How I (try to) approach process improvements and make “slightly better” decisions at my job

My job heavily involves setting up processes, monitoring them, and “behaving according to them.” This means that, even as an individual contributor to a rather fantastic team, I (luckily) have the freedom and platform to suggest and even implement process improvements that drive impactful infrastructure operations susceptible to change.

No matter what the “process” is, it can always be marginally made better.

How do we approach this problem?

Even before thinking of improvements and analyzing decisions and alternatives, we need to define “what” the “process” is. This can also be phrased as:

What problem are you trying to solve, and why?

Unless and until I can answer this question in first principles, I try not to think further. Believe me, this is waaaaaaay harder than it sounds. You really have to think and apply “cognitive strain” (more on this in the future) to answer this question deeply.

Now, considering you have a “rational” definition of the process and the reason behind implementing and following it, let’s look further and break it down even more.

So, a simple definition of a process could be: a series of actions or steps taken to achieve a particular end or result. It often involves a sequence of operations or activities that are systematically carried out according to a plan or a set of rules, mostly in a repetitive manner.

Now that we know which actions of our job can actually be called processes, and which cannot, let’s look at how we can “improve” them.

Again, what is an “improvement” to the process? Why is it required? What factors will we need to consider for it? And how will we decide if a process “has been improved”?

It is essential to answer all these questions deeply too.

For example, you might want to improve your customer service process to increase your CSAT (Customer Satisfaction) scores. Why is improvement required? (There might be an event or a metric that suggests this, maybe a low/average CSAT score in this case, or an escalation from a customer/client?) How is the CSAT score currently calculated? Why is the current CSAT score recorded in a certain way?

The question you might ask is, why do we need a metric for improving a process? To that, I flip the question and ask, if you don’t have a metric, how do you know that the process needs improvement? Does it make sense to spend a lot of valuable company time (that can potentially be spent on gathering data and evidence) improving something that does not even need “much” improvement?

The next question that needs answering is, “How will we decide if a process “has been improved”? This is where you frame the outcome. This could be: Improving (Increasing) our CSAT scores by a minimum of 10% this quarter while spending the same amount of resources. (Again, this goal is set ONLY AFTER looking at how CSAT scores are tracked currently.)

After you have set a proper outcome to be achieved, it is now time to lay out all the decisions of the CURRENT process followed (in first principles).

What I mean by that is, you write down each and every small action that takes place in the current process. Yes, they will (and should) be of hundreds of lines.

It should be down to clicks of a mouse. It is like looking at the process through a magnifying glass. This step requires a lot of patience, focus, and time. Look at each action as a decision; ask yourself “why” we need to do this and note down the answers.

After you have done this, you can suggest improvements in each of the actions by identifying MISTAKES in that specific decision. This also requires defining what mistakes in their purest form are, and how to avoid them, and then suggesting POTENTIAL IMPROVEMENTS.

An AI-generated example of Customer Order Processing for an E-commerce team, for you to understand what I mean: (Yes, I am lazy to type it myself.)

Current Process: Customer Order Processing

Open email client. Navigate to “Orders” folder. Click on the latest order email. Why: To retrieve customer order details. Read through the email to understand order specifics. Open CRM system in browser. Log in using credentials. Navigate to “Orders” module. Click on “New Order” button. Enter customer name in the designated field. Enter customer email address in the designated field. Enter the order date in the calendar picker. Select the product category from the drop-down menu. Select the specific product from the list. Enter the quantity ordered. Check availability of the product in inventory. . If the product is not available, send a backorder notification to the customer. If the product is available, proceed to the next step. Enter the shipping address in the designated fields. Select the shipping method from the options. Calculate the shipping cost using the built-in calculator. Enter the shipping cost in the order form. Review order details for accuracy. Click “Submit” to finalize the order. Receive a confirmation message. Close the CRM system. Open accounting software. Navigate to the “Invoices” module. Generate a new invoice for the order. Verify that invoice details match the order. Click “Send” to email the invoice to the customer. Close the accounting software. Archive the order email in the “Completed Orders” folder.

After you ask the correct questions, your analysis should kind of look like this:

Identified Mistakes and Potential Improvements:

1. Manual Email Handling:

Mistake: Opening and manually reading each order email

Why is it a mistake: It can be time-consuming and prone to errors (Are we tracking time/errors now? If yes, show data.)

Improvement: Implement an automated email parsing system that extracts order details directly into the CRM or order processing system. (What will be the cost to implement this? Is a SaaS solution available? If budget issues, is anything open-source available?)

2. Manual Data Entry in CRM:

Mistake: Manually entering customer and order details.

Why is it a mistake: It increases the risk of data entry errors.

Improvement: Integrate the CRM with the e-commerce platform or order management system to automatically populate customer information and order details.

… and so forth. You get the idea.

Note that these are POTENTIAL IMPROVEMENTS.

Not all potential improvements have to be implemented. It could even be worse to implement all potential improvements.

Again, this is way harder than it sounds. Some things just cannot be integrated or even automated due to a variety of reasons (time, money, etc.) and not every improvement has to be “robotic.” Even training your customer support agents to “sound friendlier” or “be more empathetic” can increase your CSAT scores! This is something you would need to take into consideration while laying out all the decisions in front of you while analyzing (as you will need to realize first that there is an improvement needed in customer empathy of the agents!)

The game is to pick the potential improvement that will have the most impact and has the highest odds to achieve the outcome you desire, and that is an actual skill which takes time to build.

What we seek is improvement to generate better results. As with all improvement, gains are achieved by reducing mistakes. In other words,

gains are reductive, not additive.

Processes are super customized and subjective. Hence, rather than searching for a ready-made template, I hope that the questions and approach mentioned in this article hopefully will make you a better decision-maker yourself. I am yet to become one!

Thank you.