Duplicate Payments: Remove the Noise of False Positives Karl - PowerPoint PPT Presentation

Duplicate Payments: Remove the Noise of False Positives Karl Andersson, Founder Phone: 508-480-8990 Fax: 781-634-0500 www.technology-insight.com

Webcast Logistics Collapse or expand the main menu Full screen or reduced screen Raise hand Send us your questions

Agenda Technology Insight 5 min Background Duplicate Payments 5 min What Can Go Wrong Assessing Your Risk False Positive 10 min Definition Impact DataShark A/P 10 min Architecture Technology & Process Demonstration 20 min

About Technology Insight Founded in 1999. ‘ Do-It- Yourself’ with DataShark A/P • secure, web-based P2P operational analysis solution developed specifically to empower A/P to drive: – Business process – Special projects – Duplicate payment identification & recovery – Risk management Recovery Audit Services • Standalone OR • Complementary to DataShark A/P • Based on each customer’s resource & needs

Duplicate Payments

What Can Go Wrong? Vendor Data Entry PO Issues Receiving Master Electronic Supplier ERP Part Master ERP Issues Feeds Issues Billing Procurement 2 Way Recurring Processes of Card Matching Payments Supplier

What’s Your Risk? invoice 90% automated 50% automated 10% processing 10% manual 50% manual automated 90% manual One One Two plus # of systems no pcard or t&e with pcard &/or t&e with pcard or t&e transaction low mid high (100,000 – 400,000) volume (< 100,000) (400,000+) -PO usage -Inconsistent PO usage -clean vendor master -dirty vendor master internal -strict data entry -Limited data entry control controls -ERP controls activated -controls not fully activated use of system -hard stop -soft stop controls Risk Lower Higher

Identifying Duplicates: The Battle Begins

False Positives What is a False Positive? • A transaction identified as a duplicate payments but is not • E.g. monthly & regular payments – Rent & lease payments – Utility – Contractual Why??? • A: Specificity of the algorithms • High specificity – Highly accurate – Software users will use the obvious few – While captures many, also misses many. • Low specificity – Cast net wider but capture a high volume of noise – Lacks refinement

Applying Algorithms # of pairs of candidate High duplicate transactions # of candidates Noise True duplicate payments Low 1 Algorithms 200

The Knock-On Effect High Resource Needs • The large volume of candidate transactions require significant resource to review. • Can review the same transaction multiple times. Limited Analysis • Reduce # of algorithms • Set high floor of transactions reviewed – E.g. $2000 or higher • Limit comparison timeframe – E.g. Compare to transactions within last 6 months

The End Result High Absolute # of duplicate payments # of candidates Cumulative total of duplicate payments recovered Audit software results Low 1 Algorithms 200

The Technology Insight DataShark A/P Process

Our Customers Maintain control of the audit process • Technology Insight performs the identification • Customers perform the research and recovery Maximize duplicate payment recoveries • Technology Insight is not limited in the number of algorithms it can use • Cast a wide net

Our Customers Minimize resource needs • Proprietary refinement technology & process removes noise of obvious false positives • Deliver 50% and higher true duplicate payment Receive comprehensive operational A/P reports & metrics • Over 200 standard operational reports • Each solution customized to meet the customer’s specific reporting needs on an ongoing basis

DataShark Architecture PeopleSoft Oracle Legacy JD Edwards SAP Lawson Any System P-Card T&E Technology Insight DataShark ERP Independent Database DataShark A/P - For A/P & Internal Audit - Over 200 operational reports & metrics - Updated daily, weekly, monthly, quarterly or annual

Duplicate Payment Identification Data Execute DataShark Normalization Technology • Data from disparate • Invoice to Invoice invoice systems: (153+ algorithms) • Invoice to P-Card / T&E SAP Oracle (150+ algorithms) PeopleSoft JD Edwards • Accounts for: Legacy cross system Any system cross vendor • Scrub & clean P-Card cross currency duplicate vendors data entry errors – all fields • Scrub & clean T&E and more… • Initial Refinement credits monthly / regular payments and more… • Assigns % likelihood of being a duplicate

Duplicate Payment Identification Analyst Review DataShark & Customization A/P Delivery • Manual process by • Delivery via secure website • TIC analyst • Duplicates prioritized by • Review 100% of output % likelihood & value • Systematically modifying • Customer manages algorithms to research & recovery with remove obvious ‘noise’ DataShark Tracking & Recovery tool • Detailed manual review • Customer results used using proprietary DataShark Dashboard to continuously refine algorithms • Provides comments based • Target 50% and higher on research results true duplicate payment • Ongoing refinement of hit rate each customer’s solution

P-Card Data • P-card and T&E data is handled in the same manner. • Vendor Name Clean-up Data from Procurement – Data is cleaned & modified to create a Card Company Parent vendor master file. – A time consuming & manual process Technology Insight classifies as “Best Buy”

The End Result High Absolute # of duplicate payments # of candidates TIC results Cumulative total of duplicate payments recovered Audit software results Low 1 Algorithms 200

Savings TIC’s Audit Savings DataShark Software A/P # of Algorithms 10 153+ Example Company: # of 80% Candidates 1,500 290 Reduction 100,000 invoices, average invoice # of True 490% size of $3,900 Duplicates 30 147 Increase # of False 90% Positives 1,470 143 Reduction $456,300 More to Savings $117,000 $573,300 Bottom Line

DataShark A/P Demonstration

Assess Your Environment Arrange a meeting with Technology Insight • Discuss needs & objectives Free Trial – No obligation to purchase • For all eligible companies • Three years of historical data – Multiple systems – P-card and T&E • Three weeks to process • Selection of Duplicate Payments to Recover – No obligation to TIC • Five Day Free Trial

Any Questions Karl Andersson CEO and Founder direct: 508.281.4697 mobile: 617.233.3525 email: kandersson@technology-insight.com www.technology-insight.com

Other Items Survey White paper Presentation