Posts tagged analytics

Key for Search Module Codes and Names

Search.gov Home > Admin Center > YourSite > Analytics

We call several different indexes when serving search results, and we present results from each of these indexes in a separate module. Each module loads when there are relevant search results to display. Each module has a code, and you’ll see these codes in the Download Details csv files available through the Queries or Clicks pages of the Admin Center analytics section.

This page lists all the module codes and their full names.

Code Module Name
AIDOC      Collections
AWEB Web Results Only (Azure)
BBG Best Bets: Graphics
BOOS Best Bets: Text
BSPEL Spelling Suggestions (Bing/Google)
BWEB Web Results (Bing)
DECOR Value-added Links for SEC Filings
FRDOC Federal Register Documents
GWEB Web Results (Google)
I14Y Web Results (Search.gov via i14y)
IMAG Image Results (Bing)
JOBS Jobs
LOVER Spelling Overrides (Search.gov)
MEDL Health Topics
NEWS RSS Feeds
NIMAG Image Results (MRSS)
OASIS Image Results (Search.gov)
OSPEL Spelling Suggestions (Search.gov Images)
OVER Spelling Overrides (Bing/Google)
QRTD Routed Queries
SPEL Spelling Suggestions (Search.gov)
SREL Related Searches
TWEET Tweets (Twitter)
VIDS Video Results (YouTube)

Analyzing Your Site's Queries

Search.gov Home > Admin Center > YourSite > Analytics > Queries

The Queries page lists the number of times a search query (that is, a word or string of words) was entered in the search box on your site by a unique searcher.

Data are shown for the present month-to-date by default. You can change the time period by selecting a different month or year and re-generating the report.

You can also toggle between filtered and unfiltered views of the data by clicking on the filter icon () in the top navigation bar. The filtered data represent our best effort to show you real searches performed by your site’s visitors. The unfiltered data include nearly all searches and clicks. Only known spiders (such as Bingbot and Googlebot) are excluded.

Data are available in the Admin Center from the past 13 months.

Top Queries

Top 1,000. You’ll see a list, in frequency order, of the 1,000 most popular queries for your selected time period.

Top 1,000 by query. Enter any specific term or phrase into the Query search box. Re-generate the report to see the top 1,000 queries that contain that term or phrase and its stemmed equivalents.

For example, below are the top 10 queries related to housing on USA.gov in August 2013.

  1. house of representatives
  2. housing
  3. speaker of the house
  4. white house
  5. house minority leader
  6. housing assistance
  7. house majority leader
  8. speaker of the house 2013
  9. the white house
  10. houses

Download Details

You can download a detailed CSV (comma separated values) file for any search term. There’s one line for each time this query was run. This CSV file gives you access to the raw search logs, including:

  • The date and time of the query
  • The query (It is embedded in the request URL. Look for the text following ‘query=’.)
  • Where the person was when they ran the search (We use the ISO-2 alpha standard for country codes and ISO 3166-2 codes for country subdivisions.)
  • The modules loaded in response to the query (Our Module Codes page provides a key of the codes and names.)
  • The kind of device and browser they were using

Clicks from a Query and Queries Leading to a Click

Select “View Clicks” to view the top clicked URLs for each query.

If you clicked on “View Clicks” for the house of representatives example above, you’d see that the top URLs clicked for that query are:

  1. www.house.gov
  2. www.house.gov/representatives
  3. www.house.gov/representatives/find

From this list of clicked URLs, you’d be able to select “View Queries” to view the top queries that led to a click for each URL.

If you clicked on the first-listed www.house.gov in the example above, you’d see that the top queries leading to a click on that URL are:

  1. house of representatives
  2. congress
  3. congress members
  4. congress representatives

You can loop through or drill down into this list for any set of clicks and queries.

Click-Through Rate (CTR)

The click-through rate (CTR) is the the number of clicks from a query divided by the number of queries (multiplied by 100 to be expressed as a percentage).

High or low CTR’s may indicate that users are not able to find relevant results. A CTR can be over 100%: if a user enters a query, clicks on a link, hits ‘back’ in their browser, returns to the search results, and clicks again, the CTR for that query would be 200% (2 clicks divided by 1 query, multiplied by 100 = 200%).


Did you know? The Monthly Report section gives a bird’s-eye view of the number of queries and clicks on your site each month. The Site Overview provides a snapshot of what has been happening on your site in the past day or so.

Analyzing Your Site's Clicks

Search.gov Home > Admin Center > YourSite > Analytics > Clicks

The Clicks page shows what searchers clicked on from your search results pages, and the number of times each of those items was clicked during the selected date range. The current month is shown by default, and the date can be adjusted using the date selectors at the top of the page and re-generating the report.

You can also toggle between filtered and unfiltered views of the data by clicking on the filter icon () in the top navigation bar. The filtered data represent our best effort to show you real searches performed by your site’s visitors. The unfiltered data include nearly all searches and clicks. Only known spiders (such as Bingbot and Googlebot) are excluded.

Data are available in the Admin Center from the past 13 months.

Reading the Report Table

Top URLs Clicked

The top 1,000 most clicked-on URLs from your results page for the selected time period, with the most popular first.

Download Details

You can download the raw click logs for any URL in the list. The CSV (comma separated values) file has one line for each time this URL was clicked. Details avialable include:

  • Date and Time of the click,
  • Query: The query they ran before clicking the search result,
  • Position: Where in the results page was the result they clicked, e.g. 5 means the searcher clicked the fifth result in the list,
  • Request: The clicked URL (It is embedded in the request URL. Look for the text following ‘u=’.) and its position on the page,
  • Referrer: The webpage the person was on when they ran the search, this column also includes the query,
  • Vertical and Modules: The modules loaded in response to the query (Our Module Codes tip provides a key of the codes and names.),
  • Device, Browser, and OS: The kind of device and browser they were using, and
  • Country Code and Region: Where the user was, physically. We use the ISO-2 alpha standard for country codes and ISO 3166-2 codes for country subdivisions.

# of Clicks

Select a hyperlinked number in the # of Clicks column to view the queries that led to the clicks for that URL (the Queries leading to a Click report).

If you then click on a number in the # of Clicks column for a query on the Queries leading to a Click page, you will see all the URLs that were clicked on from that query’s results pages (the Clicks from a Query report). This is the same report you see when you follow a # of Clicks link from the Queries page.

You can loop through or drill down into this list for any set of clicks and queries.


Did you know? The Monthly Report gives a bird’s-eye view of the number of queries and clicks each month. The Site Overview provides a snapshot of what has been happening on your site today.

Analyzing Your Monthly Reports

Search.gov Home > Admin Center > YourSite > Analytics > Monthly Reports

The monthly report gives a bird’s-eye view of the number of queries and clicks each month.

Data are shown for the present month-to-date by default. You can change the time period by selecting a different month or year and re-generating the report.

You can also toggle between filtered and unfiltered views of the data by clicking on the filter icon () in the top navigation bar. The filtered data represent our best effort to show you real searches performed by your site’s visitors. The unfiltered data include nearly all searches and clicks. Only known spiders (such as Bingbot and Googlebot) are excluded.

Monthly Usage Stats

Data for the total number of queries and clicks. Use this report to find and analyze the so-called short head (External link) of your most popular search queries and clicks, which are typically the same as your site’s top tasks.

Queries with No Results

A list of queries that returned no results. Sometimes, searchers look for information on the wrong website or use different words than what are used on the page, and so they get no results. Use this list of Queries with No Results to help searchers get to the information they’re looking for. You can set up Best Bets, or a Routed Query. You can also customize the No Results message displayed on your results page.

Impressions and Clicks by Module

Detailed data for the total number of impressions and clicks and your clickthru rate (CTR) are presented for each “module” on the search results page. We also provide the average CTR across all Search.gov customers so you can see how your rate compares to the average rate. The data in the table is sorted in descending order from the most to least number of clicks.

Use this report to inform which modules you opt to display.

Download Top Queries

View a detailed report of the number of times searchers have input specific terms and phrases by clicking on the CSV (comma-separated values) link to download the list as a text file. Once you’ve downloaded the CSV file, you can easily import it into Excel (External link) or another speadsheet program to analyze the data.

Read Understanding Your Users’ Needs By Analyzing Search Terms for tips on how to create a semi-automated report for analyzing the data in this CSV file on a regular basis.

Definitions

All data presented on this page (and other pages in the Admin Center) are IP-deduped to exclude bots and other noise in an attempt to accurately represent searchers’ intent on your site. Some definitions follow.

Queries: Number of times a search query (that is, a word or string of words) was entered in the search box by a unique searcher.

Clicks. Number of times a searcher clicked on one of your pages within the search results for a particular query.

Impressions: Number of times a module is displayed, whether it is clicked on or not. Not all modules are displayed for all queries. Each time a module is displayed it is counted as one impression.

Clickthru Rate (CTR): The rate (expressed in a percentage) at which searchers click on a module. This rate is calculated by dividing the total number of clicks by the total number of impressions. CTR is useful to measure the performance of specific campaigns, such as Best Bets for seasonal events.

Raw count: Total number of times the query was submitted. The raw count is provided in the downloadable CSV files only.

IP-deduped count: Total number of times the query was submitted by any one IP address. This excludes bots and other traffic that send in a query multiple times from one IP address. It is often a more accurate representation of “real,” human traffic.


Did you know? On the first of each month, we email you a report with data on the previous month’s queries, clicks, and top search terms.

Did you know? The Site Overview provides a snapshot of what has been happening on your site in the past day or so.

View a Snapshot of What's Happening on Your Site Today

Search.gov Home > Admin Center > YourSite > Dashboard > Site Overview

The Site Overview page provides a snapshot of what has been happening on your site in the past day. All of the analytics in the Admin Center are updated in real-time, which means that you instantaneously see the searches and clicks as they’re performed on your site.

You can opt to receive this snapshot as a daily email by clicking on the envelope icon () in the top navigation bar next to your selected site.

You can also toggle between filtered and unfiltered views of the data by clicking on the filter icon () in the top navigation bar. The filtered data represent our best effort to show you real searches performed by your site’s visitors. The unfiltered data include nearly all searches and clicks. Only known spiders (such as Bingbot and Googlebot) are excluded.

Top Queries

A list of the 10 most popular searches on your site. If you don’t have any searches, you’ll see a “Not enough query data available” message.

The data are processed to present “real” searches by humans (that is, the data are de-duped by IP address to remove bot traffic and other noise).

Top Clicked URLs

A list of the top 10 pages that were clicked thru by searchers on your search results page. If you don’t have any clicks, you’ll see a “Not enough query data available” message.

A list of the top 10 changing URLs that were visited by users on your website. If there are any pages that are trending, they’ll appear here.

Use this report to identify newly popular pages on your website. Investigate why pages are trending, if the reason is not immediately apparent.

You’ll only see the trending URLs list if you have our Javascript snippet on your web pages. Data are updated every few minutes, so you may see URLs come and go quickly.

A list of the top 10 search terms with the greatest gain between yesterday and today. If there are any queries that are trending, they’ll appear here.

Use this report to identify newly popular terms. Create new content or update existing content to ensure it’s current, accurate, and complete.

Queries with No Results

A list of the top 10 queries that returned no results.

Use this data to help searchers find your content by adding a Best Bet, updating your existing web pages, or both.

For example, on one agency’s website, a dozen searches for frostline returned no results. The frost line—also known as depth of frost or freezing—is the depth to which the groundwater in soil is expected to freeze. The agency’s pages on the topic referred to this concept as depth of freezing.

Top Queries with Low Click Thrus

A list of the top 10 searches with low click thru rates. If there are any frequently searched terms that returned results but didn’t get a click at least 20% of the time, they’ll appear here.

Use this alert about low click thrus to identify issues with coverage. You may opt to create a Best Bet, add pages that may be missing from the results, or do both to improve recall. Or, you may opt to incorporate language from these popular search terms into your page titles and descriptions, or enter common search terms as keywords, verbatim.

This Month’s Totals to Date

The total number of queries and clicks for the present month-to-date. A click is recorded each time a searcher clicks on a results link.

A graph of your site’s total search queries over time is also presented to provide an overview of how your traffic has trended from the time your agency started using our service (but no further back than the last 13 months). Hover over the trend line to see the total number of queries in a given month.


Did you know? The Monthly Report gives a bird’s-eye view of the number of queries and clicks each month.

Did you know? Click the pushpin icon in the top navigation bar to change the default site that appears when you first log in to the Admin Center.

Setting Up Analytics Alerts

Search.gov Home > Admin Center > YourSite > Analytics > Analytics Alerts

Analytics Alerts let you receive email alerts when your site visitors are getting no results for a particular query or when a query is getting a low click-through rate. Note: the Low Click-through Rate alert is not currently functional. We apologize for any inconvience this may cause.

No Results

Set up this alert to be notified if specific queries are frequently searched for but are returning no results. Help on specific fields follows, and there is an example below.

Name

Enter a name for your alert. It will be the subject line of your alert email. The name of your alert must be unique across our entire search alert system, so if you see an error message, please modify the name.

Check interval

Enter the frequency at which you’d like the system to check for queries with no results. Use m, h, d, or w to specify the minutes/hours/days/weeks you’d like between checks.

Minimum number of queries

Enter the minimum number of times, over the time window, that a given query must be entered to trigger the alert.

Time window for each check

Enter the amount of time the system should look back through for each check. Use m, h, d, or w to specify minutes/hours/days/weeks.

Time between alerts

Enter the minimum period of time between successive email messages.

Ignored query terms

Enter any query terms that you don’t want included in the alert, separated by commas.

No Results Example:

In this example, the alert system will check every 30 minutes to see if any query has been entered at least 50 times over the past 12 hours and yielded no results. The system will send an alert at most once per day if there are queries that fit this criteria, and it will not include the query ‘panda’ in the alert.

No Results Analytics Alert Example

Low Query Click-through Rate

Note: the Low Click-through Rate alert is not currently functional. We apologize for any inconvience this may cause.

Set up this alert to be notified if the results page for a query with significant traffic is not getting many clicks. If the query resulted in a click less than 20% of the time, it is considered to have a low click-through rate (CTR); however, you are able to set any percentage threshold for your alert emails. Help on specific fields follows, and there is an example below.

Name

Enter a name for your alert. It will be the subject line of your email. The name of your alert must be unique across our entire search alert system, so if you see an error message, please modify the name.

Check interval

Enter the frequency at which you’d like the system to check for queries with a low CTR. Use m, h, d, or w to specify the minutes/hours/days/weeks you’d like between checks.

Minimum search activity level (queries + clicks)

Enter the minimum number of times, over the time window, that a given query must have search activity (queries plus clicks) to trigger the low query CTR alert. Because we’re not able to differentiate between query activity and click activity, you may want to pad this number a little. However, since these queries are not getting a lot of clicks, you don’t want to pad it too much.

Maximum CTR for alert

Enter the maximum click-through rate that will trigger an alert, expressed as a percentage. You will receive an alert for a query that has a CTR equal to or lower than that percentage during the given time window.

Time window for each check

Enter the amount of time the system should look back through for each check. Use m, h, d, or w to specify minutes/hours/days/weeks.

Time between alerts

Enter the minimum period of time between successive email messages.

Ignored query terms

Enter any query terms that you don’t want included in the alert, separated by commas.

Low CTR Example:

In this example, the alert system will check every 30 minutes for a query that has had search activity (queries plus clicks) at least 100 times over the past 12 hours and has a CTR 15% or lower. The system will send an alert at most once per day if there are queries that fit these criteria, and it will not include queries for ‘lion’ in the alert.

Low CTR Analytics Alert Example


Troubleshooting tip: Don’t forget to hit ‘add’ in the upper right corner!

Troubleshooting tip: Analytics alerts are only sent to the Search.gov user who set up the alert.

How to Add JavaScript for Your Third-party Web Services

Search.gov Home > Admin Center > YourSite > Analytics > 3rd Party Tracking

Do you want your search results page to run third-party web services such as Foresee, Google Analytics, Omniture, Siteimprove, or WebTrends?

Input the JavaScript code you’d like to call from your search results page. Click submit to send us your request. We’ll input your code for you and send you an email to confirm that we’ve done it.

Some tips for commonly used third-party web services follow.

Note: We do not currently support Google Tag Manager. You will need to submit all of the scripts you are managing in your Google Tag Manager setup.

Google Analytics Tip

Within your Google Analytics account, select the option, Do Track Site Search. Set the query parameter as query. For more information, read Google’s tip, Set Up and Configure Site Search (External link).

Additionally, if you’ve requested domain masking and you want to include the analytics for your search.agency.gov subdomain with your main agency.gov domain, you’ll need to set your domain in your Google Analytics JavaScript by including _gaq.push(['_setDomainName', 'agency.gov']);.

For more information, read Google’s tip, Tracking Multiple Domains (External link).

The code you submit should look something like one of the following two scripts.

Google Analytics Code (Older Format)

 <script type="text/javascript">
 var _gaq = _gaq || []; 
 _gaq.push(['_setAccount', 'UA-########-1']); 
 _gaq.push(['_trackPageview']);
 
 (function() { 
 var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; 
 ga.src = ('https:' == document.location.protocol ? 'https://ssl'; : 'http://www') +      '.google-analytics.com/ga.js'; 
 var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); 
 })();
 </script>

Google Analytics Code (Newer Format)

 <script type="text/javascript">
 (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
})(window,document,'script','//www.google-analytics.com/analytics.js','ga');
 </script>

Digital Analytics Program Tip

Does your federal agency participate in the Digital Analytics Program (DAP)? You don’t need to do anything. We’re already fully integrated with DAP.

ForeSee Tip

Coordinate with your ForeSee representative and the Search.gov team to implement your customer satisfaction survey on your results page. The four general steps follow.

  1. Email us at search@support.digitalgov.gov to set up a CNAME for search.youragency.gov.

  2. Update the files path in your Foresee code to use an absolute path instead of a relative path.

    Find => ‘files’: ‘/fsrscripts/’,

    and replace it with => ‘files’: ‘//www.youragency.gov/fsrscripts/’,

    (Or, find => ‘files’: ‘/foresee/’, and replace it with => ‘files’: ‘//www.youragency/foresee/’,)

    in the following five files.

    • foresee-trigger.js
    • foresee-tracker.js
    • foresee-alive.js
    • foresee-qualifier.js
    • foresee-test.js
  3. Submit your foresee-trigger.js via our Admin Center. It should look something like the following script. <script type="text/javascript" src="//www.youragency.gov/library/foresee/foresee-trigger.js"></script>

  4. We’ll send you an email to confirm that we’ve set up both your CNAME and added the script for your foresee-trigger.js file.


Did you know? We use Google Analytics Web analytics software—our own tag plus the Digital Analytics Program tag—by default. Email us at search@support.digitalgov.gov if you’d like to opt out. Learn more about our site policies and terms of service.

Search Engine Optimization for Government Websites

On June 10, 2014, the Metrics Community of Practice of the Federal Web Managers Council and DigitalGov University hosted an event to honor the memory of Joe Pagano, a former co-chair of the Web Metrics Sub-Council.

This third lecture honoring Joe focused on search engine optimization (SEO).

While commercial search engines do a remarkable job of helping the public find our government information, as web professionals, it’s also our job to help the public make sense of what they find.

Ammie Farraj Feijoo, our program manager, presented on SEO for government websites and specifically talked about:

  • What SEO is and why it is important;
  • SEO building blocks for writing content;
  • Conducting keyword research; and
  • Eliminating ROT (redundant, outdated, and trivial content).

Download the slide deck [PDF] and visit the resources below to learn more.

Webmaster Tools

A Few (of Many) SEO Resources


Page last reviewed or updated:

Search Is the New Big Data

Search is easy, right? You type a term in a search box and the exact page you’re looking for appears at the top of the list of results. But search is hard and has many shades of grey.

On April 10, 2014, Loren Siebert, our senior search architect, presented on:

  • Complexities of recall and precision,
  • Popular open source search technologies, and
  • “Search magic” like stemming, synonyms, fuzziness, and stopwords.

Download the slide deck and visit the resources below to learn more.


Page last reviewed or updated:

Analyzing Your Site's Referrers

Search.gov Home > Admin Center > YourSite > Analytics > Referrers

The Referrers report shows what pages searchers were on when they entered their queries. The current month is shown by default, and the date can be adjusted by using the date selectors at the top of the page and re-generating the report.

You can also toggle between filtered and unfiltered views of the data by clicking on the filter icon () in the top navigation bar. The filtered data represent our best effort to show you real searches performed by your site’s visitors. The unfiltered data include nearly all searches and clicks. Only known spiders (such as Bingbot and Googlebot) are excluded.

Data are available in the Admin Center from the past 13 months.

Top Referrers

The top 1,000 most popular referring URLs for your selected time period, with the most popular listed first.

# of Queries

The number of queries made from a specific referring URL.

View All Queries from this Referring URL

Click on the link to see a list of all the queries that were made from a specific URL, with the most popular listed first (the Queries from a Referring URL report). You can see the number of times each query was made from that URL. You can also click to view a list of all the referring URLs for each unique query (the Referring URLs Leading to a Query report).