Using Google Analytics to Get the Most Popular WordPress Pages

README.md

This is a set of PHP scripts to fetch the most-viewed pages from a Google Analytics property.

You will need to set up API access for this project in the Google Cloud Platform Console and create an authorization JSON file.

composer.json

{
    "require": {
        "google/apiclient": "^2.7"
    },
    "scripts": {
        "post-update-cmd": "Google\\Task\\Composer::cleanup"
    },
    "extra": {
        "google/apiclient-services": [
            "Analytics"
        ]
    }
}

get-most-popular-posts-from-google-analytics.php

<?php

// Load the Google API PHP Client library.
require_once( __DIR__ . '/../vendor/autoload.php' );

// Query Analytics API for the pages with the most views.
function get_most_viewed_post_slugs() {
  try {
    $analytics = establish_client_connection();
  }
  catch ( Exception $error ) {
    error_log( $error );
    return;
  }

  // View this query in the Query Explorer: https://ga-dev-tools.appspot.com/query-explorer/?ids=ga%3ATODO&start-date=7daysAgo&end-date=today&metrics=ga%3Apageviews&dimensions=ga%3ApagePath&sort=-ga%3Apageviews&filters=ga%3ApagePath%3D~%5B%5Cd%2F%5D%7B10%7D&max-results=5
  $results = $analytics->data_ga->get(
    'ga:TODO',      // view
    '7daysAgo',     // start date
    'today',        // end date
    'ga:pageviews', // metrics
    array(          // options
      // We only need to look at the page path.
      'dimensions'  => 'ga:pagePath',

      // Sort by the most pageviews in descending order.
      'sort'        => '-ga:pageviews',

      // Get only pages with a URL that has a string of ten numbers and slashes,
      // which is what blog post URLs look like, e.g.: /2020/04/14/foo-bar-baz
      'filters'     => 'ga:pagePath=~[\d/]{10}',

      // Limit results.
      'max-results' => 5,
    )
  );

  if ( ! is_object( $results ) ) {
    throw new Exception( 'Could not retrieve data from Google Analytics.' );
  }

  // Extract the results we want.
  $posts = $results->rows;

  if ( empty( $posts) ) {
    throw new Exception( 'No rows found in Google Analytics data.' );
  }

  // Get post slugs from the results.
  $post_slugs = extract_post_slugs( $posts );

  return $post_slugs;
}

// Query Analytics API for pages that have the most 'share via email' events.
function get_most_emailed_post_slugs() {
  try {
    $analytics = establish_client_connection();
  }
  catch ( Exception $error ) {
    error_log( $error );
    return;
  }

  // View this query in the Query Explorer: https://ga-dev-tools.appspot.com/query-explorer/?ids=ga%3A56434181&start-date=7daysAgo&end-date=today&metrics=ga%3AtotalEvents&dimensions=ga%3ApagePath%2Cga%3AeventLabel&sort=-ga%3AtotalEvents&filters=ga%3AeventLabel%3D~email&max-results=5
  $results = $analytics->data_ga->get(
    'ga:56434181',    // view
    '7daysAgo',       // start date
    'today',          // end date
    'ga:totalEvents', // metrics
    array(            // options
      // We need to look at the page path and the event label.
      'dimensions'  => 'ga:pagePath,ga:eventLabel',

      // Sort by the most pageviews in descending order.
      'sort'        => '-ga:totalEvents',

      // Get only pages with a URL that has a string of ten numbers and slashes,
      // which is what blog post URLs look like, e.g.: /2020/04/14/foo-bar-baz
      'filters'     => 'ga:eventLabel=~email',

      // Limit results.
      'max-results' => 5,
    )
  );

  if ( ! is_object( $results ) ) {
    throw new Exception( 'Could not retrieve data from Google Analytics.' );
  }

  // Extract the results we want.
  $posts = $results->rows;

  if ( empty( $posts) ) {
    throw new Exception( 'No rows found in Google Analytics data.' );
  }

  // Get post slugs from the results.
  try {
    $post_slugs = extract_post_slugs( $posts );
  }
  catch ( Exception $error ) {
    error_log( $error );
    return;
  }

  return $post_slugs;
}

// Extract post slugs from the list of page URLs.
function extract_post_slugs( $posts ) {
  if ( empty( $posts ) ) {
    throw new Exception( 'No posts provided to extract slugs from.' );
  }

  // Get just the slugs from each post.
  $post_slugs = array();

  foreach ( $posts as $post ) {
    $post_path = $post[0];

    // The page URLs look like this: /2020/04/14/foo-bar-baz. We just want the
    // 'foo-bar-baz' part, which is 12 characters from the start of the URL.
    $post_slug = substr( $post_path, 12 );

    // Remove query string if one exists.
    if ( stripos( $post_slug, '?' ) ) {
      $post_slug = substr( $post_slug, 0, stripos( $post_slug, '?' ) );
    }

    // If there's a trailing slash, remove it.
    $post_slug = rtrim( $post_slug, '/' );

    $post_slugs[] = $post_slug;
  }

  if ( empty( $post_slugs ) ) {
    throw new Exception( 'Could not extract page slugs from the results.' );
  }

  return( $post_slugs );
}

get-popular-posts-data-from-database.php

<?php

function cached__get_popular_posts() {
  $transient_name = 'relpol_most_popular_posts';

  if ( false === get_transient( $transient_name ) ) {
    // echo '<p>DEBUG: No transient found. Making new request.</p>';
    $popular_posts = get_popular_posts();
  }
  else {
    // echo '<p>DEBUG: Transient found! Using it.</p>';
    $popular_posts = get_transient( $transient_name );
  }

  return $popular_posts;
}

function get_popular_posts() {
  // Store this data in a transient.
  $transient_name = 'relpol_most_popular_posts';

  $popular_posts = array(
    'most_viewed' => array(),
    'most_emailed' => array(),
  );

  // Get most viewed and emailed post IDs by looking them up via the slugs in
  // their URLs.
  try {
    $most_viewed_post_slugs = get_most_viewed_post_slugs();

    // DEBUG:
    // echo '<pre style="background:#fff;color:#000;font-size:14px;border:2px solid red;padding:12px">MOST VIEWED SLUGS:<br>';
    // print_r( $most_viewed_post_slugs );
    // echo '<br><br>MOST VIEWED POSTS:<ul>';

    foreach ( $most_viewed_post_slugs as $slug ) {
      $corresponding_post = get_page_by_path( $slug, OBJECT, array( 'post' ) );
      $popular_posts['most_viewed'][] = $corresponding_post->ID;

      // DEBUG:
      // echo '<li>' . $corresponding_post->ID . ': ' . $corresponding_post->post_title . '</li>';
    }

    // DEBUG:
    // echo '</ul></pre>';
  }
  catch ( Exception $error ) {
    error_log( $error );
  }

  try {
    $most_emailed_post_slugs = get_most_emailed_post_slugs();

    // DEBUG:
    // echo '<pre style="background:#fff;color:#000;font-size:14px;border:2px solid red;padding:12px">MOST EMAILED SLUGS:<br>';
    // print_r( $most_emailed_post_slugs );
    // echo '<br><br>MOST EMAILED POSTS:<ul>';

    foreach ( $most_emailed_post_slugs as $slug ) {
      $corresponding_post = get_page_by_path( $slug, OBJECT, array( 'post' ) );
      $popular_posts['most_emailed'][] = $corresponding_post->ID;

      // DEBUG:
      // echo '<li>' . $corresponding_post->ID . ': ' . $corresponding_post->post_title . '</li>';
    }

    // DEBUG:
    // echo '</ul></pre>';
  }
  catch ( Exception $error ) {
    error_log( $error );
  }

  // Clear existing transient data.
  delete_transient( $transient_name );

  // Save data in a transient that expires regularly.
  set_transient( $transient_name, $popular_posts, 2 * HOUR_IN_SECONDS );

  // DEBUG:
  // echo '<pre style="background:#fff;color:#000;font-size:14px;border:2px solid red;padding:12px">POPULAR POSTS:<br>';
  // print_r( $popular_posts );
  // echo '</pre>';

  return $popular_posts;
}

$popular_posts = cached__get_popular_posts();
// $popular_posts = get_popular_posts();

google-analytics-api-helpers.php

<?php

// Load the Google API PHP Client library.
require_once( __DIR__ . '/../vendor/autoload.php' );

// Establish a client connection with the service account.
function establish_client_connection() {
  $client = new Google_Client();
  $client->setAuthConfig( __DIR__ . '/google-api-service-account-credentials-TODO.json' );
  $client->setScopes( ['https://www.googleapis.com/auth/analytics.readonly'] );
  $analytics = new Google_Service_Analytics( $client );

  if ( empty( $client ) || empty( $analytics ) ) {
    throw new Exception( 'Could not connect to Google Analytics API.' );
  }

  return $analytics;
}