TLDR:

1. The supply chain has evolved significantly. However, despite the demand for quick responses, only a small percentage of organizations can execute decisions in real-time.
2. Organizations are inundated with vast amounts of unstructured and unusable data. This data overload often leads to decisions based on gut instincts and intuition due to the need for a single source of truth and shared context. 
3. The article introduces the concept of "progressive overload," borrowed from weight training, which involves gradually increasing the load on a muscle to stimulate growth. This principle can be applied to supply chain optimization through continuous improvement cycles, focusing on one aspect of operations at a time.
4. Organizations should prioritize issues and take swift corrective actions to make data actionable. 

The supply chain is no longer in its pre-historic era. We are rapidly moving through the age of digitalization. As the pandemic highlighted, today's global, interdependent, complex supply chains are riddled with disruption and challenge. The resulting need for quick decision-making is obvious. Yet, only some organizations can execute.
‍

According to Gartner, 95% of supply chains must quickly react to changing conditions, but only 7% can execute decisions in real-time.

Despite the massive shift to digitalization, organizations are drowning in unstructured and unusable data, so they've reverted to relying on gut sense and intuition to make decisions, lacking a single source of truth and a shared context to guide the organization.

In the era of digitalization, organizations are inundated with vast amounts of unstructured and unusable data. This data overload often leads to decisions based on gut instincts and intuition due to the lack of a single source of truth and shared context. To optimize supply chains, it's crucial to leverage data effectively.

At the 2023 Last-Mile Delivery Conference, I recently spoke about how businesses can stop the cycle and effectively harness and leverage data to drive more insightful decision-making and improve operational efficiency. 

In case you missed it, I'm sharing how to turn senseless data lakes into practical, applicable intelligence that helps guide better decision-making and drive growth. In the end, you'll be able to start cleaning up your organization's data mess in your day-to-day operations by:

• Creating a better data pipeline with a single source of truth and a standard set of definitions
• Setting objectives and key results (OKRs) and driving key performance indicators (KPIs) to guide the insights and focus on your data analysis
• Identifying the bottlenecks and driving more context across your organization

To help you emerge from your murky data waters with actionable insights, I suggest homework at every step so you can work out the kinks and see results quickly. 

Messy data and the paradox of choice

Just how much data are you talking about here? We generate about 328.77 million terabytes of data — EVERY SINGLE DAY!

But, as Barry Schwartz exposed in his book Paradox of Choice, we don’t always like having a choice. He says that while people find the idea of choice appealing, in reality, too much choice produces paralysis rather than liberation. There are too many options, so people find it challenging to choose.

Coupled with the inconsistent and untrustful information being fished out of murky data lakes, we’re forced back to how we operated before digitalization: gut feel, sense and intuition to lead our enterprises.

We must move past this comfort zone to keep up with the pace of change and disruption. In a way, our companies must build their brawn. How? By building data muscles.

Progressive overload: Building your data muscles in 5 steps

For world-class athletes, progressive overload is a weight training method that creates hypertrophy by gradually increasing the stress or load placed upon a muscle, thereby stimulating growth. In practice, it’s like adding weight to your bench press every week. 

We can apply this concept to the supply chain through continuous improvement cycles, where you pick one aspect of your operation and focus on it briefly before moving on to the next. 

Here’s how you can implement “progressive overload” and build your data muscles: 

1. Set a north star

To get where you want to go, you must first pinpoint precisely where you want to go. Your team needs a target to navigate that landfill of data. The more pointed that target, the better they can drive real improvement. 

Here’s how: I recommend identifying 1-3 OKRs for your team. Support those objectives with as few KPIs as possible. They should not conflict with each other. Otherwise, you risk creating conflicting data trends, leading to analysis paralysis.

The timeframe is up to you. A good rule of thumb is to set an OKR cycle on a timeframe where you are 80% confident in your organization’s performance forecast. Generally speaking, this is quarterly but could be monthly or even bi-yearly. However, annual cycles are highly discouraged to keep pace with business changes.

Homework: 

• Identify 1-3 OKRs per team
• Choose 3-5 non-conflicting KPIs
• Set a timeframe according to how far you can forecast with 80% certainty (usually quarterly)

2. Standardize data definitions and structures

Collaboration breaks down when communication breaks down. A communication breakdown is most often due to a need for more context and shared language. When employees aren’t given the full picture, they talk past each other instead of to each other. 

A singular, clearly defined vocabulary promotes collaboration. It reduces the probability of miscommunication or misunderstanding by removing the burden on employees to translate what they mean into terms their coworkers will understand. For example, ensure everyone on your team uses the same definitions to describe fulfillment vs. distribution or demurrage vs. detention. After all, you could be saying the same thing but in different ways – let’s avoid that altogether.

Homework: 

• Create shared context across your organization
• Promote collaboration with shared context and language
• Establish a singular vocabulary and definition set

3. Create a single source of truth

When it comes to data, it’s garbage in, garbage out. To avoid discrepancies and untrustful data, you need a single source of timely, error-free, trustful truth that all partners and actors can reference.

Rather than manual uploads dependent on a singular employee, data updates should be automated in real or near real-time with a systematic process to clean, standardize, and enrich the data with business logic and calculations, leaving humans to manage exceptions only.

Homework: 

• Provide real-time automated data sources
• Clean, standardize and enrich your data with business logic and calculations
• Manage by exceptions

4. Make data actionable

To minimize the impacts of disruptions to plan, you need to prioritize issues and take action as quickly as possible. Operators shouldn’t have to hunt and monitor for issues with dashboards and hard-to-read charts with hundreds of data points. Instead, proactive system alerts help identify and highlight critical issues that require operator attention for immediate corrective action and troubleshooting.

For example, a logistics analyst who must ensure shipments are on time manages hundreds of shipments, 98% of which are on track to be on time. Rather than manually verifying each shipment through many supplier portals to find the 2% that require intervention, automated tracking and prioritized exception notifications alert the operators where the problems are and which ones must be taken care of first.

Homework: 

• Improve your time-to-action
• Manage by exception with pre-set rules, thresholds and alerts
• Enable corrective actions to be taken within the same platform

5. Measure and prune

We all know that to improve, you must first measure. Analytic reports with estimated values, such as cost or time savings, empower employees with relevant, measurable data and valuable guidance for the next cycle of improvements.

Equally important is to avoid data clutter by removing unused reports and views. This ensures employees have quick and easy access to the reports they need, with the correct logic and filters, and avoids bogging them down with the ones they don’t.

And lastly, using usage analytics, you can identify irregularities and opportunities for improvement. Are there certain people who always override the system flags, which may indicate a fault in the automation logic or a chance to provide further training? Is one product consistently flagged more frequently than others, and if so, why and what can be done about it?

Homework:

• Assign a dollar value per report, such as cost or time savings
• Remove or consolidate unused report views
• Identify irregularities for improvement opportunities

Continue building data blocks for your future

Building data muscles requires a commitment to continuous improvement, mirroring the principles of progressive overload in weight training. Just as athletes gradually increase their strength, organizations must incrementally enhance their data capabilities. 

By adhering to these principles and completing the suggested homework assignments, your organizations can unlock the true potential of data, ushering in an era of data-driven excellence and resilience in the face of disruption. As we move further into the digital age, those who harness the power of data will lead the way in supply chain optimization and business success.

Where do you go from here? 

• Want to learn more about how supply chains use new data sources to gain actionable insights? Check out this article.
• Or, if you’re interested in how data can inform supply chain visibility, read this article.
• Just want to stay in touch? Subscribe to PorterLogic’s newsletter to stay up-to-date on innovative supply chain practices.