Skip to main content
School of Medicine IT

Web Traffic and SEO Trends

As we continue to evolve our public web presence across the UNC School of Medicine (SOM), understanding where and how our websites are being used can help guide efforts to improve visibility, engagement, and impact—especially among users in North Carolina, our most important audience.

2025 SOM Website Traffic at a Glance

In 2025, the School of Medicine WordPress Network recorded over 15.2 million pageviews, with 12.3 million coming from users in the United States. North Carolina users alone accounted for more than 5.3 million pageviews — approximately 35% of total traffic.

More importantly, North Carolina users were substantially more engaged than all other users, and those engagement gains increased year-over-year.

  • Average views per active NC user (2025): 5.45 (up from 4.88 in 2024)
  • Average views per active non-NC user (2025): 1.94 (down from 2.61 in 2024)
  • Average engagement time per active NC user: 2m 28s (up from 1m in 2024)
  • Average engagement time per active non-NC user: 47s (essentially flat year-over-year)

Takeaway: When we create targeted, high-value content relevant to North Carolinians, those users stay longer and explore more deeply — reinforcing the importance of locally relevant, mission-aligned content. Read more about cleaning up website content.

Search Engines Drive Discovery — Not Social Media

Organic search remains the dominant path to discovery. In 2025, nearly 76% of all sessions originated from Google search, while combined social media referrals (Facebook, LinkedIn, X, etc.) accounted for roughly 1% of sessions.

This reinforces a critical principle for all UNC SOM web authors:
Search engine optimization (SEO) is the most effective and measurable way to grow meaningful audience engagement.

Investments in:

  • descriptive page titles
  • clear heading structures
  • clean, stable URLs
  • intentional internal linking

will consistently outperform time spent chasing social media traffic.

SEO Performance and the Long-Tail Reality

In 2025, the www.med.unc.edu domain received:

  • 5.4 million Google search clicks
  • from 378 million impressions
  • with an overall click-through rate of 1.4% (1.9% among U.S. users)

While the network contains 137,789 URLs with at least one pageview, traffic is highly concentrated:

  • 97,009 URLs (70.4%) received fewer than 12 total pageviews (i.e., fewer than 1 view per month on average).
  • 53% received traffic from organic searches (Google, Bing, etc.)
  • 13,519 URLs exceeded 100 clicks
  • 1,596 URLs exceeded 1,000 clicks
  • 65 URLs exceeded 10,000 clicks
  • Just 3 URLs exceeded 100,000 clicks
  • The top 100 URLs accounted for over 20% of all network traffic in 2025.

Implication: Strategic focus on fewer, higher-value pages will drive disproportionately better results than spreading effort thinly across thousands of low-impact URLs. Cleaning up and removing low-value content will free up time to focus on important areas. Read more about how to (re)focus your website content.

AI Referral Traffic: Early Signals, Clear Implications

In 2025, we began to see early signals of how artificial intelligence (AI) platforms are intersecting with website traffic and information discovery. While AI-generated referral traffic currently represents a very small share of overall sessions, the implications extend far beyond measurable clicks and pageviews.

What the Data Is (and Isn’t) Telling Us

AI-generated traffic referrals accounted for approximately 0.64% of all sessions in 2025, with ChatGPT responsible for 73% of those referrals.

AI platforms such as ChatGPT are increasingly being used as research tools, often replacing traditional search behavior entirely. In many cases, users receive answers, summaries, or recommendations without ever clicking through to a website.

As a result:

  • We can only measure AI-related activity when a user clicks a link from an AI platform to a School of Medicine website.
  • We have no visibility into how often our content is referenced, summarized, or paraphrased within AI tools.
  • A lack of referral traffic does not mean our content is not being surfaced or used.

This represents a fundamental shift: visibility and influence are increasingly decoupled from traffic metrics.

Takeaway: The same practices that improve SEO also improve AI interpretation

Device and Browser Usage

Among U.S. users, device usage was nearly evenly split:

  • Desktop: 55.2%
  • Mobile: 43.7%

Mobile users primarily accessed content via:

  • iPhone: 78.5%
  • Samsung Galaxy: 13.1%

Browser usage was dominated by:

  • Chrome: 50.3%
  • Safari: 36.4%

This reinforces the need for mobile-first design, performance optimization, and cross-browser testing — particularly on Chrome and Safari.