5 Things Things to Know About Dirty Data | Ntara PIMpoint 2019
•Download as PPTX, PDF•
0 likes•57 views
5 Things Things to Know About Dirty Data Ntara PIMpoint Presentation 2019
Recommended
More Related Content
Similar to 5 Things Things to Know About Dirty Data | Ntara PIMpoint 2019
Similar to 5 Things Things to Know About Dirty Data | Ntara PIMpoint 2019 (20)
Recently uploaded
Recently uploaded (20)
5 Things Things to Know About Dirty Data | Ntara PIMpoint 2019
- 1. 5 thingsyouneed to know about Dirty Data
- 2. Briar DeHaven New AccountsDirector Ntara Misti White Ecommerce Manager Royal Brass & Hose Introductions Hello, niceto meet you. Aaron Jestrab TechnicalBusinessAnalyst Ntara
- 3. 1999 Full-Service Headquartered in JohnsonCity,TN year founded Asheville &Charlotte, NC Dedicated solely toinRiver’s PIM Solution digital agency
- 4. 1949 90k distributed partsyear founded In Production on inRiver iPMC Headquartered in Knoxville, TN Distribution Centers Across the South
- 7. ofbuyers saidthey would be unlikely torepeat apurchase withavendor whoprovided inaccurateproduct information. Source:ShotFarm 87%
- 9. Product Item
- 11. Part# Description Weight QtyBox 4C100HCS5 1/4-20X1HEX CAPSCREWGR5COARSE 0.0174 100 Product Item Market theProduct Sell the Item
- 12. ProductID Description HeadShape ThreadType Grade FA107-1001 CapScrew Hex Coarse 5 Part# Weight(lbs) Diameter(in) ThreadPitch Length(in) PkgQty 4C100HCS5 0.0174 1/4 20 1 100 Part# Description Weight QtyBox 4C100HCS5 1/4-20X1HEX CAPSCREWGR5COARSE 0.0174 100 Product Item
- 13. ProductID Description HeadShape ThreadType Grade FA107-1001 CapScrew Hex Coarse 5 Part# Weight(lbs) Diameter(in) ThreadPitch Length(in) PkgQty 4C100HCS5 0.0174 1/4 20 1 100 Part# Description Weight QtyBox 4C100HCS5 1/4-20X1HEX CAPSCREWGR5COARSE 0.0174 100 Product Item
- 14. Product/Item Segmentation Whythismatters: • Data placed where it belongs • Saves you massive amounts of time • Avoid data duplication • Easier enrichment • Enablesclearer & powerful product search for customers 1
- 16. Product/datateamswasteupto Thomas C. Redmon, Data Driven 50% oftheirtimedealingwithmundanedataqualityissues.
- 17. Product ID AG104-2001 AG104-2001 Series 2 4 Title Spring Lock RatchetClutch Spring Lock RatchetClutch Product
- 18. Product ID AG104-2001 AG104-2001 Series 2 4 Title Spring Lock RatchetClutch Spring Lock RatchetClutch Product 2 Series 4 Series • Same ProductDescriptions • DIFFERENT:SKU& Fitment • Used forsame Applications • Providethe same Benefits
- 19. Duplication Whythismatters: • Accurately load data • Avoid data loss • Avoid inaccuracy& confusion • Save time 2
- 22. ProductId ProductMaterial Finish AD101-8004 Zinc withYellow Chromate AD101-8005 Zinc withYellow Chromate AD101-8006 Zinc PlatedwithYellow Chromate AD101-8007 Zinc PlatedwithYellow Chromate
- 23. Consistency Whythismatters: • Reduce redundancy • Maintainclear data appropriate to usage • Establishstandards for future data loads • Established standards = accountability 3
- 26. Number Description Diameter Unit OfMeasure 571-9000 7" OVERRUNNINGCLUTCH9000LBS. 7 57112000 7" OVERRUNNINGCLUTCH12000LBS. 7 FP4725 FRICTIONPAD 2.5ID X4.75 OD 2.5 IN
- 27. Number Description Diameter Unit OfMeasure 571-9000 7" OVERRUNNINGCLUTCH9000LBS. 7 57112000 7" OVERRUNNINGCLUTCH12000LBS. 7 FP4725 FRICTIONPAD 2.5ID X4.75 OD 2.5 IN (in)
- 28. Completeness Whythismatters: • Improve data reliability • Reduce returns • Reduce customer service requests/calls • Complete datasets = increased customer trust • Data completeness sells products 4
- 31. ProductID Product Title AG103-1005 55Series DrivelineAssembly 80'CV Cat6 - North American
- 32. Short Description Long DescriptionProduct/ Item Images Benefits Features ProductID Product Title AG103-1005 55Series DrivelineAssembly 80'CV Cat6 - North American
- 33. Short Description Service and Repair Guides Long Description Benefits Features Product/ Item Images ProductID Product Title AG103-1005 55Series DrivelineAssembly 80'CV Cat6 - North American
- 34. Schematics ProductID Product Title AG103-1005 55Series DrivelineAssembly 80'CV Cat6 - North American Service andRepairGuides Short Description Long Description Benefits Features Product/ Item Images
- 35. SchematicsParts Service andRepairGuides Short Description Long Description Benefits Features Product/ Item Images ProductID Product Title AG103-1005 55Series DrivelineAssembly 80'CV Cat6 - North American
- 36. Schematics Service andRepairGuides Short Description Long Description Benefits Features Product/ Item Images ProductID Product Title AG103-1005 55Series DrivelineAssembly 80'CV Cat6 - North American Parts Application / Industry
- 37. Marketing Content Whythismatters: • Drives SEO • Marketing copysells products • Marketing copy aids newcustomer sales • Empowers customer to understand the full product story • Cross-sell/upsell/kits/product associations increase sales • Aids the job of the salesperson • Reduces customer service requests 5
- 38. Ntara’s Data Scoring Rubric
- 39. Data Scoring & Rubric Ntara’s approach to dirty data 1 2 3 4 5 Very Poor Poor Fair Good Great Data Issue Data Score Criteria Duplication Duplicate unique identifiers Consistency Various spellings, integers vs. decimals, use ofinconsistent delimiters (comma, semi-colon) & formatting issues Completeness Missing fields occurred post product/item segmentation effort Product &Item Segmentation Little tono segmentation tointernal systems orcategories which product belongs Marketing Content Minimal content tomarket products Data gathering, enrichment, or content creation required
- 40. Clientexample Royal Brass & HoseScore:Then &Now Started at a .5 Now at a 4.5+ 1 2 3 4 5 Very Poor Poor Fair Good Great
- 41. Key Takeaways
- 42. Key Takeaways • Remember: market the product, sell the item • Save your team time with clean data • Set standards for data & use them • Hold folks accountable • inRiver aids in data clean-up • Clean data enriches the power of inRiver
- 44. Q & A
- 45. Description 1/4-20X1 HEX CAP SCREW GR5 COARSE Data placement and standardization Product Description Spec: Bolt Grade Spec: Bolt Type Spec: Item Details
- 46. Description 1/4-20X1 HEX CAP SCREW GR5 COARSE Data placement and standardization Product Description Spec: Bolt Grade Spec: Bolt Type Spec: Item Details
- 47. Data placement and standardization 1/2-13X7 HEX CAP SCREW LARGE GR5 PLAIN 1-1/2X9 GR 8 PLAIN HEX CAP SCREWS 1-1/2-6X8 HEX CAP SCREW GR.8 M20-2.5X110 HEX HEAD CAP SCREW 10.9 COARSE Description 1/4-20X1 HEX CAP SCREW GR5 COARSE Product Description Spec: Bolt Grade Spec: Bolt Type Spec: Item Details MANUAL Data Cleanup / Enrichment ✘ ✔ ✘ ✘
- 48. Bad Product Data Issues Leads To: Managing multiple excel sheets Inability to group similar products to sell Forced use of custom code or scripts Slow, continuous data clean up Incorrect information across channels Inability to customize pricing by sales channel
- 49. ©2018 Ntara, Inc. All contents proprietary and confidential. 50
- 50. ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American
- 51. 55 series PTO drive shaft with a 1 3/8-6 spline quick disconnect tractor connection and minus end yoke, cross and bearing kit needed for implement connection Short Description Long DescriptionProduct / Item Images ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American Long Description, Benefits, Features The Weasler high performance PTO drive shafts are the foremost drive shaft solution in the agriculture and lawn & turf industries . Weasler PTO driveshafts are complete assemblies that from tractor to implement. They are designed for continuous heavy- duty use and meet the requirements of large farms and contractors. Weasler PTO driveshafts and CV wide angle PTO drive shafts are interchangeable with standard PTO driveshafts available in the market.
- 52. The Weasler high performance PTO drive shafts are the foremost drive shaft solution in the agriculture and lawn & turf industries . Weasler PTO driveshafts are complete assemblies that from tractor to implement. They are designed for continuous heavy-duty use and meet the requirements of large farms and contractors. Weasler PTO driveshafts and CV wide angle PTO drive shafts are interchangeable with standard PTO driveshafts available in the market. 55 series PTO drive shaft with a 1 3/8-6 spline quick disconnect tractor connection and minus end yoke, cross and bearing kit needed for implement connection Short Description Long DescriptionProduct / Item Images ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American DESIGN FEATURES • Easy Lock Guard System: Provides full coverage at maximum angles, Full 360° friction weld on guard bell, Black coloring that is durable against ultraviolet light and ozone, Cold weather impact rated to -35°C, Meets and exceeds all applicable safety standards, Quick and easy process for removal and installation. • Yokes: Interchangeable with all other Weasler yokes and standard yokes available in the market, Cast iron collars,Through-Bore keeps debris from collecting inside. • Cross & Bearing Kits: High torque capacity and longer life, Manufactured with high quality steel for increased strength, Standard kits can be upgraded to or interchanged with E-Kits. • Same component designs used for Weasler’s domestic and metric product lines. PERFORMANCE BENEFITS • Design adjustability (cut-to-length) capabilities. • Interchangeability to fit with most competitor models. • Tri-lobe, lemon, and star shaft profiles available. • Easy lock guard construction allows quick and easy assembly or removal with a simple tool such as a key, coin or screwdriver. • Available extended lubrication E-kits reduce downtime with lubrication intervals of 50-250 hours and the high temperature triple lip seal reatains grease bette • Customer dedicated engineering and sales support. Benefits Features
- 53. Short Description Service and Repair Guides ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American Long Description Benefits Features Product / Item Images
- 54. Schematics ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American Short Description Long Description Benefits Features Service and Repair Guides Product / Item Images
- 55. Schematics Parts ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American Short Description Long Description Benefits Features Service and Repair Guides Product / Item Images
- 56. Service and Repair Guides Parts Application / Industry ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American Short Description Product / Item Images Long Description Benefits Features Schematics
- 57. 55 series PTO drive shaft with a 1 3/8-6 spline quick disconnect tractor connection and minus end yoke, cross and bearing kit needed for implement connection Short Description Service and Repair Guides Schematics Product / Item Images ProductID Product Title AG103-1005 55 Series Driveline Assembly 80' CV Cat6 - North American The Weasler high performance PTO drive shafts are the foremost drive shaft solution in the agriculture and lawn & turf industries . Weasler PTO driveshafts are complete assemblies that from tractor to implement. They are designed for continuous heavy-duty use and meet the requirements of large farms and contractors. Weasler PTO driveshafts and CV wide angle PTO drive shafts are interchangeable with standard PTO driveshafts available in the market. DESIGN FEATURES • Easy Lock Guard System: Provides full coverage at maximum angles, Full 360° friction weld on guard bell, Black coloring that is durable against ultraviolet light and ozone, Cold weather impact rated to -35°C, Meets and exceeds all applicable safety standards, Quick and easy process for removal and installation. • Yokes: Interchangeable with all other Weasler yokes and standard yokes available in the market, Cast iron collars,Through-Bore keeps debris from collecting inside. • Cross & Bearing Kits: High torque capacity and longer life, Manufactured with high quality steel for increased strength, Standard kits can be upgraded to or interchanged with E-Kits. • Same component designs used for Weasler’s domestic and metric product lines. PERFORMANCE BENEFITS • Design adjustability (cut-to-length) capabilities. • Interchangeability to fit with most competitor models. • Tri-lobe, lemon, and star shaft profiles available. • Easy lock guard construction allows quick and easy assembly or removal with a simple tool such as a key, coin or screwdriver. • Available extended lubrication E-kits reduce downtime with lubrication intervals of 50-250 hours and the high temperature triple lip seal reatains grease bette • Customer dedicated engineering and sales support. Long Description Benefits Features Parts Application / Industry
Editor's Notes
- In today’s Big Data era, where massive scale and complex data reign, success is achieved by prioritizing Data Quality management. Address key takeaways in the beginning --- sets the stage for the talk & revisit at the end.
- Poll audience – C-Suite, Marketing, Technical Aaron – When I introduce myself, I can speak to: - deep data immersion with the client and the product lines. - Data Modeling - my passion about the types of products, familiarity with the product line.
- We transform institutions into digital businesses through the strategic application of digital technologies. 3 prongs to our business Consulting/strategy Technical integrations & development Creative/UX
- Specializing in hydraulic hoses, fittings, assemblies, tube bending and tube fabrication, Royal Brass and Hose distributes products and parts for mobile equipment, transportation, OEM, mining, agriculture markets, retail suppliers, industrial, mill supply and the forest and timber industry. From 2018-2019: then to now story | Drew MacDonald’s talk
- Bad data erodes trust.
- Issues with placement of data at the applicable level Simple Example: T-Shirt Product Level: Marketable Information Tee-Shirt, Crew Neck, Short Sleeve, 100% Cotton, Pre-shrunk, Machine Washable, Tag-less design for comfort, feels like your wearing a cloud. Item Level: Color, Size Blue, Large Red, Large Blue, Small Red, Small
- B/c of our tools (excel) and constraints of thinking, Only think about skus…not product/items. inRiver Whitepaper All Brands Need A Full Fledged eCommerce Strategy Whitepaper https://ecommerceandb2b.com/struggling-bad-product-data/
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- More defined Information Architecture for downstream channels like e-comm Ultimately saves customer time in enrichment Product/item segmentation allows for attribute/data inheritance onto highest level of product (parent or product) and child items inherit those data points Customer doesn’t have to individually edit attributes or data fields at every level – can just enrich/edit at the highest level Hunter Fan: ceiling fans – same fan 15 times in different colors; if you have to edit child items
- Overview: Lack of Unique IDs for every product & item or multiple copies of the same data source Simple Example: Identical sku number for a red shirt & blue shirt Identical part number for different part sizes Identical data fields for products & items within 1 system – no concept of product/item segmentation Unique IDs could be sku, Product ID, etc. Caused By Bad data in system Inconsistent data Multiple systems of record Queries & joins Product & Item data segmentation
- Two product records Right away we see duplicated product id information (CLICK) Caused By Bad data in system Inconsistent data Multiple systems of record Bad Queries & joins Inaccurate understanding of data What is causing this (CLICK) Series Work with customer (CLICK)
- Work with customer Have the same product description They are used for the same applications They both produce the same benefits BUT they have different SKUs, Series, and Fitment based on the equipment they are used on. In excel, this data is “correct”…but PIM reveals that there are issues with the Product ID Anecdote for RBH: its fine if you have a Misti or Francine who’s been w/ the company for 30 years who knows how to ship the right product every time, it could be fine. But this doesn’t scale
- Would never load data into PIM that doesn’t have unique ID b/c 2 products w/ the same unique ID would override each other. A lesser skilled agency might not realize that overrides happen. This can cause a lot of confusion & a lot of data loss. Every product must have a unique ID prior to data load for a successful data load.
- Inconsistent usage of data elements, variables, & data types within a given field If you work with suppliers, who we know won’t always play by the rules, you may have to create custom templates per supplier to make use of the supplier onboarding module within inRiver. Or you may have to write scripts to ferret out the consistency issues (and other dirty data problems) to quickly highlight the areas that need addressing. Simple Example: Gray vs. grey Red Shirt vs. Shirt Red Decimals vs. integers 3 in. vs. 3 inches
- Based on recent research by Experian plc, as well as by consultants James Price of Experience Matters and Martin Spratt of Clear Strategic IT Partners Pty. Ltd. Companies throw away 20% of their revenue dealing with data quality issues. This figure synthesizes estimates provided by Experian (worldwide, bad data cost companies 23% of revenue), Price of Experience Matters ($20,000/employee cost to bad data), and Spratt of Clear Strategic IT Partners (16% to 32% wasted effort dealing with data). The total cost to the U.S. economy: an estimated $3.1 trillion per year, according to IBM. MIT conducted a study: We had 75 executives identify the last 100 units of work their departments had done — essentially 100 data records — and then review that work’s quality. Only 3% of the collections fell within the “acceptable” range of error. Nearly 50% of newly created data records had critical errors. https://sloanreview.mit.edu/article/seizing-opportunity-in-data-quality/ Research from Experian Data Quality also found that bad data has a direct impact on the bottom line of 88% of all American companies. The averages losses from bad data was 12% of the company’s overall revenue. A Gartner Reporter also found that 27% of data in the World’s top companies is flawed. https://insidebigdata.com/2017/05/05/hidden-costs-bad-data/
- Consistency of data, or lack there of can occur in many aspects of day Data placement and arrangement Hard to automate dissection of the data Difficult to make associations between the data Within lists of data, two variables meaning the same thing but presented slightly different Poor UX
- You have to introduce governance for consistency rules that are universally applied to your data.
- Issues with fields not being fully populated across all product & item details Biggest issue for RBH and for many distributors. inRiver PIM will quickly reveal holes in your data. We often find short/long descriptions are lacking or nonexistent. Especially as we use PIM to populate downstream channels like e-commerce, Amazon, or retailer sites for big box stores. PIM highlights where data fields needs separation, which can easily lead to missing data. Simple Example: Chair May be missing length, cushion, or cushion fabric
- Data Completeness is Critical You end up with less data than you need to purchase the product.
- Incomplete data does not tell the whole story. You end up with less data than you need to purchase the product. NOTE: Some way to understand that this is the web view.
- Required fields (must be completed) to be able to save the data in the system Vs. Completeness rules (enrichment exercise) that helps to tell the full product story…critical to sell the products.
- Lacking content that supports marketing, categorization, & segmentation of products & items Includes descriptive elements like long & short description, category, product line , product features, cross sells, up sells, associated parts, kits, bundles, & configurable Additionally, this relates to images and supporting documentation, product and item associations and compatibility. Not marketing friendly Issues with marketing copy for short/long descriptions Attach PDF form or word - cannot import because its in another form
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily NOTE: Lets tie it all together.
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily This is a bit harder to define but for Break hose fittings which has the max number of attributes we have 482 items, that means 36150 data points in that dataset. Total we have 21,872 items and using the average of 52 attributes each that would be 1,137,344 data points for items. Total we have 2653 products, using the average of 31 attributes each that would be 82,243 data points for product.
- Duplication: duplicate sku #s, part #s, or product IDs create huge problem – can’t have blue shirt and red shirt with the same UPC code
- This is a bit harder to define but for Break hose fittings which has the max number of attributes we have 482 items, that means 36150 data points in that dataset. Total we have 21,872 items and using the average of 52 attributes each that would be 1,137,344 data points for items. Total we have 2653 products, using the average of 31 attributes each that would be 82,243 data points for product.
- Consistency of data, or lack there of can occur in many aspects of day Data placement and arrangement Hard to automate dissection of the data Difficult to make associations between the data Within lists of data, two variables meaning the same thing but presented slightly different Poor UX NOTES Add specs that show HOW MUCH TIME it took to clean up. Add picture of individual item.
- Consistency of data, or lack there of can occur in many aspects of day Data placement and arrangement Hard to automate dissection of the data Difficult to make associations between the data Within lists of data, two variables meaning the same thing but presented slightly different Poor UX
- Consistency of data, or lack there of can occur in many aspects of day Data placement and arrangement Hard to automate dissection of the data Difficult to make associations between the data Within lists of data, two variables meaning the same thing but presented slightly different Poor UX
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily NOTE: Lets tie it all together.
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily
- ERP Data is very limited Most of the usable content is contained within a single field “Description” What is the PRODUCT Data {CLICK} Easy to ID, not necessarily This is a bit harder to define but for Break hose fittings which has the max number of attributes we have 482 items, that means 36150 data points in that dataset. Total we have 21,872 items and using the average of 52 attributes each that would be 1,137,344 data points for items. Total we have 2653 products, using the average of 31 attributes each that would be 82,243 data points for product.