Supernova Award Category
Data to Decisions
The Problem
OEMs face increasing competition on cost and margin while market dynamics get more volatile, variable, and interrelated, complicating decision making and making savings harder to discover. Yet most sourcing organizations still use manual spreadsheets to analyze enterprise cost data, which is in turn entirely separate from fragmented sources of market intelligence. This manual approach fails to scale as procurement teams move from annual sourcing events to a continuous sourcing model, so much so that senior leaders now describe Big Data analytics (75%) and Machine Learning (44%) as important to their success.
The Solution
LevaData uses algorithms to cleanse and classify enterprise spend data, improving accuracy over time with the help of user training. The platform benchmarks this data against part-specific cost models and third-party sources to recommend savings opportunities and create negotiation playbooks for use with both contract manufacturers and raw material suppliers. This market intelligence derives from frequently refreshed data and pushes recommendations automatically to relevant users.
The results
One LevaData customer, a Global OEM in networking, achieved over $35MM in incremental savings on over $300MM of direct materials spend in its first year on the platform. A Consumer Electronics EOM achieved 98% visibility of total spend and multi-tier costs across suppliers and JDM/ODM partners within 9 months of adoption. Another customer increased scope of active management of parts and sourcing visibility by over 20% in parallel to a major transition to external manufacturing, to over 80% of total.
Metrics
All LevaData customers achieve verifiable, incremental savings within the first year of adoption totaling over $100MM in quantified savings in the last 2 years. In 2016, $50BB of total direct materials spend was annually managed on platform, and is increasing at 60% growth year over year. The LevaData Test Drive process allows for rapid validation of potential savings opportunities, based on historical spend and BOM information, usually within 1-2 months.
Insights and opportunities are constantly updated with every new update of forecasts, sourcing events, and leading indicators. The platform constantly analyzes 2MM+ data points across commodity groups, parts, and verticals for new insights.
Our mission is to deliver $20BB in incremental customer value by the year 2022 by leveraging data, community, technology, and talent.
The Technology
The LevaData Cognitive Sourcing Platform uses data streaming, AI and predictive modeling to analyze risks and opportunities for sustainable cost savings. It uses a combination of consolidated spend analytics across multiple bills or materials, commodity category trends, community benchmarking, and predictive insights based on leading indicators. It incorporates multiple enabling technologies including Google TensorFlow™, Python, Cloudera, MemSQL, Impala, and Kafka.
Disruptive Factor
Only LevaData offers a pay-for-performance engagement model that ensures a positive ROI of at least 5X in the first year of adoption. The platform is unique in that it combines enterprise spend and forecast information with market intelligence to identify emerging risks and opportunities in real-time, 24x7, to global commodity management and sourcing teams. LevaData’s insights and opportunities are constantly updated with every new update of forecasts, sourcing events, and leading indicators. LevaData continuously analyses over 2MM+ data points across commodity groups, parts, and verticals for new insights.
Shining Moment
Rajesh, Director of Supply Chain Operations at Cisco, was in midst of a challenging competitive market. The executive for a $1B product unit needed savings in the current year to protect gross margins. There were no IT resources available, so Rajesh delivered a data-driven negotiation strategy that combined all relevant internal data with external market intelligence. Rajesh delivered almost double the savings requested. That meant $60M added to the bottom line.