28) Low

   The homepage JSON-LD includes a dateModified of 2026-05-28, likely a caching or plugin quirk, which may confuse crawlers about content freshness.

   What to change: Correct the dateModified in the Yoast JSON-LD to reflect the actual last modification date.

4. llms.txt file returns 404 Medium

   The standard llms.txt file is not available, missing an opportunity to guide AI crawlers to key content.

   What to change: Create an llms.txt file at the root listing important pages (products, about, AI innovation) to help LLMs discover content.

5. Cold LLM knowledge does not reflect current AI repositioning Medium

   A frontier LLM describes Upland as a unified platform with BlueVenn as a key product, but the site now emphasizes 'AI and agentic operations' and BlueVenn is absent. This mismatch means AI-generated summaries are outdated.

   What to change: Publish authoritative content (e.g., a dedicated AI page, updated product descriptions) and ensure it is crawlable and well-structured to reshape LLM priors.

6. AI Innovation page last modified August 2024 Low

   The /about/ai-innovation/ page describing the AI Council and partnerships has not been updated since August 2024, while the homepage aggressively promotes AI enablement. This inconsistency may confuse crawlers.

   What to change: Update the AI Innovation page with current initiatives, partnerships, and recent content to align with the homepage messaging.

7. No FAQ or Article schema on blog posts Low

   Blog pages lack Article schema, and no FAQ schema is present anywhere, reducing the chance of rich results in AI and search.

   What to change: Add Article schema to blog posts and consider FAQ schema for relevant content.

8. Low external discussion and review presence Medium

   Web searches for reviews, Reddit mentions, and AI visibility returned zero results, indicating limited off-domain conversation that shapes AI model priors.

   What to change: Encourage customer reviews on G2/Capterra and engage in relevant online communities to build external signals.

What's working

• Robots.txt allows all AI crawlers except standard exclusions — The robots.txt has no AI-specific blocks, allowing GPTBot, ClaudeBot, and others to crawl freely.
• 10 of 11 tested AI crawlers return 200 — All major AI crawlers except OAI-SearchBot receive full content, ensuring broad AI visibility.
• Yoast JSON-LD with Organization schema on every page — Every page includes WebPage, BreadcrumbList, WebSite, and Organization schema with logo and social links, providing solid entity grounding.
• Anthropic domain verification token in DNS — The DNS TXT records include an anthropic-domain-verification token, indicating active relationship management with Claude/Anthropic.
• Homepage provides rich content with case studies and statistics — The homepage contains 386 words with specific ROI data and case study callouts, offering strong material for AI summarization.
• Blog publishes frequently with recent posts — The blog shows posts dated May 2026, indicating active content creation that can keep AI models updated.
• Sitemap includes 80 URLs across 29 product sub-sitemaps — The sitemap is well-structured with separate sub-sitemaps for each product, aiding crawler discovery.