At last week’s Supply Chain Summit, Crisp’s Head of Supply Chain Barry Bradley gave brands a step-by-step guide to running a data-driven operations team.
It’s an incredibly exciting time to work in the supply chain. The pace of change is increasing exponentially, and there are innovations and learnings being shared across the industry every day. One of the big catalysts for change right now is the acceleration of data and analytics in supply chain management. But while data is critical to operating your business, it has to be used thoughtfully to fully capture value and help teams achieve their strategic goals. At last week’s Supply Chain Summit, I shared how brands can use data to drive sales growth, increase supply chain efficiency, and reduce waste.
The Power (and Pain) of Data
Why is data so critical in supply chain management? It all comes down to visibility and transparency. After all, we can’t fix a problem we don’t know we have, and we can’t respond to changes unless we know they’re there. A business that waits for customers to tell them an order didn’t arrive is ultimately going to have fewer customers than one that’s using data to proactively measure shipments.
Using data to understand where the problems lie helps your team prioritize investments and figure out what to solve first. A recent survey of procurement leaders showed a 15-20% increase in ROI when using data and analytics to make decisions. In the retail business where you’re scraping for every piece of margin to fuel growth, that’s a big advantage.
But getting that data hasn’t always been easy. For one, it’s time consuming. That same survey showed that up to 50% of procurement teams’ time was spent looking for data. This is a huge commitment, especially for smaller companies. Compounding that problem is defining what the ‘right’ data is. How do you know what data points are a true signal, and what’s just noise?
This all begs the question: how can you get the right data, and do it without a ton of pain? Here, I’ll outline how to build a strong data readiness process to maximize the potential for data to drive action in your organization.
The Four Steps to Data Readiness
Step 1: Set the strategy
This is the first step to any major project, data or otherwise. Start by identifying the goals of your supply chain team: what will separate your organization from the competition? How will you know if you’re hitting these goals? Establish a set of KPIs and set performance targets to measure your progress.
Let’s take the example of a supply chain team at a major retailer. They might say one of their core goals is to keep products in stock for customers. A good KPI to track for this goal might be a low out of stock (OOS) rate. But on the other hand, retailers also want to maintain a low average inventory to avoid excess costs. This is an inherent trade off, and therefore a decision to consider when you set your strategy. Being deliberate and transparent with these priorities and tradeoffs will keep your team from making suboptimal decisions, like large purchases of seasonal products that may prevent you from being out of stock, but also leave you with excess inventory.
Step 2: Define the process
Once you’ve set your strategy and KPIs, it’s time to use process mapping to define the steps that will help you achieve those goals. Work backwards from the desired outcome to see what steps happen leading up to it, documenting the data points available to measure progress along the way. When going through the steps of each process, ask “how do we know we’ve completed this step correctly and on time?” These ‘driver metrics’ can give you an early signal to act on before your outcome is impacted.
Let’s go back to our retailer example. Reduced OOS is our outcome, but we still need to build out the driver metrics to understand steps leading up to it. Here are some questions to ask:
- Before product gets to the shelf, it has to arrive at the store. Can we measure when the product arrives, and how much?
- What are the stock levels at the warehouse (before the product gets to the store)? Is the warehouse shipping goods on time?
- Before the warehouse, the supplier has to ship the product or raw material. If we can measure supplier fill rates and know if a supplier hasn’t shipped what you ordered, that’s a leading indicator of OOS. Can we order from a different supplier to meet forecasted demand? Should we remove promotions or special events that would increase demand before getting more shipments?
Step 3: Collect your Data
Once you’ve mapped your process and identified measurement opportunities, data collection can begin. Identifying information sources and how you’re going to use them will help you capture data correctly.
When collecting and aggregating data, it’s important to consider a few data qualities to make sure the measurements created are useful:
- Frequency: How often is the data pulled or updated? How often will it be reviewed? Make sure the frequency of review matches the frequency of updated data to maximize the impact of review meetings.
- Granularity: Are you collecting data at the right level? Lower granularity could mean more insights — but also more noise. A sales team may want hourly sales information to better target customers, but procurement will likely only look at weekly sales when purchasing products and materials.
- Normalization: Understand what needs to happen to make the data useful. Examples include transforming dates to match fiscal calendars, removing duplicate values, and inputting naming conventions.
- Source: Where is the data stored, and who owns it? Is this internal data saved in our cloud, or is this data from partners coming in via email, or pulled from a portal?
Data can be found in multiple places, and you’ll want to consider all of the possible options available to you in process mapping. Possible data sources include:
- Internal data (sales, manufacturing, procurement, etc.)
- Customer data (orders, forecasts, promotions)
- Supplier data (what’s on order, ETAs, lead times)
- Partner data (anything from transportation, to brokers, to media partners)
Don’t forget data hygiene
Once you have the data collection in place, the next step is to keep the data integrity high so it can be useful. That’s where data hygiene comes into play, which entails:
- Understanding who in the organization is responsible for data collection, accuracy, and calculations
- Regular auditing: Check in to ensure your data sources are still high quality, if processes have changed, or where metrics need updating
- Consider manual vs. automated data collection, cleaning, and auditing. Based on your organizational capabilities, it may make sense to use a data platform like Crisp to automate the process.
- Remove silos to ensure the whole organization (from sales, to supply chain, to finance) is working off the same inputs and toward the same goals.
Step 4: Build Your Action Plan
Now that you’ve built intentional KPIs and collected reliable data, you can build solid plans so the team can quickly take action when one step of your process is breaking down. Good action plans have a few things in common:
- A clear signal: What will trigger your various teams to act? With clear reporting on your new outcome and driver metrics, you can now map roles and teams to each KPI.
- A clear understanding of the process map to identify likely root causes. For any given metric, there are likely a few common issues affecting performance that you can identify ahead of time.
- Once the common root causes are known, script out the best actions to take when one of them pops up. The goal is to make your response simple, effective, and repeatable.
Setting up these action plans may sound like hard work, but your teams already have the expertise you need. It’s now a matter of aligning on what the process should be and the best actions to take. A quick meeting of the minds to write it down will save time each week when issues pop up or new people need to be onboarded.
With these steps in place, you’ll be well on your way to leading a data-ready supply chain organization that can easily meet goals and respond to disruptions.