@nitobuendia

Nito Buendia

@nitobuendia

Solving Challenges with Tech
Skip to main content

Case study: Automating a critical workflow for an online business school

How a quick automation fix saved resources, drove improved quality and sped up new revenue avenues.

By on · (1488 words)8 min read
Image by rawpixel
Image by rawpixel

During near two years, I worked for @IEBSchool, as a product manager for @OpenIEBS (a course-based open e-learning platform, today called Citius School).

During my time there, I had the challenging task of increasing the portfolio of available courses with minimal investment; just by tapping into the already existing pool of resources of the company.

The challenge: bandwidth was a growth constrain

Very early, I learned that this was not as easy as it seemed. In general, the limitation of having to stick to idle capacity made me explore all the different processes and teams of the company looking for bottlenecks which prevented the new platform to get the resources needed to grow.

There was a large variety of teams involved in the process: from Customer Care to Finance, from Engineering to Marketing. All of them were at a large occupation. However, from all of them, our content producing team was the busiest and had the least available bandwidth.

This team is responsible for validating and creating the materials used for each lesson. This is a critical process for the business. The content is consumed by our students and it is of paramount importance for their learning experience.

The team receives the actual content from our various instructors, professors and guest speakers. Then, the content is made available on the platform and adapted to our quality and style guidelines.

As the portfolio of the company is quite varied; the team was tasked with preparing a large amount of lessons each week and had very little capacity to do anything else. This was a huge problem because the new platform required a large amount of courses to be successful.

While the school had the content to deliver new courses, the team was unable to prepare them at the pace required. In other words, the platform could not grow at the speed we wanted because there was a bottle neck in the material preparation process.

Exploring the workflow of content creation

I decided to immerse myself with the team and help them in the process of creating new contents. I soon realised how manual and tedious large part of the process was.

This is how the main workflow looked like:

  1. Professor or guest lecturers will send materials for each lesson to the team. The materials will come in any format / style the creator would consider.

  2. Our team would review the materials to ensure the quality was good. If it didn’t meet our standards, recommendations would be sent to the instructor to improve the content.

  3. Once ready, the team would add the content to the e-learning platform, following our style standards and tools.

  4. The team would then create a PDF version, going again through the same process of applying style and using a different set of tools.

The process could take about somewhere in the range of 2–4 hours per lesson; depending on the length of the materials. The validation task is the most important one, as it ensures the quality meets our students needs. However, the two most expensive tasks were the creation of the materials, on both the platform and PDF version. These were the least value added processes, and yet, the most costly tasks on the workflow. These two steps combined could take up to 80% of the time (1.5h–3h).

More importantly, this process was not only for our new platform, but for all the masters and programs that were done on the Business School. As a result, this workflow was one of the main costs to expand our portfolio.

Analysing the cost of the process

While the actual numbers cannot be shared, let’s run some quick numbers.

If the business school had a portfolio of 20 masters, with an average of 15 modules each and 4 lessons per module; this means that 1,200 lessons would need to be created per semester, or 2,400 lessons per year.

Assuming that this would take an average of 3 hours, this is 7,200 hours per year only for preparing the materials. Assuming an estimate of 1,800 hours per year per worker, this means that you would need 4 full time employees working only on preparing materials (7,200 hours / 1,800 hours per employee per year). While not huge, it is a significant investment for a relatively trivial, but critical, process.

This investment was alone for the master programs. The new platform was requiring hundreds of new additional courses, which could have driven the cost up significantly.

How could the School reduce this investment?

First approach: crowdsourcing and collaboration

The first investment we decided to take was investing in crowdsourcing the creation of materials. We opened our content creation tools so all instructors could edit their own materials on the platform directly and even collaborate with others.

This already had a significant impact on the process and worked very well in some courses in OpenIEBS. Our team could work over the first input submitted by the instructors, and simply adapt it to our requirements.

This reduced the time required for our team to get the materials prepared. Instead of adapting to any formats received and submit it to our e-learning site, the content was already in the platform ready to be reviewed and modified.

Despite this, a lot of tasks were still needed, like creating the PDFs or validating the content. Let’s not forget that the PDF creation still could take over an hour per lesson.

Additionally, this brought other issues.

The team had to invest resources on teaching the instructors how to use the platform; which was far from free. Even after this, some instructors were unhappy about the lengthy process and were unlikely to create more materials due to the high cost for them.

The loss of control on the review process made it difficult for the content to have the full quality control we wanted. Instructors could make changes at any time without our team noticing; which could lead to unreviewed material, bad style guides or discrepancies with the PDF copies. This was specially not acceptable for any of our programs.

We needed a different approach.

Second approach: automating redundant tasks

When reviewing the creation process, one could easily spot the problem: the content creation task was duplicated! The team had to add and style the materials on our platform; and then do exactly the same for the PDF version! This felt very redundant when I shadowed the team. Yet, no one had thought of this duplication before.

Once the content was input on the system, this was already available on a database. Why would we need to apply the same steps again just to create a PDF copy? Once the content was digitalised, there were multiple libraries that could convert the text and style into PDFs of the max quality with no additional effort.

Our engineering team took less than a week to incorporate the feature request to automate the PDF creation. Since this task was about half of the time, this allowed us to produce materials at double the speed than before. Not only this, but on the crowdsourced model, this ensured that the PDFs were always up to date and of the best quality as its creation was fully automated.

The impact of automating PDF creation

Time savings: about half of the original cost

Before, we talked about the cost of this process to create 2,400 lessons per year was 4 full time employees. If we managed to save 1.5h per lesson, this saved us 3,600 hours (2,400 lessons per year x 1.5 h) per year; or, 2 full time employees (3,600 hours / 1,800 hours per year per employee).

While this represents half of the team, this story is not just about time savings. Half of our content creation team was not fired overnight.

Instead, we doubled the investment in high quality value tasks like reviewing the materials and making proactive suggestions to our instructors, which really improved the overall quality of the materials.

Additionally, the bandwidth was no longer as constrained, and allowed us to create additional capacity to tackle the expansion of our new platform, which required a lot of their help to be successful. This way, the initial savings transformed into a new avenue of revenue for the company, and some successes down the line.

Conclusion

The business school was able to save important resources by automating a part of the workflow of material creation (one of our critical tasks).

The savings were not only significant in time and economic resources; but the additional capacity allowed us to increase the overall quality of the programs and explore new avenues of revenue quicker.

All by simply stopping, thinking critically about the work, realising about the small numbers adding up and solving the quick win.

Read more on Automation My Work Productivity