Much presale information was treated as literature rather than content. Customers needed to search on any device for products that had the dimensions and styling they needed to fit their bathroom or kitchen, but many of the critical dimensions were buried and unsearchable. So Kohler Company created an XML database using web-based X-Forms for input, X-Queries for search, XSL for output, and Windchill roles for data access. Now, changes are made only once by the data owner. Content is accessible by user role. People can search for data without opening PDFs, and outputs can be tailored to the user’s task and preference.
How To Utilize Calculated Properties in your HubSpot Setup
Lavacon 2014: But will it fit in my bathroom? Creating a single interface for product presale data
1. But will it fit in My Bathroom?
Creating a single interface for product presale data
@MarkPeterson4 #LavaCon
Mark Peterson
Kohler Co.
2. @MarkPeterson4 #LavaCon
Why am I here
• We help users buy more of our stuff
• We built a portal to take down silos/barriers
• We let content owners do their own data
• We built structure to enforce consistency
• We increased accessibility to the data
• I’m sure it could be done better
3. The “napkin drawing” that got people to get it
@MarkPeterson4 #LavaCon
Single Source of Truth
for Product Presale Data
A E D
C
B D = Automation of the Process
E = Integration of Content
systems
4. What’s cool about our latest system
• Role-based access
• Lights-out updating with SAP manufacturing data
• Mass update tools
• Feeds to downstream systems
• Automated generation of formatted data
• Flexibility adding fields and new canned values
• Build-your-own and pre-formatted searching
@MarkPeterson4 #LavaCon
5. Your preferences for a “case study”
@MarkPeterson4 #LavaCon
1. Demo
2. Worst “mistakes”
3. Greatest scoring moments
4. Financial justification
5. Attacking silos
6. Tools and architecture
7. Searches and reporting
8. Establishing customer needs
9. Building what nobody could
articulate
10. Automated workflow
11. Automated output
12. Being a “radical” who always talks
about “structure”
I can tailor this to your top items
6. @MarkPeterson4 #LavaCon
Our DataMart
• The DataMart is a data superstore upstream of
Kohler.com and the “web builder”
Web-based input interface
Data owners and data lockdown for each field
Powerful searching tools for any user
Reports aligned to the needs of downstream users
•
7. @MarkPeterson4 #LavaCon
DataMart Project Goals
• Single source of truth for pre-sale product data
Internal access to all pre-sale data before general customer access at product launch
No more opening “documents” to answer a question
Change roll-out and change tracking
Reliability of data as a key trust factor
Incremental sales
Net cost reduction
Aligns with mobile needs
8. @MarkPeterson4 #LavaCon
Reasons
• When data changed, the right people and
systems were not notified
Data lacked consistency
There was too much time spent proofing
Current systems could not add fields fast enough and keep them
aligned
Searches could only be done by a small set of people
9. Average # information sources used to get information and/or advice regarding the specific plumbing
products you needed for this specific project:
For considering/exploring options:
@MarkPeterson4 #LavaCon
Consumers are now in control
•
4.2 3.9
2012 2013
For narrowing choices:
2.6 3.0
2012 2013
10. @MarkPeterson4 #LavaCon
How customers interact with us
• Customers spend more time planning details of their kitchen and
bathroom than any other rooms
• Matching the brand, color, and details matters to them
• And they plan rooms around a few key fixtures
25%
33%
2012 2013
68%
75%
2012 2013
68%
75%
2012 2013
Important that all bathroom plumbing
products are the same brand
Extremely/Very Important
Important that all bathroom plumbing
products have the same color/finish
Extremely/Very Important
Important that all kitchen plumbing
products have the same color/finish
Extremely/Very Important
13. Proof and reproof and never locked down
• Specification Sheets were proofed by four different areas
• People spent more time marking up and proofing changes than just
@MarkPeterson4 #LavaCon
making live changes themselves
• Other silos manually pasted spec info, which then needed more
proofing
• Catching stale data generated another proof cycle
• No method to notify all silos of specific data changes
14. Cost savings beyond incremental sales
@MarkPeterson4 #LavaCon
• Proofing time
• Time spent looking for data
• Pasting time
• PM time
Current
Proposed
Savings
2010 Total Cost All
Spec Sheets
Engr
Art/TP
Text/TP
Prf/TP Mrkt
Prf/Mrkt
Prf/Codes
Proj Man/TP
Prf/Engr
Comm PCI/DS
15. The business value and long-term potential
• Customers are less confused by our conflicting data
• We don’t turn on a fire hose when they ask for a drink
• Reduced internal support costs via:
Creating a central authority for technical answers
Reducing the number of times engineers get asked the same question
Eliminating multiple data entry points for any one piece of data
The request form is the data entry form…there is no retyping to proof
No stale data
Eliminating labor-intensive formatting of tables within our current spec sheets
Reduced language support costs and new-language
@MarkPeterson4 #LavaCon
16. @MarkPeterson4 #LavaCon
What changes for internal users
• Content owners type into a user interface…there
is no retyping to proof
Eliminates Paper/PDF mark-ups for spec sheets
Gatekeepers by content ownership area
Internal users can access data by queries rather than by opening
spec sheet PDFs
17. @MarkPeterson4 #LavaCon
Scale
• 20,000 products x 10 colors x 4 changes/year
Historically this was partially-captured in 4,000 spec sheets
Product changes several times right before/after launch due to ongoing enhancements
Multiple silos grabbed that data manually
Multi-month lead time for the simple changes…requestors just gave up
People were sure certain sources were the original/master, but they were not
There was more mythology than proof about how our people used our publications
Most legacy data was not field-based nor in a format suited to automated capture
18. @MarkPeterson4 #LavaCon
Timing
• One year from approval to launch
Live one year as of Oct 2014
One year’s worth of data pre-loaded (eat our own cooking)
Soft-launch 90 days before full launch
90-days to a functional prototype
Lengthy licensing approval
Financial approval
Pilot project
Content analysis
Needs analysis
19. @MarkPeterson4 #LavaCon
Before setting the costs
1. Customer data needs analysis and modeling
– What new data fields/content needed, what can go away, who uses what
– What is the source of the data, who owns it, who else links to or copies it
2. Data flow and storage
– What can accomplish the above needs and model
– How is new data received and vetted
– How to feed any necessary systems while retaining one master data source
– When is a data set considered ‘released for use’, either internal or external
3. Consumer formats and output tools
– Who generates internal data reports and what format serves their online needs
– Appearance of the main external “pub” format to meet short-term needs
– Accommodation of future language outputs
– Testing of capability for additional formats as they are defined
•
20. Data from current spec sheet page 1
Proposed Data Source
(uses stay same as current)
Current Data Source
Owner/originator: Marketing
How its received: Mark-up, E-mail
Where stored: Word, SAP, Contenta, CPDB
Who else uses it: Web Page, EQ Spec,
K-800 Design Manual,
Price Book, Resellers
Owner/originator: Marketing
How its received: Mark-up
Where stored: SAP, Contenta, CPDB
Who else uses it: Web Page, EQ Spec,
K-800 Design Manual,
Price Book, Resellers
Owner/originator: Codes & Standards
How its received: Mark-up
Where stored: Word/Excel file, Contenta
Who else uses it: (Partial) Web, EQ Spec,
K-800 Design Manual,
Price Book, Resellers
Owner/originator: Engineering, Marketing
How its received: Mark-up
Where stored: Contenta
Who else uses it: N/A
Owner/originator: Engineering (with Pubs
clean-up)
How its received: ASM file
Where stored: Windchill (ASM and
pubs file)
Who else uses it: Price Book
Owner/originator: Marketing
How its received: DataMart interface
Master/central storage: DataMart or SAP
Owner/originator: Engineering
How its received: DataMart interface
Master/central storage: DataMart or SAP
Owner/originator: Codes & Standards
How its received: DataMart interface
Master/central storage: DataMart
Owner/originator: Marketing
How its received: DataMart interface
Master/central storage: DataMart
*Replaced with web photo when avail.
Owner/originator: Communications
How its received: Data file from CPDB
(for image file number)
Master/central storage: DAM (Digital
Asset Management System)
21. @MarkPeterson4 #LavaCon
Basic process flow
Data Mart Store
Interface
Codes and Standards
Engineering
Marketing
Staging
Area
Not
Released
Released
Report of missing info
to each Interface
group
Add Override
Add info
or
override
Search
Field-based
Full-text
Exact matches
Greater-than, less-than
By role
Custom
Publish
1-Page pdf
Dynamic Spec Sheets
Code listing
RSS Feed – Mobile
Field-Based Data
Report
Changes over time
Difference list
Retail data file
Price book file
Etc.
22. @MarkPeterson4 #LavaCon
Worst mistakes
• Listening to “don’t worry about money now just be on time”
• Having the wrong people spec/approve design changes
• Not having enough test plans
• Underestimating how much people will work for a workaround
• Letting the data drift due to other issues
23. @MarkPeterson4 #LavaCon
Highlight reel moments
• From scratch to live in less than a year
• Saying yes to “can it do this?”
• Playing “stump the search guy”
• Management giving us money because of a demo
• 20+ launch presentations in one week
• Fields requested by users doubled scope (reduced silos)
• New internal customers we’d never heard of
24. @MarkPeterson4 #LavaCon
Thoughts on attacking silos
• Show the inconsistencies
• Find the true upstream source
(it can be like trying to find the fountain of youth)
• Connect, don’t compete
• Focus on getting the data out and their maintenance issues
• Get several silo owners together in a lock-in (several times)
• Focus on the outside users
• Show a vision and a really great demo, then let them talk
25. Thoughts on searches and reporting
• The search sells the rest
• Reducing cycle time for things like price books opened other doors
• One “free” byproduct saved somebody 20 hours a week
• All management groups were pre-motivated to reduce time their staff
@MarkPeterson4 #LavaCon
spend looking for data
• All management groups were pre-motivated to reduce time their staff
spend answering repeated questions
26. Thoughts on establishing customer needs
• Question the old stale anecdotes
• Being a customer
• Watching customers
• Monitoring social
• Popular search terms
• Competitive evaluation
• Parallel industries (such as kitchen appliances for us)
@MarkPeterson4 #LavaCon
27. @MarkPeterson4 #LavaCon
Contact
• Mark.Peterson@Kohler.com
• @MarkPeterson4
• 920-918-5361
• Vendor partner info also available upon request
• Experience Gracious Living online at: http://www.KOHLER.com
• Experience "Destination Kohler" for yourself online at
http://www.DestinationKohler.com.
28. @MarkPeterson4 #LavaCon
About the Speaker
• Mark is a quarter-century member of one company, having navigated through
typesetting to desktop publishing to early adoption of XML and content
management.
• Formerly a staff and department manager, he now focuses on identifying and
implementing technical solutions to meet trending consumer needs.
• Before settling in “technical publications,” he was in electronics sales, journalism,
training, and even a little building construction.
• Working with a major international plumbing manufacturer with thousands of
products, he sees lots of data to connect and silos to decommission.
• He also gets an excuse to spend time in Home Depot and to play with new
computer technology.