Skip to main content

​
Metrics & Charts Guide

To trust your analytics, you need to understand exactly how each number is calculated, where the data originates, and how edge caching influences your counts. This guide provides an exhaustive breakdown of every card, graph, histogram, and breakdown available in GitAscii Pro.

​
1. Primary KPIs & Stat Cards

​
Total Views vs. Unique Visitors

  • Total Views: Represents the total number of HTTP requests processed by GitAscii’s SVG render API for your profile during the selected time period.
  • Unique Visitors: Calculated via Redis HyperLogLog registers (PFCOUNT). Each unique device token (computed via daily rotating salt) is counted once per time window. If the same person views your README 15 times in one day, it counts as 15 views and 1 unique visitor.
  • Trend Percentage: Compares current period performance with the immediately preceding equal time period (e.g. current 30 days vs. previous 30 days): Trend (%)=(Current Period−Previous PeriodPrevious Period)×100\text{Trend } (\%) = \left( \frac{\text{Current Period} - \text{Previous Period}}{\text{Previous Period}} \right) \times 100

​
Cache Hit Ratio & Response Statuses

  • Cache Hit Ratio (e.g. 98.2%): The percentage of requests where the client or GitHub Camo proxy had an up-to-date cached SVG and GitAscii responded with 304 Not Modified (or CDN edge cache hit) instead of re-rendering full SVG trees (200 OK).
  • Why High Cache Ratios Are Good: Indicates rapid loading speeds for your README visitors and minimal network transfer time.

​
Average Render Latency

  • Average Latency (ms): Measured from the instant the HTTP request arrives at GitAscii’s serverless handler to the final SVG stream completion.
  • Typical latencies range from 15ms to 35ms, ensuring zero perceptible delay for anyone visiting your GitHub profile.

​
What It Displays:

  • Total Views Line (Lime): Daily volume of SVG requests.
  • Unique Visitors Line (Cyan): Daily unique visitor counts.
  • GitHub Camo Proxy Area: Differentiates views coming through GitHub’s proxy vs direct hits.
  • Interactive Tooltip: Hovering over any point displays:
    • Exact date (YYYY-MM-DD)
    • Views count
    • Unique visitors count
    • Cache hit percentage
    • Camo proxy vs Direct request split

​
3. Hourly Traffic Distribution (24-Hour Histogram)

​
How to Use This Data:

  • Shows which hours of the day receive peak traffic.
  • Helps developers determine when to push major README updates, release open-source projects, or publish articles to maximize initial impressions.
  • Displays both direct browser traffic and proxy pings.

​
4. Weekday × Hour Activity Grid (Heatmap Matrix)

A 7-day ×\times 24-hour matrix (168 cells) color-coded by visitor intensity:
  • Rows: Days of the week (Sunday through Saturday).
  • Columns: Hours of the day (0 to 23 UTC).
  • Cell Color: Gradient from dark background to vibrant lime (#c5ff4a) based on normalized density: Intensity=min⁡(100,Views in CellMax Cell Views×100)\text{Intensity} = \min\left(100, \frac{\text{Views in Cell}}{\text{Max Cell Views}} \times 100\right)

​
5. Geographic Demographics & World Map

​
How Country Geolocation Works

When an SVG request arrives at the edge:
  1. The server extracts the standard ISO 3166-1 alpha-2 country code from edge headers (cf-ipcountry or x-vercel-ip-country).
  2. The country code is recorded in a Redis dimension hash for that date.
  3. The raw IP address is discarded immediately without writing to disk.

​
Map Visualizations:

  • World Map Chloropleth: Continents and countries shaded according to their relative share of total profile views.
  • Top Countries Ranking: Country names, ISO flags, view count, and percentage share.
  • Continent Breakdown: Donut chart grouping traffic across North America, Europe, Asia, South America, Africa, and Oceania.
  • Timezone Distribution: Common timezones represented by your audience.

​
6. Technology & Runtime Dimensions

GitAscii Pro parses coarse User-Agent headers to provide audience technology insights:

​
7. Multi-Profile Performance Comparison

If you manage multiple profile configurations (e.g. default, minimal, stats):
  • The Profiles Table aggregates metrics side-by-side:
    • Total Views per slug
    • Unique Visitors per slug
    • Camo vs Direct ratio
    • Active error status
    • Relative contribution to total portfolio impressions

​
8. Data Reliability & Trust Checklist

GitHub repository traffic (/traffic tab) only counts visits to repository source code pages. GitAscii measures impressions of your Profile README (github.com/username), which typically receives significantly more views from social media, GitHub search, issue discussions, and PR reviews.
GitAscii sends proper Cache-Control headers and handles If-None-Match / ETag validations. Whenever GitHub’s Camo proxy revalidates an SVG with a conditional request (304 Not Modified), GitAscii records the view and increments the cache-hit metric.
Known search engine crawlers (Googlebot, Bingbot, GitHub-Hookshot) are tagged separately under the Bot / Crawler device dimension so they do not artificially distort developer engagement insights.