Further to this blog post where we discussed 13 points to think about during your automation journey, the foremost of them was understanding the importance of data. In this post, we will discuss how to use data to decide ROI and roadmap for your IT Automation journey. We will start with data collection, data sanitization and data analysis. We will also discuss Anakage ITSM Log Analyzer tool and see how it can make this process simpler.

IT Infrastructure is complex for any decent size company. The introduction of any new tool or automation initiative to an organization gives a nightmare to the team. Besides solving a few specific problems, any change also brings maintenance, support, and compliance overheads it adding to the total cost of ownership. Before starting any automation journey you should ask below questions

  1. What is the issue with the current setup?
  2. What are we losing if we do not bring in this tool or do not think of automation?
  3. Will it improve our IT Team KPIs like the number of tickets resolved, number of tickets reduced, time saved to resolve a ticket, time saved for servicing a request, time saved by users to resolve a ticket?
  4. Will the change increase the workload of IT Team vs ROI?

This post will help you discover how data can be used to define ROI or roadmap of your IT automation journey. To answer above questions, we will follow 3 step process –

Collect data –

Before collecting data refer your company or client policies regarding how you can use these data. Find out if –

  1. Whom you can share these data with? If they are vendors are they covered by Non-Disclosure Agreement.
  2. Can you take this data outside the environment?
  3. Are you allowed to use online tools to analyze data? If yes then doubly make sure that the data is sanitized properly. You do not want your organization/client data to start appearing in search results or start showing in logs. Refer terms of use and privacy policy of the website/tool regarding data handling.

As part of your data collection be aware of seasonality of tickets. In any organization incidents and ticket requests are always changing due to

  1. Change management initiatives like introduction of new tools, printers, processes
  2. Issues from OEM side and their patches
  3. Cycles like windows 10 migration
  4. Process cycle such as appraisals, people joining organization and need of onboarding.

The automation journey should be aware of these seasonality’s. Selection of timeline of data which has been taken for analysis is very important. If Windows update cycle has happened 6 month ago so do not take data for 6 months since Windows related issues will dominate in past but they are not present now. The automation roadmap should also think in future, how quickly can we take care of those issues.

IT Team uses lot of tools. Contact tool owners and find out what kind of data you can get from them.

  1. Incident and Service Request data from ITSM tool is a reliable source.
  2. Data and Insights from your existing tools like event viewer logs, monitoring tool logs, Anti-Virus logs, Product Licensing logs etc.
  3. Collate all the data from past new tool rollout phases. They could be initial teething issues, user queries on mail/chat etc., time to train IT Team on the tool, time to onboard users on new tool
  4. Any other data you can get hold on.

Sanitize the data –

You will realize that you have lot of data and might become overwhelmed by it. IT Team core competency is not data so start small. Based on the results you can increase the scope of your data and results. Follow below steps –

  1. Create a new excel sheet.
  2. Since original data sources will have lot of extra information so remove all the redundant data that is not needed for analysis. Columns like data, time, person, teams, category, subcategory etc can be removed. Only keep problem and resolution notes to begin with. Your aim is to find out what are the topmost issues where you spend time and how was it solved which will give you an idea about how it was solved.
  3. Remove all the Personally Identifiable Data (PII)
  4. Remove all the confidential and client specific data,
  5. Try to fit in other sources of data in same format. For example – if you have data from chat keywords put them in problem description column, if you have data from monitoring tools provide a description for them for patterns they are showing.

Analyze data –

There are lot of basic tools which can be used to quickly find if the data is showing any actionable pattern. You can proceed in following way –

  1. Manually understand each row and tag with comments in excel. Create another sheet where you write down issues and frequency in different column. Problem with this approach is that there are thousands of rows in any data which is primarily based on ITSM logs. This will take lot of your time.
  2. Another approach is to use word frequency counter or word cloud generator tools to find the frequency of words in your data. You will discover that it will give unexpected patterns and high frequency patterns you can use to do filter in Excel and focus on those rows. But you will have to do another round of sanitization on data before this step. Words like It, have, are and a whole category of words known as “Stop Words” – more can be found here
  3. Alternatively you can use Anakage ITSM Log Analyzer tool (this is in beta stage and access is provided on request) which we will cover in next post. It goes beyond plan search, categorize and find frequency. It uses AI/ML to find patterns/sub patterns and business rules. It also understand IT jargons so you do not have to train it before using. And with each use it continuously keeps retraining itself. This tool takes ITSM logs as input and provides a detailed report after analyzing data automatically. You will get reports like –
    1. What are most frequently occurring issues in your organization along with numbers.
    2. How many of them can be resolved by Anakage Platform and detailed break up.
    3. If you extend Anakage platform and customize it for your organization, how will final number be.
    4. Based on these data you can quickly get insights like – What is ROI of tool, What is roadmap of deployment, How much can be cover from day one and in future updates, what is cost saving, how much of admin time can I save, what is impact on user time to solve a issue, what are processes that are not in Anakage platform scope but should be part of my roadmap for Automation.

    After this exercise your IT automation journey will no more be black box or dependent on story telling skills of anyone or futuristic picture shown by a marketing heavy product. It will give you more realistic picture of ROI and roadmap for future which you can predict and track.

    The same thought process you can extend for other cases like –

    1. Need of any new tool. Validate its values and ROI unless ROI promised by the tool cannot be measured using the given data set.
    2. Identify/Validate processes that can be automated
    3. Identify/Validate processes that should go through Root Cause Analysis process
    4. Validate assumptions or filter out ideas that are collected out of any innovation camp.
    5. Rule out Human bias during decision making process
    6. Set realistic and measurable goals

    Contact us to help you discover or finalize your IT Automation Plan.



    Leave a Reply

    Avatar placeholder

    Your email address will not be published. Required fields are marked *