4 Key Challenges to Managing Meaningful Spend Analytics

To remain healthy and competitive, large enterprises require management and oversight over corporate spend throughout the organization. However, large multinational corporations continue to struggle with managing spend analytics to gain a consistent and meaningful unified view of spend throughout different functional groups within their organization. Due to large amounts of data silos and operational complexities in implementing data analytics, true insights for unified spend are either always untrustworthy or impossible to maintain.

The root cause to this problem can often be traced back to these 4 challenges:

1. There is no low-effort path towards consistent integration of spend data sources over time
2. Taxonomies describing spend are inconsistent, and often not meaningful in aggregate across systems
3. Classification of spend transactions is highly manual with long lead times and delays
4. Insights derived from spend analytics are often not trustworthy or interpreted poorly

To tackle these challenges, the Tamr platform has been used by many of our customers to help bring more spend under management, and instate trust, consistency, and actionable insights into their enterprise-wide spend analytics.

1. Tamr provides a platform to consistently integrate data sources over time

Depending on the size of functional groups and business units within the organization, large enterprises may have varying ease of access and ability to integrate spend data across ERP and procurement systems. Spend data, data systems, and domain knowledge are often siloed within specific groups of teams. The Tamr unification platforms provides a very extendable system for enterprises to wrangle, reconcile, and onboard disparate data sources over time with little additional overhead due to Tamr’s best-in-breed machine learning approach to data unification.

As a result, many of our customers have been able to unify spend-related data across the organization to empower a true 360-degree view of spend for procurement teams to explore and manage.

Spend Analytics Powered by Tamr: Explore the Embedded Tableau Analytics above. The dashboard can also be accessed on Tableau Public here.

2. Tamr enables the application of consistent, meaningful taxonomies to describe spend

One of the key steps in commanding a unified view of spend across enterprise teams is to ensure the organization has a consistent “language” in describing their spend. Having a standardized taxonomy to describe spend may often be challenging for many reasons as described in a previous blog post here. Not only does Tamr help with providing a way to build a common language to describe spend, but the platform can support the ongoing implementation of a unified, mutually exclusive, collectively exhaustive (MECE) taxonomy such as UNSPSC.

Applying a unified, MECE taxonomy like UNSPSC across all systems make organization-wide analytics far more meaningful. Note how Tamr’s implementation of a unified category impacted how indirect spend were characterized in the siloed legacy systems.

Finding the ideal taxonomy to consistently categorize spend across all transactions is subjective and difficult. However, working towards a unified taxonomy is critical if an organization is to have an accurate grasp of its enterprise-wide spend.

Observe how fragmented the unified category was described in legacy systems previously. By implementing a more unified nomenclature for spend, organizations have a much more accurate view of managed spend.

3. Tamr’s machine learning approach provides the ability to instantly classify spend transactions and avoid human error and delays

Many global enterprises continue to classify spend transactions by implementing rigid, rules-based formulas or sending data sets out to third-party vendors for manual labeling. Often times, rules-based formulas tend to be inaccurate overtime as data systems encounter unexpected data variety and cleanliness. On the other hand, outsourcing data for manual labeling introduces huge delays, stale data, and long time-to-insight. In addition, both approaches expose risks of error and exploitation inherent in manual intervention and processing.

With Tamr, all spend can be programmatically classified under one taxonomy with Tamr’s human-guided machine learning. With an approach that reproduces human behavior using data-driven machine learning models, new spend transactions feeding into a configured Tamr platform can be automatically classified within seconds without human bias or errors. As data variety changes over time, Tamr’s machine learning models can be easily monitored and updated to fit the most up-to-date texture of spend data across the organization.

Legacy approaches to spend classification leave gaps for rules-based and human errors. In the above example, legacy rules may have classified Desk Paper Clips as “Desks”, where Tamr, using all the data context in the system, would have classified it under “Office supplies”.

4. Tamr enables trustworthy, actionable insights from spending patterns and behavior

Organizations need transparent, trustworthy data to accurately plan their procurement strategy and operations, but simple questions are hard to answer due to an overwhelming amount of disparate data silos. Tamr solves this problem by providing a transparent process to harmonize spend data from different data sources and classify spend under comprehensive, standardized taxonomies that align with the enterprise’s view of their data.

By doing so, Tamr empowers trusted, actionable insights to answer strategic procurement questions of what, where, how and from who spend is being sourced. Tamr’s unified data easily extends to custom analytics and insights to help procurement teams manage both efficiency and risk.

With a trusted source of unified data, enterprises can confidently build analytics to drive operational and strategic insights in procurement that may not have been possible before.

Conclusion

As we have discussed in-depth in a previous blog post on Supplier Analytics, implementing meaningful and actionable data analytics to be adopted across large enterprises is often more difficult than expected. Without building a streamlined approach to capture and unify trustworthy spend data across disparate data sources, analytics can have very low adoption or, when consumed, lead to inaccurate and potentially consequential insights. The Tamr platform has been used by many of our customers to tackle the fundamental data problems inhibiting the powerful analytic insights promised by their digital transformation initiatives.

To lean more about how customers like GE realized $80 million of cost savings within a year through Tamr, reach out to schedule a demo.

‍