The prevalence of technology in everyday life has changed the job market considerably for students of Library and Information Science. Traditional roles have shifted and morphed to keep up with the times. While a traditional librarian role may be what comes to mind when thinking of library school, there are other jobs you may not have considered that you may already be qualified for. This is my account of a non-traditional library role as an online reporting/web analytics intern for a publishing firm.

One day I came across a job posting for an internship at a large publishing firm. I applied for the internship position as an online reporting/web analytics intern for the web marketing team. I wasn't quite sure what to expect. I found that my MLIS and Information Management (IM) certificate fit the requirements of the internship nicely and I was ready for an adventure.

Online Reporting / Web Analytics

As the online reporting/web analytics intern at a publishing firm, I am responsible for creating dashboards using Adobe's Omniture Digital Marketing Suite and sending them to the stakeholders. Dashboards are created by stringing together a series of reports that help provide insightful information regarding web usage. In other words, Dashboards are a series of web usage reports.

I create these dashboards through requests that are sent to me either through email or SmartQ (a project organizer application). There are two types of requests that come in for the creation of dashboards.

• The first type of request is specific. The specific request tells me what reports to include in the dashboard.

• The second type of request is scenario based. The scenario request doesn't provide you with the specific reports to include. Instead, the scenario request (as the name suggests) provides you with the purpose of how the information will be used.

The scenario requests are the ones that require the most critical thinking and communication. To give you a better idea of a request and the purpose behind creating dashboards, I will give you an example of a scenario.

Dashboard Request Scenario

One day a web designer asked me to pull together a dashboard of the mobile user. She wanted this data to help support her proposal that responsive design (when a webpage automatically adjusts to the screen size of the user) should be included in the plan for the corporate wide website redesign. She needed data that provided her with different computing specifications of the user. The two sites she was concerned with tracking data for were the catalog and the corporate website. So the types of reports that were added included operating systems, monitor resolutions, and screen size. She also added that she wanted to include mobile users.

In her proposal, she wanted to prove that the number of mobile users is increasing and therefore responsive design should be a priority in the redesign initiative. To illustrate whether the number of mobile users accessing the corporate website and catalog was increasing, the reports needed to show a trend (a 6 month period was sufficient for her purposes) instead of rank. Below is an example of the Mobile Screen Size report for the catalog that was included.

Figure 1: Mobile Screen Size Report for the company's catalog

In addition to creating dashboards, I keep an eye out for data anomalies. When running multiple reports, you can spot a trend in one report and have an idea of how the data will be represented in another. When I spot a data anomaly, I report it to my colleague who begins an investigation into the coding. There can be many reasons for a data anomaly such as incorrect tracking coding.

There are also data sets that look like anomalies but are actually correct representations of the data. For example, when I was running reports for a community site I spotted what looked like an anomaly between the "Unique Monthly Visitors Report" and the "Monthly Page Views Report". Unique visitors are individuals who visited the site and are represented one time in the data. So although the same individual may visit the site many times, the individual is counted once for the time frame that is specified in the report. Therefore the unique visitor provides a better representation of the number of people that actually visit a site. The "Monthly Page Views Report" is a count of the total number of times a page was viewed for the site.

Therefore it would stand to reason that the "Monthly Page Views Report" would produce data that closely resembles the "Monthly Unique Visitors Report". However, the two reports for this site couldn't be more different. According to the "Monthly Unique Visitors Report", there were only 16 visitors for the month of April. The "Monthly Page Views Report" shows a total of 700 page views for the month of April. The total number of visitors to the total number of page views seems incredibly high. Each visitor would have to view an average of 43.75 pages from the beginning of April until April 20^th (when the report was actually run).

Figure 2: Monthly Unique Visitors Report

Figure 3: Monthly Page Views Report

This community site was reported and after an initial investigation, it was determined that the data is an actual representation of the site activity. So what was the reason for 700 page views and only 16 unique visitors? Both internal and external visitors were counted. Internal visitors are individuals who access the site through the company's IP address. External visitors are individuals who access the site using an IP address that does not belong to the company. So the 700 page views can be explained in terms of developers working on the site and constantly needing to refresh the page to ensure the code works, object placement, etc.. If a filter were to be included to separate the internal data from the external data, the data would be a better representation of a customer's activity on the site. It is worth mentioning that although this filter is not currently in place, it is in development.

I keep a close eye on any dashboard whose reports have caused me trouble. I keep a list of the dashboard names, the name of the report(s), and schedule of when they are to be sent to the stakeholders. By checking up on the dashboards, this ensures the quality of data and a speedy response to remedy a problem should one arise. It is important that the data is accurate because of the impact it has on decision-making.

The community site dashboard whose "Monthly Unique Visitors" and "Monthly Page Views" reports made me suspicious of data anomalies is one such dashboard that I've checked back on. So as mentioned before, the "Monthly Unique Visitors Report" showed that as of April 20, 2012 there were 16 unique visitors for the month of April. When the report was run again on April 30, 2012, the data showed only 2 unique visitors for the month of April. Data should not be decreasing.

Figure 4: Monthly Unique Visitors Report created April 20th, 2012

Figure 5: Monthly Unique Visitors Report created April 30th, 2012

There can be two possible reasons for the decreased number in unique visitors for April. The first reason is perhaps the filter to separate internal and external visitors was implemented. The second reason is perhaps there is something wrong with the tracking code or platform. After reporting the problem and consulting with my colleague, I discovered that the filter is still in development. So there is no possibility of a filter separating the data. Therefore an investigation into the dashboard and community site has been initiated.

The creation of dashboards and spotting of data anomalies is only the tip of the iceberg of what I do. The internship has been both challenging and rewarding. I was surprised at how much I've relied on the skills I've learned in my courses (LIS 7460, LIS 7490, LIS 7491 in particular). Most importantly the discovery of relating MLIS skills into non-traditional job roles has been particularly valuable and one I hope that other LIS professionals might take into consideration.

Click on the button if you would like more information about the Information Management Certificate and/or the Information Analytics Specialization.

(function(){
var hsjs = document.createElement("script");
hsjs.type = "text/javascript";
hsjs.async = true;
hsjs.src = "//cta-service.cms.hubspot.com/cta-service/loader.js?placement_guid=d3baab63-cd68-481d-8031-99f027f83233";
(document.getElementsByTagName("head")[0]||document.getElementsByTagName("body")[0]).appendChild(hsjs);
setTimeout(function() {document.getElementById("hs-cta-d3baab63-cd68-481d-8031-99f027f83233").style.visibility="hidden"}, 1);
setTimeout(function() {document.getElementById("hs-cta-d3baab63-cd68-481d-8031-99f027f83233").style.visibility="visible"}, 2000);
})();

What do you think? Please add your comments to start a discussion!