Berkeley*IEOR v. W.W. Grainger — Profitability-Calculation Patents Fail the Alice Test

Background

Berkeley*IEOR owns three patents claiming computer-based methods for calculating the profitability of individual business objects. Its infringement suit accused Teradata and several Teradata customers, including W.W. Grainger and DHL entities. The district court treated claim 1 of U.S. Patent No. 7,596,521 as representative and held the asserted claims eligible under 35 U.S.C. § 101, reasoning that an unconventional use of parallel computation improved computer performance. The court separately entered judgment for the defendants on other infringement grounds.

The Court’s Holding

The Federal Circuit reversed the eligibility ruling and affirmed judgment for all defendants on the alternative ground that every asserted claim was patent-ineligible. At Alice step one, the panel held that the claims were directed to the abstract idea of calculating profitability: they collected and processed information, performed mathematical calculations, and used a generic computer to do so. Faster computer execution did not turn that abstract process into patentable subject matter.

At Alice step two, the asserted improvement—parallel processing—could not supply an inventive concept because the claim language did not require calculations to be performed in parallel. The parties and the district court agreed that the claims also covered sequential calculations. An optional implementation described in the specification cannot rescue a broader claim that may be practiced without that implementation. Berkeley’s alternative argument that performing calculations inside the relational database was inventive failed for the same reason: its alleged benefit depended on the unclaimed parallel-processing feature.

Key Takeaways

• An improvement described in the specification cannot provide the inventive concept when the claims do not require it.
• Using a computer to perform profitability calculations faster remains an abstract information-processing idea.
• A patent owner cannot rely on a preferred embodiment to narrow broad claim language during the § 101 analysis.
• The court did not reach Berkeley’s challenges to the customer-suit stay, indirect-infringement judgment, or other dismissals because ineligibility disposed of every count.

Why It Matters

The nonprecedential decision reinforces a practical drafting and litigation lesson for software patents: a claimed technical improvement must appear in the claims themselves. Describing parallel processing or another performance enhancement in the specification is not enough when the asserted claims also cover conventional implementations lacking that feature.