Manufacturer saves millions with Snowflake-powered analytics strategy

Strategy-first approach to analytics saves manufacturer millions of dollars and thousands of man-hours.

Article 3 in a three-part series

As we discussed in our previous articles, getting the most impact and value from Snowflake, or any other cloud-based data platforms and tools, hinges on having a modern data and analytics strategy in place. The strategy sets analytics priorities that are aligned to business goals—and enables business and IT leaders to make technology investment decisions that advance those goals.  It can take time and discipline to design the right strategy first—especially when powerful new platforms and tools, like Snowflake are so easy to stand up and quick to deliver an initial pay-off. But when an organization commits to a strategy-first approach, it pays operational and business dividends for the both the short and long term.

One of Aspirent’s clients, a US-based manufacturer of building products, had to make this very same decision: Leverage technology for a short-term fix to its analytics frustrations, or map out a phased strategy that could flex to meet IT and business goals over time.

An unsustainable situation

Towards the tail end of a multiyear, enterprise-wide SAP implementation, our manufacturing client faced several challenges to accomplishing its analytic objectives. Data couldn’t be easily integrated from within or outside SAP. The HANA views created for business unit reporting suffered from performance issues that made ad hoc analysis cumbersome and ineffective. Processes were riddled with redundancies and inefficiencies.

Simply put, our client had maxed out the analytic capabilities of its SAP system. The data was there for the most part, but largely inaccessible to the areas of the business— sales, finances, logistics, and manufacturing—that needed it to make strategic decisions.

Obstacles to business insights—from all directions

Amid the multiple data sources and custom hierarchies, IT teams struggled to enrich, extract, or transform data in any meaningful way. Key ERP and transportation management databases in SAP couldn’t easily talk to each other. And when teams tried to bring in third-party data to increase visibility in specific areas of the business (like shipping logistics) the process required even more time and manual effort. Severely limited reporting options and a lack of visualization capabilities meant “reports” were delivered via rudimentary data tables in Excel spreadsheets. Insights on problem areas or trends over time were difficult to spot, track, and address.

The obstacles and limitations involved in pulling and analyzing data made it impossible for the organization to have a single of source of truth. Reports of different shapes, sizes, and data points proliferated across IT and business functions. Case in point: Approximately eight different business teams were pulling reports on product shipments, all rife with inconsistencies and inaccuracies. And because even these sub-par reports were so painful to create, they were done on a weekly basis at best. The business was flying blind in between their Monday status meetings—and not flying anywhere close to 20/20 the rest of the time.

Gaining clarity on the best path forward

The company knew that getting on the right analytics path would require more than just technology, so they turned to Aspirent for help mapping out a modern data and analytics strategy. Yes, they wanted to make quick progress on the analytics front, but they also knew they needed a better and more scalable environment and data foundation for their future-state analytics ambitions.

Before they could move forward, they needed to know where they were starting from. That meant conducting a thorough and objective assessment of their existing data and analytics capabilities, technology limitations, and pain points. The process produced a detailed understanding of their current analytic process, the metrics used, decisions made—and it clearly spelled out the unmet needs. With this complete picture in hand, the Aspirent team got to work defining a target architecture and identifying the tools to drive a next-generation data model.        

Establishing the foundation: Snowflake’s value-add

Using Snowflake as a key enabling part of the strategy had been a possibility from early in the assessment, but the client wanted to understand how best to leverage its capabilities—and combine it with other technologies—for maximum impact. Its unique elastic database functionality would provide the accessibility and scalability the organization needed to accommodate different user hierarchies and workstream needs on demand (and manage costs in the process).

Snowflake’s multicloud support would also ensure the platform freedom to essentially plug and play the cloud and analytics technologies that would best support the client’s near-term goals—in this case Microsoft Azure, Power BI, and Databricks—and provide a robust foundation for maturing their analytics capabilities.

The icing on the cake was Snowflake’s data sharing and data marketplace capabilities. Being able to easily find, integrate, and access third-party data sets specific to their industry and their business would enable them to improve visibility and conduct more advanced analytics without the headaches of their existing process.

A multi-tech strategy comes to life

What does the Snowflake-powered multisource technology solution ultimately look like?

• We created an end-to-end scalable analytical solution based on Snowflake and utilizing cloud resources.
• We extracted 200+ tables from the client’s existing SAP system using Azure Data Factory and landing raw exports into Azure Data Lake Gen2
• We built a curated multidimensional layer using Databricks that loads into Snowflake; combined multiple input sources and applied business logic, custom hierarchies, and record-level attributes.
• We identified third-party data sets for integrating from Snowflake that would enrich existing data and decision-making

Future State: Target Architecture

The analytics engine takes shape

With a coherent strategy in place and the data easily accessible, the team could focus on putting accurate and actionable analytics in the hands of business users. We created 12 certified data models, 20 certified Power BI dashboards, and trained 150-plus users how to create and consume reports. And to keep things moving smoothly on the back end, we provided the client’s IT team with skills development on data pipelines, platform, and data warehousing, implemented release pipelines and code repositories, and helped onboard their new Snowflake and PowerBI hires.

New efficiencies fuel new investments

In the process of bringing this solution to life, we helped the client understand how to track and manage the use and associated costs of their new cloud-based data and analytics environment. We identified the savings they could achieve by reducing or eliminating the use of their less efficient platforms and extracting greater value (in the form of both hard dollars and saved time and resources) from Snowflake and PowerBI specifically. These savings and the shifting of the funds they enabled—along with an estimated savings of 6,000 man-hours—helped offset the cost of the client’s overall cloud data strategy investment.

Building on the benefits

Just a year ago, our client couldn’t imagine a reality where they had easy access to the data they needed—from SAP and third-party sources—to support their business, let alone accurate and insightful reporting at their fingertips. But thanks to a refreshed data and analytics strategy with Snowflake as the cornerstone of a multisource technology ecosystem, the client can ingest and analyze data faster, easier, and with greater accuracy and impact. Here are just a few examples of the impact:

• Freight lane optimization: By replacing the team’s time-consuming manual data collection and analysis process with an auto-feed of freight lane rates from the Snowflake Data Marketplace, the team was able to extract timely and verified data, engineer it, and compare it to existing freight spend. From this process, we could identify inefficiencies and excess costs in their shipment network, which equipped our client to negotiate lower freight contracts with shipping partners—saving the manufacturer an estimate $3M- $5M in annual savings. We’re now working with the logistics team to engineer a more efficient tendering process, which will optimize how they assign their 1,000-plus daily shipments across five carriers. The new process will replace the existing chronological approach with one that identifies and automates the best possible configuration of carriers with shipments to optimize costs.
• Improved reporting and visibility: Remember those eight different shipment reports? They’re now in one beautiful dashboard that’s just a click away for business users across departments and hierarchies. Business teams can pull indicator reports from Snowflake on a regular basis and adjust their pricing, sales, customer service, or manufacturing strategies to maximize opportunities.

These analytic product solutions are just the beginning. Snowflake will provide the scale and functionality to enable the client’s continued move up the analytics maturity curve—looking at causals in the data, understanding why things are changing, and ultimately building predictive and prescriptive analytics capabilities that will help the company identify issues and solutions before they become problems. In the wake of all this progress, what’s the biggest challenge still facing our client? How to meet the surging internal demand for analytics and reporting. Now that business teams have visibility into what’s happening in their departments, and the insights to make informed decisions, they can’t get enough.

Just a reminder as we close the loop on this series: You’ll get more business impact from Snowflake (which already has a lot to offer) when you evaluate and implement it as part of a modern data and analytics strategy. Speed to stand up is a good thing, but speed to value is even better. And in case you haven’t yet read the first two articles in this series, you’ll find them here: “What is a modern data and analytics strategy and how can Snowflake help you deliver on it?” And “How to use Snowflake to its maximum value in your data and analytics strategy.”