Master the Mainframe

08.06.201316 Min Read — In Mainframe

Update: I’m pleased to announce that the IBM Academic Initiative Team reached out to me as a result of this post, and many of the issues that this articles brings up are being addressed. The search engine optimization (SEO) issue is currently being fixed in anticipation of the launch of several key 2013 Master the Mainframe contests, including North America. Part of the issue turns out to have been related to a recent move between hosting providers, which any webmaster knows is killer for SEO. Nonetheless, the leadership is committed to getting this resolved ASAP. I’ve also been invited onto the Master the Mainframe planning committee to help drive participation in the contest and brainstorm about possible improvements. My personal goals are to drive greater participation from schools that have hitherto had little to no penetration by partnering with millennial mainframers that have recently graduated from those schools, as well as trying to identify inhibitors that prevent students that finish Part 1 from moving onto Part 2. I am also hoping to find novel ways to use the Millennial Mainframer site to unite the different elements of the mainframe education ecosystem into a cohesive whole: e.g. mentor and encourage Master the Mainframe students, help them connect with mainframe education, help them find internship opportunities, get them to blog to build their personal brand and thought leadership, help them land jobs, and then keep them engaged as mentors.

This is the third post in our five-day series on Mainframe Education. On Monday of last week, we posed the question, “Do Millennial Mainframers actually exist in the flesh in any real numbers or are they the Unicorn or Loch Ness monster of the IT World… much talked about, but little seen in the wild?” On Tuesday of last week, we looked at the success of our own blog in reaching out to Millennial Mainframers around the world, and we made some stark conclusions, including that our mainframe content was fundamentally failing to reach a millennial viewing audience. After taking a break to demonstrate how to optimize a MacBook for use with 3270 terminal emulation, today’s post continues the five-part series to examine one of the more well-known efforts to promote the mainframe among millennials: the IBM Master the Mainframe Contest.

When millennials begin to investigate a topic for the first time, they often begin by executing a Google search query. Web developers understand this well, which is why the search advertising and SEO industries have become so lucrative. It ts therefore troubling that searching the phrase “Master the Mainframe” produces jumbled results. While one might expect the official website of the IBM Master the Mainframe contest to be the top search result, the official page doesn’t actually show up in the top Google search results at all.

Top Search Results for Phrase “Master the Mainframe”

  1. IBM Press Release entitled “IBM Master The Mainframe Contest Prepares Students With In Demand IT Skills For A Smarter Planet.”
  2. Wikipedia Article for “Master the Mainframe”
  3. Master the Mainframe Facebook Page
  4. Master the Mainframe Australia Facebook Page
  5. Master the Mainframe Twitter Page
  6. IBM-sponsored blog “Building a Smarter Planet” hosting an post written by Master the Mainframe winner Miles Nosler entitled “How Wrangling Big Data Helped Me ‘Master the Mainframe’
  7. A short Dataversity post called “IBM Announces Winners of Master the Mainframe Contest”
  8. Millennial Mainframer’s Forum for discussing Master the Mainframe. **heh heh**

Wow. If I were a contest participant with the average Millennial’s expectation of instant gratification, I would start getting frustrated. While these resources describe the Master the Mainframe Contest and IBM’s efforts to combat the mainframe skills shortage, none of these links easily channel a student to a link to sign up for the contest. For goodness sake, even that crappy Millennial Mainframer blog pops up in the top ten before any official IBM Master the Mainframe Contest.

When researching this article, my research began with this exact problem. Mainstream search engines could not locate the official IBM Master the Mainframe Contest page, forcing me to perform fifteen minutes of old-school AltaVista-style URL jiu jitsu to get where I wanted to go. By rephrasing my query as “Master the Mainframe 2012,” the DeveloperWorks article on Master the Mainframe – India 2012 popped up in spot six. Deep in that page, I found the contest link to the Master the Mainframe India site, which gave to enough data to use my advanced computer science training to reason out the hierarchical URL structure of the Master the Mainframe site. Based on the fact that the Indian site was at, I reasoned that I could chop off the India sub-directory to reach a landing page to pick a country. Bingo! I made it to and reached this page:

Screenshot of the IBM Master the Mainframe Landing Page

The fact that it took fifteen minutes of advanced web searching techniques to get to the proper page for the Master the Mainframe contest is a huge problem. I suspect that an average millennial showing a modicum of interest in this contest would have likely given up by this point. Perhaps I need to offer some IBMers the Sean McBride: Master the Internet Contest. The top prizes include search engine optimization and increased exposure to your contest. After all, if a bozo like me spending $4 per month on web hosting can out-SEO a multi-billion dollar corporation like IBM, something is probably wrong. But I digress…

The Sun Never Sets on the Master the Mainframe Contest

Once on the actual Master the Mainframe site, things look great. This landing page lists out all of the contests for 2013 and the past contests for 2012. Apparently, we’re already getting into Master the Mainframe season, as the first 2013 contest was Poland, which wrapped up on June 2nd. This facet illustrates an interesting recent development in the Master the Mainframe Contest: it’s gone global. Although the Master the Mainframe contest was originally designed as a means to support the development of mainframe education in rich-world areas, like the United States, Canada, and Europe, it has since spread out to other areas, including out-sourcing destinations, like India, which already have significant numbers of mainframe academies and training centers. On this world map, the countries highlighted in red participate in some form of Master the Mainframe Contest.

A world map illustrating the counties that participate in the Master the Mainframe Contest

This move to expand the contest globally seems a logical choice. IBM mainframes have always been a global phenomenon, and recent sales to new customers in Brazil and China have increased the need to skilled mainframers in those areas. The contest is probably not needed in India, as there are already numerous mainframe schools there, like this one in Chennai, but for the sakes of equity, they might as well be able to compete as well.

Despite this globalization, the Master the Mainframe Contest follows the same format in each of the competitions around the world. The contest gives high school and college students the chance to solve a variety of systems programming and application development tasks on the z/OS mainframe operating system (there are no such contests for other mainframe operating systems, like z/VM, z/VSE, or z/TPF).

The contest is broken into three parts of increasing difficulty:

  • Part One essentially involves downloading a 3270 emulator, connecting to a mainframe, logging into the TSO/ISPF environment, customizing ISPF user preferences, locating datasets, creating members in a dataset, entering data into a member using ISPF editor, using prefix commands to copy and move text in ISPF Editor, running a REXX script that counts the lines in your dataset member, and then submitting that report to IBM for automatic grading. The key lessons of part one are that although the mainframe uses a very different sort of interface from other computers, it is still a computer that has core concepts like a file system (albeit a very different one than Linux or UNIX or Windows), an editor, facilities for running custom scripts, and ways of connecting to other computers over the internet. Everyone that completes this basic mainframe orientation successfully receives a very swanky mainframe t-shirt.
  • Part Two shifts to a series of actual tasks, including submitting pre-written JCL and observing the status of the batch job, fixing basic JCL errors, learning more advanced ISPF Editor functionality, FTPing into a mainframe Linux environment to transfer files to an z/OS dataset, using JCL to submit a SQL statement to DB2 for z/OS, writing small programs in C, modifying a REXX script, creating an ISPF panel, examining system details through the OMEGAMON XE performance tool, going into z/OS UNIX to compile and run a Java program. These Part Two tasks do a good job of giving examples of day-to-day tasks in z/OS, but they also emphasize that the skills that students actually learn in school are still applicable to the mainframe, including SQL and relational database theory, programming languages like C or Java, UNIX and Linux operating environments, basic computer scripting (REXX isn’t all that different from Python or Ruby), and tab-based presentation layers (ISPF panels don’t look all that different from static HTML). Typically, the contest participants that complete this part get some sort of prize pack, which has traditionally included hoodies, bumper stickers, and that sort of stuff.
  • Part Three of the Master the Mainframe contest is quite a bit more intensive, as it simulates problems similar to those experienced by actual systems programmers. The scope of this project requires between several weeks and months of effort, often involving more advanced mainframe tools, such as the Rational Developer for System z Eclipse environment. Given the significant amount of time required for this portion of the contest, the prizes are also quite a bit better than other parts. The top five Part Three submissions typically win a tablet computer, such as an iPad or an Asus Transformer. In the North America competition, the top three winners also get a all expenses paid trip to beautiful Poughkeepsie, New York. Woo Hoo!

Master the Mainframe Successes

From my perspective, the contest has been fairly successful in a three key areas. Firstly, the contest gives students temporary hands-on exposure to a mainframe system. Given the fact that hardly any universities offer mainframe courses, this is a critical avenue for young IT professionals to investigate whether mainframe employment may be a good fit. Secondly, IBM seems to understand that the mainframe cannot successfully “sell itself” to young technologists. Given the fact that most millennials have no idea what a mainframe is, attracting millennial talent to mainframes requires more than just providing z/OS login credentials and shooting out corporate-style education over e-mail. Thankfully, the Master the Mainframe team has done a pretty good job at leveraging social media to reach out to the IBM Academic Initiative schools. On Facebook, they have 2,530 likes, many of whom are actually in the target 18-24 demographic. Additionally, IBM’s willingness to turn their flagship mainframe outreach program into a contest with tablets and mainframe swag reflects a pretty good grasp of the concept of “gamification.” The application of techniques explain why this contest has successfully grown to 4,600 participants in the 2012 North America Contest. Third, as someone that has completed Parts One and Two of the Master the Mainframe contest within my first year of learning mainframes, I can attest that the author of this contest has done a fantastic job at explaining mainframe concepts in ways that millennials are likely to understand. The instructions use terms like “memory” and “disk” rather than “central storage” and “DASD,” which may seem trivial to long-time mainframers, but is actually a great boon to millennial comprehension. Additionally, the tone of the contest seems to be quite goofy and fun, more like O’Reilly’s Head First Java than some of a ponderous IBM manual, information center, or redbook. Additionally, the exercises aren’t too easy or too hard. They seem to be well tailored to building skills developed in preceding modules to stretch the minds of young technologists. Based on these strengths, I consider the Master the Mainframe Contest the gold standard of hands-on introductory mainframe training.

Now to the issues…

Despite these advantages, I think that there are a number of issues with this contest. Many of these issues relate less to the contest itself, and more to how it is marketed and integrated into larger IBM Academic Initiative efforts.

At a high level, the Master the Mainframe contest plays the role of recruiter. It is intended to give high school and college students a fun and non-intimidating first exposure to mainframes in order to channel them into IBM Academic Initiative courses and mainframe jobs.

Ideal Flow of College Students into Mainframe Jobs

  1. Student hears about Master the Mainframe Contest from friends, professor, internet
  2. Student signs up for the Master the Mainframe Contest
  3. Student completes Part 1
  4. Student completes Part 2
  5. Student may or may not complete Part 3, due to difficulty and time commitment
  6. Student decides to sign up for mainframe course based of experience with Master the Mainframe.
  7. Student completes mainframe coursework and signs up for systemzjobs website
  8. Student gets hired into mainframe position.

The issue with this model lies in any roadblocks to moving engaged end-users through this channel. Let’s step through each of these steps to consider potential problems.

How effective is the Master the Mainframe Contest at exposing high school and college students to the mainframe?

One of the most important elements of this model is making students aware of this contest and signed up to compete, which is fundamentally a marketing effort. In the North American Master the Mainframe contest, around 4,600 folks were sufficiently engaged to the point of enrolling in the contest. It’s probably safe to assume that around two thirds of the students that sign up are in community college, undergraduate college, or grad school. The remaining one third of students are likely either high school or technical trade schools.

While most of these students likely completed the hour-long Part One to earn their cool mainframe T-Shirt prize, it seems that many students failed to proceed to the ten-hour Part Two. Based on the North America Part Two Wall of Fame, only 8.6% of contest participants (397) actually complete part two of the contest. It’s unclear what happens to these other 4,203 participants, but this number likely includes people that sign up for the contest but then don’t do anything, people that just wanted a Master the Mainframer t-shirt, and students that were unable to complete the Part Two exercises.

Looking more closely at the list of Part Two winners leads to a number of other realizations. Most striking was the fact that Master the Mainframe contest participation is definitely dominated by a few key schools. Although 98 educational institutions had students that completed Part Two of the Master the Mainframe contest, ten of these schools are responsible for two thirds of all Part Two winners. Even more surprising is that fact that Lake Brantley High School is responsible for 186 contest participants that complete Part Two, which totals nearly half of all participants at this level. CSU Dominguez Hills, Georgian College, and the University of South Carolina also have 27, 22, and 14 participants respectively, suggesting that IBM Academic Initiative courses at this schools require participation in the IBM Master the Mainframe contest. All of the other 94 schools had quite a few less students on the Part Two Wall of Fame. 64 schools had only one student on this list, and the other schools had between 2 and 4 students. Out of the 98 educational institutions on the Part Two list, 18 were high schools, 1 was a adult education center, 3 were community colleges, and 75 were college/universities.

While it is impressive that 208 high school students completed Part Two of the Master the Mainframe Contest, this means that only 189 college students in North America completed Part Two of the contest. By comparison, a recent CRA reports that the United States and Canada have around 67,850 undergraduates majoring in Computer Science or Information Technology programs.

When looking at the university names, it’s also clear that students at top Computer Science institutions are not enrolling in the contest. Out of the 75 colleges with students that made it onto the Part Two Wall of Fame, only three programs ranked on the US News’ list of top 50 Computer Science programs (26 of which are in North America). These schools are UT Austin (with two in Part Two), UC San Diego (with two in Part Two), and New York University (with only one in Part Two). This suggests that students at MIT, Stanford, Carnegie Mellon, Berkeley, Harvard, Princeton, Cal Tech, UCLA, Cornell, University of Toronto, Columbia University, U Washington, Urbana-Champaign, U Chicago, Yale, and other elite institutions likely have no idea about what a mainframe is or the critical role they play in the global economy, let alone ever being willing to consider employment in a mainframe job.

The Master the Mainframe site also lists the Part Three winners. A total of 28 students completed part three, which is a small 0.6% of those that signed up for the contest or around 7% of those that completed Part Two. 24 of these students were in a four-year college program or graduate program, none of these students were in community college, and four students were in high schools or non-college adult education schools.

How effective is the Master the Mainframe Contest at channeling college students into mainframe employment?

One might think that the students that completed Part Three might be the most likely to pursue mainframe positions, given the fact that they spend 100+ hours on the Master the Mainframe contest. However, this is not necessarily the case. While most of the contest winners are still students, a LinkedIn search suggested that some of the students that completed and even won the Part Three contest have chosen not to accept mainframe positions. Out of the winners of the 2012 contest, it appears that at least two have chosen not to accept mainframe employment. That means that even an IBM-funded trip to Poughkeepsie may not be enough to lure students to accept mainframe positions. #gasp. Looking at other contents participants that completed Part Three, there are clear examples of student technologists that decided to work with other technology companies, such as I think it’s fair to assume that the willingness to consider mainframe employment is directly related to a student’s dedication to push forward in the Master the Mainframe contest, suggesting that the students that only completed Part One are less likely to pursue or gain mainframe employment than those that completed Parts Two or Three. This suggests that only a small portion of Master the Mainframe contest participants eventually accept mainframe employment.


While the Master the Mainframe Contest is a masterfully designed training program for millennials, the contest seems to fail to attract sufficient millennial talent to replace the numerous senior mainframers that will be retiring from IT departments in the near future in North America. Based on the significant drop of participants from Part One to Part Two and from Part Two to Part Three, I would guess that less than fifty contest participants eventually make it into mainframe employment in any given year. This would equal about 1% of all contest participants. Perhaps this is due to lack of interest, but I suggest that this may have to do with many of these college students being unable to gain additional mainframe training beyond the Master the Mainframe contest due to the limited number of colleges offering mainframe coursework. After all, is a few weeks of Master the Mainframe experience really sufficient to be hired as an operator or systems programmer? Further troubling is the fact that top ranked computer science research schools seem wholly unrepresented in the Master the Mainframe Contest. Assuming that these students are more likely to exert influence on the direction of the IT profession, this may be a troubling precedent.

Based on my experiences searching the term “Master the Mainframe” on Google, it’s clear that there is substantial room for improvement. First and foremost, IBM should create a centralized Master the Mainframe landing page and social media presence for all countries and use basic SEO techniques to ensure that it is the top listing for mainframe education queries. Next, IBM should allocate an actual budget to buy ads on facebook, twitter, and google targeted to IT and CS students to ensure greater exposure to the contest outside of the powerhouse Academic Initiative Schools. Next, IBM should try to build a social presence or partner with a Mainframe-themed social network **hint hint** to better build a sense of comradery and support among millennials thinking about mainframes. On this network, young mainframe employees and older mainframers interested in mentoring should engage with this talent to give them support and convince them that the mainframe is worth learning. While it’s critical to partner with IBM Academic Initiative schools, IBM should also set up training infrastructure for the numerous students that aren’t at Marist College. This should include free ongoing access to z/OS during the student’s entire academic career and MOOC content similar to that offered by Udacity. This could even by accomplished by partnering with Marist College more deeply to drive down the cost of z/OS IDCP courses via scholarships or other arrangements. Finally, the success of these programs should be directly assessed according to the number of new millennial mainframers entering the workforce. I suggest the heretical thought that perhaps the System z executive team should have their bonus plans tied to these metrics as well, as what’s the point of selling mainframe hardware or software if clients are going to be unable to run them in a few years due to a mainframe skill shortage? It’s not too late to fix this problem, but time is running out, and this is too critical to the future of the mainframe to ignore.

If this article resonated with you, please use one of the social-login buttons to join the Millennial Mainframer site. Our team is dedicated to taking a more active role in promoting the Master the Mainframe contest and helping to mentor and network those students into mainframe positions. You don’t have to be a millennial to join, as long as you’re willing to help millennial mainframers in their difficult journey. Let’s fix this together.

This post was originally hosted on the Millennial Mainframer blog

© 2018 by Sean McBride. All rights reserved.
Last build: 15.07.2021