When making something, quality is always a consideration.
It is something that should be guaranteed and improved upon. However, in striving for it, we want to avoid putting too much strain on anyone or anything, or causing them to suffer losses.
We'll create a ruler called "QA" so that everyone can have a win-win situation.
Former Yoyogi production director becomes new checker in Thailand
Hello, I'm Tanaka, a checker and QA at Monosus Thailand (Monotai).
I used to work for Monosus in Japan for about four and a half years as a director and coder at Coding Factory (CF). Those of you who worked with me on projects may know me. I am grateful for all your help during that time.
I was approached by Monotai to work as a checker, and transferred to the company in February of this year. I've come all the way to Thailand.
By the way, what is a checker? Monosus's checking system
If you are not familiar with Monosus's inspection system, please see the link below.
Check system to ensure "delivery quality" | Coding Factory - Coding Agency
The role of the "checker" is to check whether the deliverables produced by Monosus meet the delivery standards, give instructions for corrections, and judge them as OK.
The checking process was originally created to reduce the burden on customers of checking products after delivery.
This system works by having directors, designers, coders, and others who take a step back from the people who create, and who only "view" the finished product, double-checking it from a third-party perspective to ensure that only those that are approved are delivered.
Although the checkers don't usually appear in the public eye, they are involved in every project as "separate eyes and minds from the production team" and are now involved in not only CF but most of the final products produced at Monosas.
reference:
Changes in Quality Control Department Verification
CODING FACTORY TOPICS Vol.128 June 2019 issue
http://coding-factory.com/cftopics/cf128.html
When I was a former member of the CF club and was involved in production, I naturally asked for checks every time.
Now, he is in a position where he is asked to check the work by Monotai members and other production members.
Although I have some knowledge of production and am flexible in some areas, I am still a newbie as a checker. Furthermore, I am the only checker at Monotai, so I began to move forward by trial and error, relying on the experience, wisdom, and help of the members of the quality control department at Yoyogi Monosus.
If this keeps up I won't be able to earn anything!?
The distorted reality of "allocated amount < estimated man-hours < actual man-hours"
The job of "checking" has always been something that has not been far from me. I thought I knew what it entailed, but when I actually started doing checking work as my own job...
It was harder than I thought.
What I felt after actually checking it out
①It takes longer than expected
First and foremost, this.
Checking takes time. Of course, it takes less time than production, but in the end it took much longer than we expected.
In addition to the time it takes to check, the number of projects that are difficult to implement is increasing these days. Various documents are deployed depending on the project, and time is also required to interpret them (preparation for checking). Also, since each staff member is responsible for more projects than the production staff, it is very difficult to understand the specifications of all the projects, so before each check, they check the in-house document "Check Sheet Site Map" that summarizes the check requirements.
In addition, when measuring the productivity of checking work, there is an "estimated labor hours" that indicates how long each task will take.
However, sometimes the estimated man-hours are not enough to carefully check that the specifications given by the director and coder are being followed and that nothing is overlooked.
In other words, the "estimated man-hours < actual man-hours" formula was amazing. From my experience, it was about "estimated man-hours <<<< actual man-hours".
② I feel very tired
When I check things like that, it ends up being really tiring.
Not only are my eyes tired, but my shoulders are also stiff. Since I've been concentrating for so long, my head is gradually starting to hurt.
I try to take frequent breaks to maintain my concentration, but this can be a problem with the amount of work I have to do in a day.
If it takes 1 hour to actually do one page from design to coding, then an 8-hour shift would be 8 pages, but that's without a break. If you take breaks as needed, you won't be able to check 8 pages unless you work overtime.
I once tried to see how far I could go without a break, but after three hours I was exhausted. When I talked to my seniors about it, they scolded me, saying, "You're doing too much. You can't guarantee quality at that rate."
The checker is the last (or very last) line of defense to ensure that the quality of the work is deliverable by a third party. If that is not done properly, there is no point in having a checker.
Rest is essential to maintain a good level of consciousness, so it is important to take proper breaks during work and avoid overtime.
When I think back to the time when I was at CF and was busy with schedules and tasks every day.
I thought it was great that all the checkers could take proper breaks and go home on time, and I wish they could return more checks in that time.
Ah, I want to chop my younger self off. I want to hit him with the corner of a dictionary.
Even so, this is harder than I thought...
Having experienced firsthand the various aspects of checking that I had never experienced during my production days, I was left scratching my head.
Because there was another problem.
Further problems arise.
The inseparable topic of "money"
Working in a company comes with a personal budget.
As a checker, I also have a budget and aim to achieve it every month.
Before starting a new job, I was briefed on the budget and calculated how much work I would need to do per day.
Of course, I understood that achieving this goal from the first month would be quite difficult and that some pushing would be necessary, but there was an even bigger "distortion" here.
In the case of Monosus, the profit allocation percentage for each section involved in the project is decided. I won't go into the details, but if we assume that there was a project worth 500,000 yen, the estimated labor hours were about 1.7 times (!?) the amount of labor that could be spent within the allocated amount.
This is merely a comparative calculation with the allocated amount, so it is different from an estimate.
In fact, even at the time of the "estimated labor hours" stage, it was nearly twice the labor hours equivalent to the allocated amount.
This estimated man-hours is only for the checking work. It does not include the "direction man-hours" for compiling specifications and confirming them.
The next thing to consider is whether the actual labor hours are nearly equal to the estimated labor hours.
Within Monosus, a ranking is assigned to each page that is checked.
A: New pages. Coding is done from design to production, and each page needs to be checked in detail. Wording and links are also checked based on the design file and WF, so it takes time and effort.
B: Light new pages. This applies to mass production migration projects using parts collections. Since it is assumed that the coding is consistent, it does not require as much effort as A.
C: Correction. Check pages A and B and recheck if any corrections are found.
Checks A and B are called the first round, and C is called the second round. The first round is a content check + specified browser check, and from the second round onwards the areas and browsers with problems are checked.
The "estimated labor hours" corresponding to each rank are approximately 0.85h (50 minutes) for A, and 0.25h (15 minutes) for B and C.
However, when I actually tried working on it, it took about 1 to 1.5 hours to complete page A.
Also, depending on the length of each page, whether it has JavaScript built in, etc., it can take more than 2 hours. Sometimes it took half a day to check just one page.
This is also true for B-rank pages, which are considered "light." Even C, which should require the least amount of work, can take longer than estimated depending on the number of revisions made in the first round.
In other words, in order to achieve the monthly budget with the current allocation percentage, the actual labor hours, which tend to inflate, must first be reduced across the board to the equivalent of the estimated labor hours, and then further compressed to nearly half that amount in order to make a profit.
Even I was shocked by the harsh reality that was derived from the distorted reality of "allocated amount < estimated labor hours < actual labor hours."
Standing in the way of productivity
The quest for "QA" that is not just a checker begins!
If the amount allocated for the actual work is insufficient, you should organize and explain the situation and work to have the appropriate amount for the checking work included in the estimate.
But before that, there are some things we need to do regarding the work at hand.
In fact, until last year, the checking work was an indirect department and had no budget or sales.
Starting this fiscal year, the barriers between direct and indirect departments have been eliminated, and the structure has been changed so that each department can generate a certain amount of sales.
Having worked in the direct sales department, CF, for a long time, one of the major themes I pursued in my work was how to maximize profits with the minimum amount of time and effort given a given budget.
Now that I'm working as a checker and my checking work has a budget and revenue, I need to improve my productivity in the same way.
Why does it take longer than the estimated man-hours? Why is it so taxing on the mind and body that you can't work without frequent rest?
Why has such a distorted relationship diagram come into existence?
As I tried to trace the results back to the causes, I came to an idea.
Of course, it is essential to streamline and automate checking methods.
But isn't the fundamental problem also in the current workflow?
"Quality Assurance = QA" is not just about checking deliverables, it is about guaranteeing, managing and improving the quality of the entire service
When I was told about the transfer, the job title I was given was "Quality Assurance" (QA).
At the time, I thought it was just a cool way of saying it, but when I suddenly looked up the term "QA," I was embarrassed to find that it was a job title that actually existed and I just didn't know about it.
Just like a Monosus checker, they check the results, see if they have affected other areas, and give instructions for any corrections that need to be made.
Furthermore, analyzing the results and improving workflows and manuals is actually a very important job for QA; doing so is the job of "quality assurance and management" that ultimately leads to improved quality of services provided by the entire organization.
Reference No. 48: The role of quality assurance in improving call center quality | CRM Solutions / Call Center | OKI Software
https://www.oki-osk.jp/product/crm/column/2017/c48.html
What is the difference between QA (quality assurance) and QC (quality control)? | landgather
https://landgather.com/qaqc
As a former production person, I was able to see what I could do as "another set of eyes and a different mind" and began my exploration of "QA" at Monosas.
Doing this more would make it faster. It would be more accurate. It would reduce the burden on the director and the coder while increasing the efficiency of the checker. It would make the collaboration between the two smoother.
Up until now, I have been on the "creation side" at Monosus. Now that I am on the "viewing side," the overlap between my two skills has given me many ideas for improving workflow that will lead to improved service quality at Monosus!
This is the slide I used when I gave a presentation at the "corder party" at Monotai the other day. Improving the productivity of checkers should lead to increased productivity for directors and coders.
What I am going to write from now on is my "path to becoming a QA expert" which I am pursuing freely and independently, without getting anyone's permission.
Next time, I would like to consider the various problems that hinder productivity improvement that I have found while carrying out my work as a checker, from a micro perspective of problems with checking methods and a macro perspective of problems with workflow.
TANAKA Natsumi