<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Jire Nonso on Medium]]></title>
        <description><![CDATA[Stories by Jire Nonso on Medium]]></description>
        <link>https://medium.com/@jirenonso?source=rss-c0e8312f7538------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*ZKTH8ePLr0e4Qd1wwjwOTw.jpeg</url>
            <title>Stories by Jire Nonso on Medium</title>
            <link>https://medium.com/@jirenonso?source=rss-c0e8312f7538------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 06:40:38 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@jirenonso/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Ticket to Resolution: Analyzing IT Support Performance]]></title>
            <link>https://medium.com/@jirenonso/ticket-to-resolution-analyzing-it-support-performance-93e266ab8747?source=rss-c0e8312f7538------2</link>
            <guid isPermaLink="false">https://medium.com/p/93e266ab8747</guid>
            <category><![CDATA[it-support]]></category>
            <category><![CDATA[tickets]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[it-companies]]></category>
            <category><![CDATA[data-analytics]]></category>
            <dc:creator><![CDATA[Jire Nonso]]></dc:creator>
            <pubDate>Sat, 03 Oct 2026 05:23:30 GMT</pubDate>
            <atom:updated>2026-10-04T18:10:03.814Z</atom:updated>
            <content:encoded><![CDATA[<p><em>How I analyzed an IT performance data to investigate ticket volume, recurring issues, support queues, and resolution patterns in an IT support environment.</em></p><p><strong>Every IT support team has the same fundamental challenge:</strong></p><p><em>A ticket comes in. Someone needs to solve it.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/693/1*RpGxCLl2ibvR9nXoIOCKZg.png" /></figure><p>But between those two points is an entire operational story.</p><p>Some tickets are resolved quickly. Others remain open for days. Some issues appear repeatedly. Certain support queues may carry significantly more workload than others. And behind every unresolved ticket is a user waiting to get back to work.</p><p>This raised an interesting question for me:</p><p><strong>What can ticket data tell us about how an IT support operation is performing — and where might there be opportunities to improve it?</strong></p><p>That question became the foundation for my project, <strong>“Ticket to Resolution: Optimizing IT Support Performance.”</strong></p><p>Using a dataset of more than 1,000 Jira Service Management-style support tickets, I built an Excel-based analytical dashboard to explore ticket volume, resolution patterns, recurring issues, support queues, and operational workload.</p><p>But the goal wasn’t simply to build a dashboard.</p><p>The goal was to use data to understand the story behind the tickets.</p><h3>The Business Problem</h3><p>IT support teams operate in a high-volume, service-oriented environment with constant influx of requests, high ticket volumes and recurring technical problems received on a day-to-day basis. This can be overwhelming, hence leading to increased workload and frustration. This project aims to analyze IT data in order to resolve tickets faster, improve user productivity and enhance proactive prevention of recurring issues.</p><h3>Data Structure &amp; Initial checks.</h3><p>The dataset contained 11,923 records and 14 fields, spanning across a period of one and a half year (1/1/2024–31/05/2025). The fields included ticket IDs, creation and resolution dates, ticket descriptions, priority, support queue, ticket type, and category-related tags.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S2_zq4feOh0sFpE7gVU_gg.png" /><figcaption>IT Dataset.</figcaption></figure><p>I examined the structure of the dataset, checked the fields and considered how each variable could contribute to answering the business questions. I also defined the grain/granularity of the data(level of detail): which represented a unique ticket made by a customer, their location, the time at which it was created and resolved, as well as the support queue that handled the ticket (with metadata about that particular ticket, i.e. primary tag, category tag, documentation tag, etc).</p><p>The first step wasn’t visualization.</p><p><strong>It was understanding the data.</strong></p><h3>Methodology</h3><p>The analysis followed a structured data preparation, validation, transformation, and visualization process using <strong>Microsoft Excel</strong>. The objective was to transform the raw data into a reliable analytical dataset that could be used to evaluate resolution patterns, recurring issues, support queue performance, ticket priorities, and workload distribution.</p><h3>Data cleaning &amp; Preparation</h3><p>Before I began analysis, I made sure the data was free of inconsistencies that could skew analysis and insights because a bad data is equivalent to bad analysis and thats not something we want here.</p><p>I cleaned and transformed the data using Power query. I chose this tool because it wasn’t just appropriate for large datasets like this but also supported automated cleaning tasks, so if there was new data to be updated, it would simply follow an automated cleaning process to avoid time wasting rather than being approached manually.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/235/1*vcRxJas2lKQCcLBUGc4wMg.gif" /></figure><p>Then, I began profiling the data to check for nulls, errors, inconsistencies across every field. I also trimmed the data to ensure that there were no extra spaces; the data types for each column was also inspected properly to avoid wrong formatting of the data. No duplicates were also found as all entries were unique and belonged to 11,923 customers. There were a lot of inconsistencies in some of the fields, with some represented as upper case, lower case,etc. in a single field. I handled this by standardizing them into one form (i.e. lower case), this was because “USA” is different from “US” even though they’re the same thing, the program would treat them as two seperate entities which could skew or interpret false results during analysis; So i standardized them into “USA”.</p><p>Another challenge i encountered while cleaning was — some fields had more than one entry, one field could contain 4 to 7 entries seperated by a comma and i noticed this pattern throughout the data. For instance, one entry could have “Billing”, “Feature”, “technical”, “Social media, “digital” repeated throughout the dataset. So I filtered the data to identify all the entries that had this issue before splitting them according to their appropriate columns so “Billing, feature, technical, socialmedia, digital) are split into columns — Primary tag, secondary tag, category tag, Technical tag, Status tag). This same step was repeated for the other entries.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jlBSkalP8_y8QZ6_D0jN8A.png" /><figcaption>Messy data with fields having more than one entries.</figcaption></figure><p>After this was done, some entries had null because there was insufficient data to split the remaining text to the last column. I also found anomalies e.g. additional tag recorded “2019” which was inappropriate and was considered an error. I made sure they were all handled to ensure data integrity and improve the quality of the data because good data is equivalent to good analysis.</p><h3>Analysis &amp; Approach Basis</h3><p>Here, I established that a typical support operation might receive tickets ranging from:</p><ul><li>Software bugs</li><li>Feature requests</li><li>Technical issues</li><li>Access requests</li><li>Billing questions</li><li>Service outages</li><li>Performance problems</li><li>General inquiries</li></ul><p>At first glance, a simple ticket count can appear useful.</p><p>But ticket volume alone doesn’t tell the whole story.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*t3JDeYBUiNyz8xnTRmg57Q.gif" /></figure><p>Imagine two support queues:</p><ul><li>Queue A handles 500 tickets.</li><li>Queue B handles 200 tickets.</li></ul><p>It would be easy to conclude that Queue A is performing worse because it has more tickets.</p><p>But what if Queue A resolves most of its tickets within a day, while Queue B handles fewer tickets but has a much longer resolution cycle?</p><p>The data needs context.</p><p>That led me to structure the analysis around four business questions:</p><h3>1. How can we solve tickets faster?</h3><h3>2. What issues are recurring most frequently?</h3><h3>3. Which types of requests require the most time or effort?</h3><h3>4. Where can workflows be improved to support better customer experiences?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*qP438huoLI8SdiKBAsBf7A.gif" /></figure><p>Then I determined the key performance indicators: Average resolution time, Total tickets volume, % of high tickets (i.e. tickets that were attributed to high priority and expected to be resolved in a relatively shorter time frame).</p><p>One of the most important derived metrics was <strong>No. of days to resolve.</strong></p><p>This was not originally in the dataset so I feature engineered this field by subtracting the creation date (i.e. date the ticket was created) from the resolution date (calculated expected date the ticket would be resolved).</p><p><strong>Days to Resolve = Resolution Date − Creation Date</strong></p><p>This provided a basic measure of how long tickets remained open.</p><p>From there, I began familiarizing with the dataset by exploring the relationship between:</p><ul><li><strong>Ticket volume</strong></li><li><strong>Resolution time</strong></li><li><strong>Support queues</strong></li><li><strong>Priority</strong></li><li><strong>Ticket types</strong></li><li><strong>Categories</strong></li><li><strong>Tags</strong></li></ul><p>The objective here was to move from, <strong>“What happened?”</strong> to <strong>“Why it happened?”</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*suHst5pc1IBcZ6etMdE2dw.gif" /></figure><h3>Key Finding 1: Seasonality in Ticket volumes</h3><p>To understand whether support demand changed across the year, I visualized monthly ticket volumes using a line chart and compared the patterns across 2024 and 2025.</p><p>A recurring seasonal pattern emerged. <strong>Ticket volume dropped by 12.6% in February 2024 before increasing by 13.8% in March</strong>, after which volumes remained relatively stable through the following months. June recorded the highest ticket volume overall, with 735 tickets<strong>,</strong> indicating a period of particularly high support activity.</p><p>Interestingly, the same February–March pattern appeared again in <strong>2025</strong>. February experienced another decline in ticket volume, followed by a noticeable increase in March. This recurrence across two consecutive years suggests that the changes were not isolated to a single year and may reflect a recurring seasonal pattern in support demand.</p><p>Ticket volumes also showed a moderate decline toward November and December, indicating relatively lower support activity during the latter part of the year. Overall, the data suggests that Q1 and parts of Q4 generally experienced lower ticket volumes compared with the higher-activity periods in the middle of the year.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/642/1*VgtnHLwzgwcbV4FCfHt7fA.png" /><figcaption>Seasonality of trends in Ticket volumes</figcaption></figure><p>The recurring pattern provides a useful basis for support capacity planning. Periods with historically higher ticket volumes, particularly around <strong>June</strong>, may require greater staffing capacity and resource availability, while lower-volume periods could provide opportunities for teams to focus on backlog reduction, documentation, process improvements, or other operational activities.</p><h3>Key Finding 2: Ticket distribution across support queues.</h3><p>One of the things I examined was the distribution of tickets across support queues. First, I considered the metrics (ticket volumes) and key dimensions (Ticket type and support queue) before pivoting the table for analysis. I also considered using conditional formatting for this visual so it any non-technical stakeholder would see how that distribution is spread out across support queues. The data showed that some queues handled substantially more tickets than others.</p><p>For example, <strong>Technical Support accounted for approximately 28.6% of tickets</strong>, followed by Product Support and Customer Service.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/607/1*s5jr0GKx2qFEdfJGsNv1XA.png" /><figcaption>Conditional formatting showing the distribution of ticket types and volumes across support queue.</figcaption></figure><p>This immediately revealed the workload on major support queues like Technical support and Product support.</p><p>But there was an important analytical caution:</p><p><strong>A queue handling more tickets does not automatically mean that the queue is performing poorly.</strong></p><p>High ticket volume may simply reflect the nature of the issues that team is responsible for.</p><p>This distinction is important because analytics should help us ask better questions rather than encourage premature conclusions.</p><p>Instead of saying:</p><p>“Technical Support is the problem because it has the most tickets.”</p><p>A more appropriate question is:</p><p><strong>“Does the workload handled by each queue correspond with differences in resolution time, ticket complexity, or recurring issue types?”</strong></p><p>That is a much more meaningful analytical direction.</p><h3>Key Finding 3: Recurring Issues Can Reveal Opportunities for Prevention</h3><p>Another important part of the analysis was understanding the categories and tags associated with tickets.</p><p>The <strong>Category Tag</strong> analysis showed that the most frequently occurring categories included:</p><ul><li>IT</li><li>Performance</li><li>Tech Support</li><li>Feedback</li><li>Bug</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/440/1*oPzn52CyJQLoxBZlEM97nw.png" /></figure><p>The top three categories alone accounted for approximately <strong>39% of the categorized tickets</strong>, while the top five represented nearly <strong>48%</strong>.</p><p>This was particularly interesting.</p><p>If a relatively small number of issue categories account for a substantial proportion of support activity, there may be an opportunity to investigate whether some of those issues can be prevented rather than repeatedly resolved.</p><p>For example, recurring issues could potentially inform:</p><ul><li>Knowledge-base articles</li><li>Self-service documentation</li><li>User training</li><li>System improvements</li><li>Product fixes</li><li>Preventive monitoring</li><li>Automated troubleshooting</li></ul><p>This is where support analytics moves beyond measuring workload.</p><p>It begins to support <strong>problem prevention</strong>.</p><h3>Recommendations</h3><p>The following recommendations were given based on the key findings above, we would recommend the IT support team &amp; management to consider the following:</p><ul><li><strong>Support demand consistently increased after February and reached its highest point in June.</strong> The support team should use historical ticket patterns to plan staffing and resource allocation ahead of expected high-volume periods, particularly around mid-year, while using lower-volume periods for backlog reduction and process improvement.</li><li><strong>Technical Support, Product Support, and Customer Service handled the largest share of tickets.</strong> Management should review workload distribution across these queues and assess whether staffing, escalation procedures, or knowledge resources are sufficient to prevent bottlenecks during periods of high demand.</li><li><strong>IT, Performance, Bug, Technical Support, and Outage tags appeared frequently across the dataset.</strong> The team should investigate the underlying causes of these recurring issues and prioritize preventative measures such as system improvements, troubleshooting guides, knowledge-base articles, and user self-service resources.</li><li><strong>Ticket volumes showed recurring increases and decreases across the same periods in different years.</strong> Support planning should incorporate historical seasonality rather than relying only on current ticket volumes, allowing teams to prepare capacity and resources before predictable changes in demand occur.</li><li><strong>Resolution time was closely aligned with ticket priority in the dataset, limiting its usefulness as a standalone performance measure.</strong> Future data collection should capture actual timestamps for ticket creation, first response, assignment, escalation, and resolution so that metrics such as first-response time, mean time to resolution, and SLA compliance can be measured more reliably.</li><li><strong>The analysis identified workload patterns but could not fully explain the root causes behind every recurring issue.</strong> Future analysis should combine ticket data with additional information such as customer segment, product/system affected, escalation history, SLA status, and resolution category to identify the operational factors driving support demand.</li><li>The organization could establish a recurring reporting process that tracks ticket volume, queue workload, recurring issues, resolution performance, and seasonal trends over time, making it easier to identify emerging problems before they become larger operational issues.</li></ul><h3>Key Takeaways</h3><ul><li><strong>Ticket volumes followed a recurring seasonal pattern.</strong> February recorded lower ticket volumes in both 2024 and 2025, followed by increases in March, while June recorded the highest overall volume at <strong>735 tickets</strong>. Ticket activity also declined toward November and December.</li><li><strong>A small number of support queues handled a large share of the workload.</strong> Technical Support accounted for <strong>28.62%</strong> of all tickets, followed by Product Support at <strong>18.72%</strong> and Customer Service at <strong>15.59%</strong>. This concentration highlights where much of the support workload was being handled.</li><li><strong>Recurring issue categories represented a significant portion of support activity.</strong> The most common tags included <strong>IT, Performance, Bug, Tech Support, and Outage</strong>, suggesting that recurring technical and performance-related issues contributed substantially to the overall ticket workload.</li><li><strong>High ticket volume did not necessarily mean the same thing as high resolution performance.</strong> Comparing support queues and resolution days showed differences in ticket-handling patterns, but the analysis also revealed that the resolution metric was closely associated with the priority values in the dataset.</li><li><strong>The data contained an important limitation that affected performance interpretation.</strong> High, Medium, and Low priority tickets consistently corresponded to approximately <strong>2, 3, and 4 resolution days</strong>, respectively. This suggests that the resolution-time field may have been generated or structured based on priority, limiting its usefulness as an independent measure of actual support-team efficiency.</li><li><strong>The analysis demonstrates the value of combining workload, issue, queue, and time-based analysis.</strong> Looking beyond total ticket counts provided a clearer picture of where support demand was concentrated and where operational improvements could potentially have the greatest impact.</li></ul><h3>Limitations</h3><p>Perhaps one of the most valuable lessons from the project wasn’t a chart.</p><p>It was a limitation in the data itself.</p><p>While analyzing <strong>Days to Resolve</strong>, I noticed an unusual pattern when comparing resolution time against priority.</p><p>The resolution-day values appeared highly constrained by priority, with observed values clustering around:</p><ul><li>High priority — 2 days</li><li>Medium priority — 3 days</li><li>Low priority — 4 days</li></ul><p>This raised an important question: “<strong>Is the resolution-time metric actually measuring operational speed?”. </strong>If the resolution duration is derived from a process that effectively assigns or determines resolution time based on priority, then it may not provide an independent measurement of actual support performance. That means a simple statement such as: “High-priority tickets are resolved faster.” could be misleading.</p><p>Another limitation was that the analysis could not account for a full year-year comparison because the dataset did not span two complete years.</p><h3>Final Thoughts</h3><p>This project started with a simple dataset of support tickets.</p><p>But the deeper I went, the more I realized that the interesting part wasn’t the tickets themselves.</p><p>It was the <strong>story behind them</strong>.</p><p>Every ticket represents a problem.</p><p>Every category represents a pattern.</p><p>Every queue represents a workflow.</p><p>Every resolution time represents a process.</p><p>And every pattern in the data creates an opportunity to ask a better question.</p><p>That is what I’m increasingly learning to appreciate about data analytics:</p><p><strong>The goal isn’t simply to find numbers.</strong></p><p>It’s to understand what those numbers represent, question the assumptions behind them, identify meaningful patterns, communicate them clearly, and ultimately help people make better decisions.</p><p>And sometimes, the most valuable insight isn’t an answer.</p><p>It’s discovering that we need to ask a better question.</p><p>Here’s a view of the dashboard built using Excel:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/917/1*U52X5_FHFtepRLyBRzOYAw.png" /><figcaption>IT Performance Dashboard.</figcaption></figure><p><strong>Tool &amp; Skills used:</strong> Microsoft Excel, Power Query.</p><p><strong>Techniques:</strong> Data cleaning, exploratory data analysis, Pivot tables, slicers, KPI development, Dashboard development, Conditional formatting, Data analysis, Stakeholder communication, Data Story-telling.</p><p>I’m continuing to build my skills by working on projects that connect data analysis with real-world business and operational problems.</p><p>Feel free to connect with me and lets talk data!</p><p>Linkedin: <a href="http://www.linkedin.com/in/jirenonso">www.linkedin.com/in/jirenonso</a></p><p>Email address: jirenonso@gmail.com</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=93e266ab8747" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>