Why a Complex Data Model May Be the Simplest Solution

How to best structure product data and metadata is a challenge facing many companies today. Take, for example, “Memory Organization,” a complex attribute found in Memory ICs. Memory cells are organized into rows and columns and one chip can even have two different organizations of its cells. This makes it possible to have two values for the attribute.

Memory Organization: 256 K x 16 bit; 512 K x 8 bit

What is the optimal architecture for a complex attribute like the one above? The old adage of “Keep It Simple” persists and leads many to capture “Memory Organization” in one text field.

Solution A:

Memory Organization: 256 K x 16 bit; 512 K x 8 bit

Solution A may simplify the initial process of data modeling and capture, but over time it proves inflexible and virtually guarantees inconsistent and unnormalized data. Manufacturers detail specification information in different ways and Solution A does not provide a structure for how the information should be captured and stored. Additionally, because all of the elements of the attribute are captured in one field, any manipulation of the data has to take place manually.

An alternative model might look like the following:

Solution B:

Memory Organization No. of Units 1: 256 K

Memory Organization Unit Size 1: 16 bit

Memory Organization No. of Units 2: 512 K

Memory Organization Unit Size 2: 8 bit

(Concatenated) Memory Organization: 256 K x 16 bit; 512 K x 8 bit

Solution B treats the attribute as having two values, each composed of two elements: number of units and unit size. It is the superior solution for a number of critical reasons:

1. Consistency: Name, value and unit of measure are broken out for each element of each attribute value. Limiting what is found in each field ensures consistency. Where appropriate, restricted values can be defined for the value and unit of measure fields to further promote normalized data that can be effectively maintained over time.

2. Search: Each “Memory Organization” can appear separately in the dropdown list of a faceted search menu, reducing the number of unique values a customer must sort through. As search technologies advance and navigation attributes employ searchable textboxes, a customer will later be able to search on the element of the attribute that is of most interest. Perhaps the customer is concerned with finding a 32 bit organization and less worried about the number of units. Solution B enables more effective search for both internal and external users and can easily feed increasingly sophisticated systems.

3. Flexibility: Solution B is a flexible structure that can be manipulated in a number of additional ways. Conversion and other mathematical operations on values and units are possible. Elements can be added or removed as the data needs change. Perhaps the test conditions of a particular attribute value were captured—“120 V at 50 A”—and it is later determined that only the voltage is of interest. If the entire string was captured in a single text field there is no simple way of deleting the “Test Condition.” If, however, the model parses out each element of the attribute value into “Voltage” and “Test Condition,” the latter is easily deleted or altered across an entire data set.

4. Integration: A data model that fully defines of all of its constituent elements and their relation to one another is a model that can more easily integrate with other systems both internal and external to the organization.

Gina Bulatovic

Sr. Data Solutions Consultant

Why Should We Care About Taxonomy?

Taxonomy, or product classification, is the basis of online navigation, attribute normalization, and meaningful analytics and reporting. Taxonomy is the ultimate expression of a distributor's ability to keep their products organized and is often the most over-looked data element in the ecommerce experience.

Signs Your Taxonomy Might Not Be Optimal:

• Leveraging a third-party data structure that is too broad in some places and too narrow in others.
• Product attributes are not normalized at the terminal taxonomy node (the lowest level of classification).
• Print catalog index repurposed as web taxonomy.
• Taxonomy nodes are not intuitive or are redundant.

Best Practices:

• Dedicate resources to maintaining a custom taxonomy based on the unique product breadth and depth.
• Leverage taxonomy as a basis for detailed financial reporting and analytics. For example, a more granular taxonomy can allow an organization to report sales trends of wrenches by type (e.g. "combination wrenches," "box end wrenches," "adjustable wrenches," etc.), rather than being limited to reporting and analyzing "Wrenches." Increased sales of specific types of wrenches could indicate trends within specific customer segments.
• Optimize taxonomy for meaningful online navigation. In tandem with keyword search, taxonomy allows customers to quickly navigate to and isolate relevant products.
• Route frequent search terms directly to relevant nodes of taxonomy to ensure search result relevancy.
• Use terminal taxonomy nodes as the basis for defining and normalizing product attributes.

Eli Cooley

Solutions Architect

Confessions of a Data Geek

As a self-proclaimed datavangelist, I often find myself drawn into (or leading) conversations about the value of data. I work with many companies that begin with an understanding that product data is important, but are unable to articulate precisely why. These are some typical scenarios by which data classification, cleansing, and normalization can provide clearly demonstrable ROI to an organization.

For manufacturers, cleansed data can provide a complete picture of what they buy to support their operations. The data provided by suppliers is often incomplete or cryptic, making it impossible to analyze in any consistent way. The size of the investment can be measured against the relative size of the spend. I've worked with companies with lots of plants and facilities, spending hundreds of millions annually to sustain their organization. It's not atypical of these companies to achieve 5 to 15% savings in annual spend, just by better understanding of the data. This can translate into tens of millions of dollars of cost reduction opportunity.

The "low end" estimates are achieved with minimal effort and impact to business as usual. The high end estimates require more tools and technology, but also enable the sustained maintenance of cleansed data. In other words, if you fix the problem on the cheap, you'll probably need to fix it again soon.

For distributors, enriched product data drives sales. eCommerce is a way of life, and customers cannot buy what they cannot find. Your products need to be classified, cross-referenced, attributed, cross-listed, and have images and other support collateral. Enriched product data isn't optional anymore, it's an expectation from an increasingly demanding (and large) audience. In speaking with many VP's of eCommerce, I've noticed a resounding theme. There is often a resignation to do their best with the data that the rest of the organization provides. This paradign needs to change; eCommerce needs to drive data standards throughout the rest of the organization. Tying search logs into an effective master data management strategy can often increase online sales by 10 to 15%, depending on the technologies in place at the time.

Both areas of improvement require little to get started, and can provide incremental value to more than justify the cost of getting started.

Yes, these are very real conversations I have at work, on the train, and even at parties. I'll post more information on where, when, and how to party with datavangelists in my next blog entry...

Rob Stowell
Strategic Solutions

ByteManagers Webcast Produces Bevy of Q&A

Webcast panelists provide written answers to select attendee questions

Several attendees from our June 25th webcast submitted questions, but as is always the case, there wasn’t enough time to respond to all.

Here, Eli Cooley from ByteManagers and Greg Palmer from Reid Supply provide in-depth responses to a handful of unanswered questions.

Q: Are keywords and categorization related and how?

A: (Greg Palmer) There should be a direct correlation between your product categorization and keywords. One of the keys to web page indexing is relevance. If your categorization and keywords match, the search engines can connect the dots easier. If your category is "widgets" and your keyword is "widgets", when you get indexed by search engines, they can determine that you sell "widgets" as one of your main product categories.


Q: How do they normalize the data if the vendor won't provide the data?

A: (Eli Cooley) This is a great question. It’s hard enough to normalize data across hundreds of vendors, but you'll often find that a handful of vendors will not provide crucial data elements or will not provide data at all.You are left with two options before accepting a gap in information:
1) Develop an escalation process by which Senior Management gets involved to resolve content gaps. We are seeing this scenario more and more frequently. We have even seen distributors ending relationships with vendors who cannot provide necessary product content.
2) Shift from the mindset of "asking" for data to "getting" data. This can involve sourcing from the vendor website(s) or catalogs when necessary. Some distributors go so far as to take product measurements or images themselves.


Q: Greg, did you have to re-engineer your process for product data, that is getting what is needed from suppliers, from internal product managers etc. Today we struggle with this, and are often told the supplier cannot give us the data.

A: (Greg Palmer) Our product data collection process has been refined over the years. Once your data is clean and your attributes are normalized, create a spreadsheet of the attributes you want all suppliers to give you in each product category. Some will provide what you need, others may not. For the ones that do not, you may need to assign internal resources to translate what the supplier calls an attribute versus what you want to call it.


Q: I would like to get Eli's and Greg's opinion on ThomasNet

A: (Eli Cooley) I will assume that this question is referring to ThomasNet.com, and not the organization as a whole. The problem with assessing a site like ThomasNet is that, on one hand, it is arguably the industry-leading vertical search engine. I can't point to another resource that is demonstrably superior. On the other hand, many people I've spoken with are frustrated by ThomasNet's content and usability.


Q: How would you go about researching search-terms that customers use for each product?

A: (Greg Palmer) Your search logs and analytic software should be able to tell you what your customers are searching for. You will find out a lot of important information in your search logs.


Q: How important is design of a web site? What suggestions would you make to get your team on board for a web site redesign?

A: (Greg Palmer) Web site design and layout are critical elements to web site success. If your customers can't find what they are looking for quickly, they will leave and not come back. The fastest way to get your team on board is to ask your customers what they like and don't like about other sites. Make sure you get a good cross section of customers and design your site that makes sense. Some suggestions we received, we decided most customers may not want or need what was suggested. You need to also look at other best-in-class sites and go from there.


Q: I assume paid ads work. I am interested in particular how paid ads relate to industrial distributors.

A: (Eli Cooley) It really depends on the product portfolio of the industrial distributor. One thing is certain for any industrial distributor: some of the paid ads will work, and some won't. It's important to have a metrics-driven system that will maximize campaign effectiveness over time. For example, you might find that advertising to manufacturer part numbers is highly profitable for some brands and a waste of money for others. A system needs to be in place that will stop running the ads that are losing money and increase the ads that are working. Given the number of brands and product types most industrial distributors deal with, an analytical framework is critical. Even for a mid-sized distributor, the universe of potential keywords is often in the hundreds of thousands, with significant portions of opportunity lying in the "long tail."
Another important factor is the Distributor's brand awareness. If I'm seeing your name every time I search for certain types of products, your brand is in my head, whether I know it or not. This can be powerful for niche and regional distributors looking to expand their footprint. Of course, if your website is poor all bets are off.


Q: Greg, did you focus on stock material or both stock and non-stock for normalization?

A: (Greg Palmer) Both. If you are selling it on your web site, it doesn't matter to the customer where the products are stocked. They need to be able to easily find whatever you are selling, stocked or not.


Q: Is a UNSPSC code taxonomy considered good, fair, reliable or . . . ?

A: (Eli Cooley) UNSPSC is reasonably effective as a global classification scheme. However, it would likely look silly if you tried to use it on your website. It would be too broad in some places and too narrow in others... kind of like a bad suit. Effective ecommerce requires customization. I would even go so far as to say "style." Even if an organization is classifying all products to UNSPSC behind the scenes to facilitate other business processes, ecommerce leaders tend to have a customized customer-facing ecommerce taxonomy.


Joe Walsh
Manager, Customer Development

ByteManagers Webcast a Success!

Survey generates solid industry insight

On June 25th, ByteManagers sponsored a webcast with Industrial Distribution (www.inddist.com) and our customer Reid Supply (www.reidsupply.com). The one hour webcast focused on the importance of quality product content and site search to help drive online revenue. Greg Palmer, Director of Marketing at Reid Supply, reported how recent efforts to overhaul their website led to a 17% increase in online revenue. The webcast drew 192 attendees from a total pool of 410 registrants. The profile of attendees and registrants was a broad cross section including a heavy dose from the Industrial Distribution market with company sizes ranging from under $25Million to an excess of $1Billion.

At the midpoint of the webcast, we conducted a survey that asked attendees "What percentage of your sales is generated online?" As you can see, over 27% of companies surveyed are generating more than 5% online, and 15% are generating more than 10% of company sales online!



The webcast is archived and is available for viewing at:
http://www.inddist.com/webcast/ALL_WEBCASTS/2069-Building_a_Profitable_Online_Strategy_in_Distribution.php

We'll continue to post interesting data that came out of this webcast in future posts.

Joe Walsh
Manager, Customer Development

Building a Profitable Online Strategy in Distribution

Improve online search and customer satisfaction and boost your bottom line.

Your company may have a great Website, with valuable information, but can your customers always find the products they are searching for?

Is your product data structured in a way that is useful to your customers and does it enable a complete online "self-service" experience?

Join industry leaders for the one hour webcast sponsored by ByteManagers, Inc. as they discuss the following key concepts: the importance of quality product data, driving online revenue with better search capabilities, building a "customer-centric" online strategy, and steps to implementation and customer adoption.

Speakers include:

Jack Keough
Editor
Industrial Distribution

Greg Palmer
Director of Marketing
Reid Supply Co.

Eli Cooley
Strategic Solutions
ByteManagers, Inc.

Date: June 25, 2009 Time: 2pm ET

Sign up for this webcast today.

Introducing Product Data Insider

Welcome to the first installment of Product Data Insider. The purpose of this Blog will be to provide education, best practices, and to aid subscribers in producing and sustaining world class product content. In addition we welcome questions and comments about topics you'd like to see covered. It is our intent to post new Blogs at least monthly and more if we find subscribers are checking back more often.

Some housekeeping notes. You will regularly read a switch in terminology between the words "data" and "content". Although, the IT community, depending on who you talk to, may have distinct difference between these terms, Product Data Insider, uses them interchangeably. Other terms you may see exchanged are Taxonomy vs Categorization, Item vs Product, software vs technology, among others. When in doubt, please email and ask for a clarification and I will be more than happy to oblige.

We hope to develop a community that welcomes the sharing of ideas and thoughts as well as challenges that professionals face so as to enable education as well as practical suggestions and solutions to content challenges for the MDM, IT, Marketing, e-Commerce and Product Information Management communities.

Welcome to Product Data Insider,

Jeff Bertrand
Sales & Marketing