Now You Find the Bottleneck and Intro on Hamilton's Counter Positioning - Data and Decision #6
Interesting application of bottleneck's idea from ToC and knowing more about strategy that can defeat big companies in this 6th edition of Data and Decision.
Hello there! Welcome to the latest issue of Data and Decision!
We've got a few new subscribers around here (welcome everyone :D). So, I think it'd be better to re-share my reason for creating this newsletter.
I write this newsletter because I've grown tired of reading the majority of content in the data sphere. Most of it consists of things like "Company with the highest Data Scientist salaries" or just another listicles such as "Top 10 BI Tools You Need to Learn."
This newsletter is different. I've chosen to follow my curiosity and collecting valuable skills for my day-to-day role as a data analyst. And as you may know already, I share what I learn here.
Some of the things I wrote are pretty conventional, like my last post a few weeks ago about Amazon's hybrid approach on setting the right metrics. But it can also get so niche (and little bit weird), like my hypothesis that we can adopt ideas such as "Tactical Periodization" and "Decision Space" from football and modify it for analytics field.
However, the goal remains the same: exploring interesting concepts, see if these ideas can enhance our work, and test them in real world.
That's my story in a nutshell. Without further ado, let's dive into this week's content!
2 Insights
Now You Find the Bottleneck
As I mentioned a few weeks ago, one of the books on my current reading list is "The Goal" by Eliyahu Goldratt. Here are two important points that I've gathered so far while reading this book:
First, you need to really understand your system’s “true goal”. For instance, every business that compete in the market have to make money, no matter what industry they operate on. Eliyahu’s mention these three metrics that can translate “make money” into business action: net profit, ROI, and cash flow. In manufacturing, the “make money” can be translated also into three metrics: throughput, inventory, and operational expenses (to know more about the definition of these three, you can read my post about it here).
Second, every system must harbor a bottleneck, a component tighter than the rest. This bottleneck, defined by the component with the highest gap between things produced and demand, wields considerable influence over the system's output—so much so that any effort aimed at improving elements other than the bottleneck is essentially a wasted effort.
The first point is quite straightforward. I can’t add anything to it.
However, the full effect of the second insight is still something that I need to work on. Though I haven't fully articulate its implications, two obvious notions arise:
If you don’t know the location of your system's bottleneck, identifying it should be your primary task. In fact, this is the very first step taken by the book's main character upon realizing his own ignorance regarding his factory’s bottleneck. Analytics can be a powerful tool in this endeavor, presenting a rare opportunity for us to shine and aid the team. In the book, this part is also the very first time word “analytics department” being mentioned.
List all activities that you did to “improve the system” and flag whether they work on “bottleneck” and “non-bottleneck” to each of them. If the majority of them relate to “non-bottleneck”, you already know there’s something wrong.
I'll illustrate both of these points by sharing my own mistakes.
Ever since college, I've dreamt of owning a printer. This is like an obsession during my college year. The idea of reading interesting research papers, sending them to a Bluetooth printer, and hearing the satisfying 'brr!' sound felt like pure magic to me.
I've had the means to purchase one for a while now, and last week, I finally convinced myself to do so. I believed it was necessary to enhance my reading system in terms of both experience and quality. But nonetheless, I want to do “ritual” that The Goal’s describe: find your bottleneck and have you work on it.
Imagine my surprise when, after conducting an analysis exactly like what I prescribe above, I discovered my true bottleneck: dedicated reading time.
Yeah-yeah I know. It's a familiar story for many of us. The shortage of reading material isn't my real limitation. I already have a stack of book and articles waiting to be read from the past few weeks. Even though this is just straight common sense, I don’t know about this - not until I start the exercise.
And so, my bottleneck is my dedicated reading time, or mimicking what Jonah’s from the book said, the “most precious resource” of my system.
Should we attempt to eliminate this bottleneck? No. The book even argues that not all bottleneck should be eliminated. However, prioritizing efforts to make it more efficient and produce output with higher quality should be our focal point.
The book provides a blueprint for achieving this:
Eradicating idle time for bottleneck components (in the book, by ensuring continuous operation of the machine),
Implementing quality checks for materials entering the bottleneck to ensure that you don’t waste bottleneck resource for producing garbage,
Make sure that the bottleneck’s output has an actual demand for it. In the book, this basically means don’t produce things that you want to sell 9 months in the future - produce what you know that had market demand today.
Now, let's apply this recipe back into the context of my example above:
During my dedicated reading hour, I must prevent any distractions from intruding and disrupting it, even for a single minute (paralleling the “elimination of idle time” that the book mentioned). This, of course, easier said than done.
I should meticulously assess the books I intend to read. Perhaps I can check reviews on Goodreads first, see the discussion around it in Hacker News, or explore Patrick Collison’s bookshelf (side note: also the same place where I discovered another one my favorite book, '7 Powers'). In essence, this constitutes a form of quality checking.
It might come as a surprise, but when committed to such an endeavor, reading ceases to be purely for leisure. Every hour is committed to responding to the existing 'demand'. In my context, this translates to reading what interest me and I can use in my work or this newsletter. While I yearn to read into Cervantes' 'Don Quixote', I shouldn't allow it to fill my dedicated reading hour. Of course like what you might predict, I still failed quite a lot on this part.
"What you have learned is that the capacity of the plant is equal to the capacity of its bottlenecks. Whatever the bottlenecks produce in an hour is the equivalent of what the factory produces in an hour. So, an hour lost at a bottleneck is an hour lost for the entire system.’’ - The Goal
Woah this already a long post. But you might be wondering, can we apply this to our business, or at least to specific business lines you're involved in? Absolutely.
Have your business already identified your “true goal"? Most business teams has some forecasts or budgets to measure targets and costs. Utilize these to define your "demand."
Identify the bottleneck component(s). There might be more than one. As mentioned earlier, the analytics team can play a significant role in this process.
Catalog every activity you've undertaken in the past 30 days and map each to its corresponding component: bottleneck or non-bottleneck. What percentage of these activities is linked to the bottleneck?
Based on my understanding from the book, strive to enhance the bottleneck by eradicating idle time, introducing quality checks before material/input being processed by it, and using the bottleneck exclusively for actual demand rather than allocating it to future use that might not realized in the long run.
In the upcoming edition, I will try to finish my summary on The Goal and wrote the complete review of this book. Stay tuned!
Why You Should Read More about Counter Positioning
“7 Power” might just be the most underrated book I've come across in the business category. I asked one of my friends, who shares the same nerdy interests as I do, about it, and he simply shrugged his shoulders. Most Twitter influencers who ironically discuss competitive advantages extensively still seem oblivious to the existence of this book.
Contrastingly, Reed Hastings wrote the foreword for it. Notable figures like Daniel Ek, Patrick Collison, and Peter Thiel have also expressed their praise.
Truly, this book has the highest signal-per-hype ratio.
One strategy from Hamilton’s 7 Power that I found particularly intriguing is what he defined as “Counter Positioning.”
According to Hamilton, this Power can be wielded to “defeat an incumbent who appears unassailable by conventional wisdom metrics of competitive strength.” This is precisely what Netflix employed to dethrone Blockbuster's dominance. It also explains how index-fund pioneer Vanguard managed to navigate a crowded and adversarial market of active funds.
Perhaps the most compelling aspect of Counter Positioning is that this strategy doesn't hinge on the incumbent being foolishly incompetent to be effective. In reality, such conditions are rare, as most competitions between companies "typically take place over extended periods and with great thoughtfulness on all sides."
Furthermore, Hamilton states, "The only bet worthwhile for a challenger is one in which, even if the incumbent plays its best game, it can be taken off the board."
If you'd like to peek a bit of what this book offers, I recommend listening to the Invest Like Best podcast episode featuring Hamilton."
Personal Update - My Discussion with Mas Adiva Shabana Last Week
Every other week I have a chance to talk with mentors that I found from ADPList and Linkedin. Until now, all of the mentors (yes, 100%) that I discussed with both had high curiosity about their field and really competent at what they do.
One of them that I had a opportunity to discuss with is Mas Adiva Shabana.
Mas Adiva now works as a Senior Product Analyst in Remitly, an UK-based payment company. During our call, he kindly shared his tips for working with product and engineering team and the data quality issues. We also lightly jokes about what is the usual problem that data team faced when collaborating with others and what can we do about it.
Some of the talking points that we had:
The role of an analyst in the UK is already highly specialized, which tends to confirm my hypothesis that technical tools and capabilities like coding, SQL, and dbt will eventually yield diminishing returns. Analytics team will have to add value from their business acumen.
When it comes to collaborating with engineering and product teams:
Always strive to understand their decision-making processes and constraints. This understanding will guide you toward providing them with valuable insights.
Read books that other teams are also reading. This insight is resonates with me. I'm currently delving into "What CEOs Want You to Know" and "7 Power" to grasp the frameworks that business teams usually utilize. I need to find books for other team as well.
Find ways to alleviate the engineering team's workload. For instance, consider crafting your own data application using DBT. Seek monitoring tools that are easy to manage and don't strain the engineering resources. Can’t help but try to relate this to our previous discussion: most company’s bottleneck is engineering resource. Make sure they only do what has the most impact.
Regarding the skills analysts should cultivate during this tech winter:
Adapt to accomplishing more with fewer resources—acquire skills that enable you to handle multiple tasks effectively. Effective because in this winter, team size will shrink but the task will kept increasing due to the company’s urgency.
Familiarize yourself with all the significant output metrics (financial, product, user growth) that hold importance for your company. With this, you can gauge how’s the market going so far.
That’s all for this week issue, folks! See you around!