RFID Tags

RFID tags have been in use for some time by some major retail companies (e.g. Walmart) and the US DoD (resulting in new standards: EPC). The trend is to replace barcodes with RFID tags.

 

RFID stands for Radio Frequency IDentification.  The tags can be embedded like a strand of string in clothes, packaging, and within the plastic base of a customer loyalty card.  These tags can be read from a distance and do not need to be in the line of sight.  There are currently little to no regulations in North America regarding their use in this manner.

 

Several frequencies are in use.  The application and distance from the reader usually dictates the frequency which should be used.  RFID Systems currently operate in the Low Frequency (LF): 125 KHz or 134KHz band, High Frequency (HF): 13.56 MHz band, Ultrahigh Frequency (UHF): 868 MHz (Europe) or 915 MHz (North America) or 930 MHz (Asia) bands, and Microwave: 2.45 GHz or 5.8 GHz bands.

 

The practical range of Low Frequency RFID systems is limited to less than 1.5 feet.  They are most applicable when you need to read through water and metal.  Common applications are animal tracking, proximity pass cards (subway passes),

 

The practical range of High Frequency RFID systems is limited to less than 3 feet.  They can be difficult to read through water or metal.  Common applications are tracking library books, hospital patients, and airline baggage.  Several standards are in place including ISO 15693.

 

The practical range of Ultrahigh Frequency RFID systems is limited to less than 9.5 feet.  They can be hard to read through water or metal.  This is the frequency used for meeting the EPC (Electronic Product Code) standard.  Common applications include tracking inventory all through the supply chain and distribution.  It is also used for electronic highway toll collection and parking lot access.

 

The practical range of Microwave Frequency RFID systems is limited to less than 3 feet.  They are hard to read through water and metal.

 

RFID technology is now being considered in some prototype retail applications to determine who is or profile the person that entered your store as a customer/prospect without your customer service rep needing to ask them for their name, address, credit card, email, or phone number.  This has the potential for instant recognition of good and bad customers. 

 

RFID technology is enabling real-time Point Of Sale customized marketing and up-sale (or down-sale) approaches for the customer currently engaged with a sales rep.

 

Recent developments in the RFID reader multi-frequency and multi-protocol devices should open up the use of RFID clouds to identify customers and relationships at point of sale.

 

Imagine the solid relationships you can develop and use to support customer retention just because it appears that the salespeople in your store remember their name, as well as how important a customer they are, and how they like to conduct business (even though your salespeople may deal with hundreds of customers a day)!

 

Does this sound pretty good?  How does this work you ask?

 

RFID tagging technology has gotten so cheap and ubiquitous it is being put into many retail items to track inventory and prevent theft.  This includes items of clothing and store credit/loyalty cards.

 

The appropriate RFID Reader can be placed at egress points and strategically located as well as handheld.  The Reader can transmit the tag information it reads wirelessly to your company’s internal computer network.  The RFID tags (depending on the protocol and your store's internal product nomenclature) self identify to the reader the unique inventory id tags accompanying the customer.  This is called an RFID cloud.  Depending on the products associated with the detected RFID tags this cloud could be used as a unique identifier for the customer/prospect.  An RFID reader can read the tags from several feet away while they remain hidden from view.

 

A search through your company's database can reveal the items previously purchased from your store as well as who purchased them.  Due to reader errors, and the nature of clothing, it is best to use a fuzzy search and preponderance of tags to identify / profile a customer.  But, such a match can be highly likely based on fit, specific type of clothing, and that only one person would wear the same article.

 

The loyalty credit card (still in their purse of pocket) RFID tag (as well as the cloud of clothing tags) can also be used to quickly look up the customer's information and profile the purchase history of the customer.

 

Other items the customer may have in their possession purchased from other stores may also contain tags that can be read.  For clothing stores this can help you to identify sales opportunities based on what percentage of your competitor's products may be in use by the customer facing you.

 

Although the cloud of unique tags may indicate who the person is most likely to be, the entire set of tags could be assigned to/purchased by someone other than the wearer.  If a person other than the original purchaser is wearing them, there could be a relationship that you could silently attribute and use in your marketing efforts (should the purchase have been within the last year) given additional information such as gender, last name, address, age, et. al.

 

Isn't this an unethical use of this data?  Isn't this an invasion of privacy? 

 

Technically in North America, it is not yet.  Europe does have some restrictions, but North America is still the wild, wild, west.  The responsibility for safeguarding private information still rests on the expressed relationship between the business and the individual who has entered your property.  If you provided them the opportunity and they haven't opted-out, you can use the information to market to them. You may still customize your offerings based on what would most appeal to a given customer.

 

I recall a movie starring Tom Cruise that took place in the future.  Everywhere he walked storefronts knew him by reading his eyes and cameras watched his every move.  His character had his eyes removed and replaced with eyes from someone from Japan just to throw off the tracking devices of the authorities he was trying to avoid.  Talk about invasive!

 

If requested, RFID Tags can be easily disabled with a blow from a metal hammer.  The folks who prefer not to be recognized can take such precautions.  They may easily retain their privacy from RFID cloud detection without needing to pull their eyes out of their sockets. 

 

As a responsible and customer sensitive business owner, you may even offer to freely provide or sell your RFID reader services to identify and permit your customer to disable the tags.

 

But let’s also face facts, cameras and software to recognize faces already exist.  They are in use by government entities.  Freedom and privacy need those who enjoy them to maintain eternal vigilance in order to keep them.  This requires you be armed with the same technology that can be used to infringe on your freedom and privacy; hence, the second amendment to the US Constitution.  It is better to be armed than just forewarned.  Leaving this technology just in the hands of the government and your competition puts you and your customers at a disadvantage.  RFID tags are a reality and will continue to expand due to the reduction in theft they impose and the cost savings in labor along the supply chain. 

 

Let’s also remember there are many folks that will greatly appreciate the fact that your salespeople will respond to them in the manner they want them to.  Some folks like a full explanation.  Some just want to know features.  Some only look at price.  Some prefer to avoid specific salespeople.  Everybody is different.  They all have different reasons for coming to your store.  Your ability to recognize and respond appropriately will improve your sales.  It is a competitive market.

 

There are some data quality issues to be overcome with this form of data collection and marketing approach.

 

1. Multiple recognition protocols (standards) are supported by RFID technology.  Different frame layers are used to store unique identifiers.  Next generation RFID readers are simplifying this task.

 

2. Multiple product nomenclatures (SKU#) are supported by RFID technology.  Different formats for unique identifiers are used.  Standards are coming out to support universal tagging standards (EPC) but some data cleansing/pre-processing may be indicated based on internal corporate product identification standards.

 

3. Individual matching using SKU# and RFID tags requires a preponderance of fuzzy matched tag evidence including purchase history and personal attributes.  The acceptable matching number of known products will vary per business and the business’s market penetration.

 

4. Household or family relationship determination requires real-time fuzzy matching of names and addresses.  You should never rely on RFID alone.  Doing so in a vacuum can introduce many inconsistencies.

 

5. Relationship types change. Consider: Friend, Spouse, Mother, Father, Daughter, Son, Sister, Brother, Brother-In-Law, Sister-In-Law, Cousin, Girlfriend, Boyfriend, or Co-worker.

 

Overall this approach can get you another step closer to being as appreciated as the neighborhood grocer down the street that could recognize every customer by a name or a face or what cut of meat to put up front in the display case when he saw you walk into the store.

 

Current contact data quality tool suites such as that from Trillium Software employ fuzzy matching algorithms that are applicable for matching non-name and address data such as SKU# and unique RFID tags and determining in real-time a preponderance of evidence thus recognizing a customer by the articles in their possession.

Personal Naming Conventions - USA

In the previous posting, I discussed House-holding. I did mention there were some challenges to be overcome in the data as well as the data collection process.

For the purpose of this discussion assuming the data is reliably entered in the correct format (which is neither a likely nor safe assumption), one of the challenges remaining is using it to identify family members and roles.

As an example, consider personal names. The United States of America and Canada follow traditional western style naming conventions with some minor additions, e.g.

<Personal title> <Given name> <Christian name> <Surname>

When grouping US or Canadian data by family household, we are most interested in the Surname.

The Surname equates to the family name (derived from “sire”). It is usually from the father’s side of the family, but there are some exceptions. After the 1960’s, more liberties started to be taken with the format for the family surname. Some families now use hyphenated (and non-hyphenated) versions of the mother’s maiden name and father’s name. Yet other families take the mother’s maiden name, and some families have decided on completely new Surnames. Still other folks have removed the family name for their children altogether in preference of a fashionable name such as “Freedom”, “ Liberty ”, or “Tree”. (Reminds me of that tattoo you once thought was cool and now you are stuck with).

Associating people with the same family name is a way to group people in the same family household. Assuming we can successfully match those with the same Surname or valid permutation of the Surname at an address as a member of the same family, that leaves us with the members of the family that do not have the same last name, e.g. Mother-In-Law, Father-In-Law, Daughter-In-Law, or Son-In-Law, and adopted children.

A data collection technique that could be leveraged to learn the possible Surname of in-laws is using the Mother’s maiden name for account verification. However, the data collection process could be improved by requesting both the Wife and Mother’s maiden name as well as Husband’s first name.

Adopted children can be difficult to identify as being in the same household. However, collecting additional demographic info such as the date of birth on an account application can be used to identify minors at an address and thus a presumably subordinate relationship to the head-of-household at an address.

The Given name is the personal name that the individual is called in common conversation, such as Peter or Paul or Mary. Identifying the correct Given name does come with challenges as well.

In the US and Canada , nicknames are common. Nicknames can be derivations of a root name such as Pete for Peter or Jon for Jonathan. Nicknames can also be derived from pop-culture phrases such as: Wingman, Wings, Chickie, or associated with particular career choices such as: Cookie, GI, Ears, Gunner, and include paternal or maternal nicknames such as Daddy, Papa, Pops, Pa, Mama, Mimi, Mommy, Mom.

Complicating the match process is the use of the Christian name (sometimes called the middle name(s)). It is called the Christian name as the assignment of a middle name was done primarily to honor and identify the Patron Saint for a baptized person. Now-a-days some parents, based on their modern beliefs, assign “middle-names” to their children to honor a relative or close friend in addition to, or in lieu of, a patron saint.

There are people that use their middle name in common conversation and only use their first name in legal correspondence such as “Arthur William McCullock” (who goes by “Bill”). There are also folks with multiple middle names, such as “John Henry Rip Van Winkle Franklin Solomon Clark”. Some folks even go solely by last name in common parlance such as “ Clark ” above.

The correct treatment for Given names is most relevant for matching individuals but some of the non-root name based nicknames can be used to identify gender and based on the profession lend credence to a match that includes a business title. Identifying the Head-Of-Household by inference can greatly help to identify which is most likely the family name.

In the USA , the Surname can be followed by a Generational title (either “Sr” for Senior, “Jr” for Junior, or roman numerals such as: III, IV, V, VI; or numeric abbreviations such as 1st, 2nd, 3rd, 4th, 5th, 6th, etc.). This can sometimes be used to clarify parental relationships.

There are also additions to this format such as a separate line for Business Title (such as President, Vice President, Secretary, Director, Trustee, Customer Service Representative, Sales Manager, Account Manager, et. al.) that can follow the Personal Name as well as Professional Suffix (such as MD, PHD, ESQ, et. al.) that could follow Generation. e.g.

<Personal Name>: <Personal title> <Given name> <Christian name> <Surname> <Generation> <Professional suffix>
<Business Title>

The Personal Title can identify a marital relationship such as Mr & Mrs, Dr & Mrs, Mr & Dr, et. al. The Personal title can also help to qualify certain business titles. Although in the USA military officer titles stay with the person even after they leave the active duty service as commissioned officers are always held in reserve. Other titles such as Reverend, Bishop, Cardinal, Deacon, Prior, Pastor, Sister or Brother denote a religious order relationship. Still other titles such as Sir Knight, Lady, Gentleman, Master, King, Queen, Prince, Duke, Duchess, Baron, Baroness, Princess denote permanent affiliations. In many cases personal titles that indicate relationships can be dissolved but the intent of the title is permanence.

On the other hand, Business Titles are specific to a job or role. Almost nobody retains the title of President for a corporation upon retirement. The notable exception to that is for those whom the title of President was conferred in the course of elected high office. This is the case for President of the United States , Senator, et. al. Alternate personal descriptors can be attached as well in lieu of Mr or Mrs such as “The Honorable” (abbrev Hon), “The Excellent”, and “The Most”.

So far, I have only discussed the basic western style conventions used in the “new world” of “English-speaking” Northern America . There are over 250 countries in the world. There are many different cultural naming conventions.

Most companies do not do business in all countries, but many companies do business in a subset of countries. With the emergence of the European Union and other international treaties that have relaxed restrictions for foreign workers, many folks are now crossing country and cultural boundaries to find work. Sometimes they adopt the personal naming norms of the country they are in, and sometimes they do not. In any case, folks that cross national boundaries usually bring something of value to the new country and are in need of goods and services in the new country. These people are usually good prospects for new business.

In a future posting, I will continue on this subject. (This was just the tip)

Don’t forget to Stay tuned..!

Householding

In every business transaction, what you know about the person you are selling to or buying from can make all the difference in the world. We all have business or personal relationships with others and it is through these relationships that we make things happen. A business transaction is a social contract that in many cases is between more than just the individual buyer and seller. Sometimes the real decision makers are in the background.

A customer's long term value to your business is most predictable and reliable by what relationships they have, not just with you, but with others.

One of the goals of Contact Data Quality is to help identify these relationships. We call this "householding".

Technically speaking householding can go much further than just identifying persons within families or persons within businesses. Both of these do identify important relationships and both of these can require overcoming some significant challenges in your data and data collection process.

Many companies also household at the contact point level, such as address, phone, and/or email address. Companies with savvy marketing staff identify "survivors" at the contact type level as well.

Address level householding is used by many companies to identify customers based on their distance from their distribution sites so they can better provision their resources and delivery of goods and services. This is acutely the case for utility companies that need to lay cable or telephone lines.

However, householding at the address level is important for the distribution of any good or service that supports an entire destination address, e.g. septic pumping, landscaping, trash pickup, to electric, heating oil or gas. All of these companies need to do business with the decision maker for the entire property. At a minimum, householding at the address level can help you to minimize mail costs. Of course, the cost of not reaching the decision maker with your sure-fire proposal is the highest cost to pay of all.

Companies that household at the phone level want not only to minimize telephone costs but contact the decision maker. This is usually not the small child that answers the phone and says "bye bye". There are also several companies that provide services delivered over phone lines including telecomm related call services, DSL, television/entertainment.

Householding at the email address level is done because it is a contact point and because many folks share an email address. Although we realize that it does not cost more to have your own email address, many people still share an address for other reasons. It can be for business reasons such as a fall back email handler when someone is out of the office traveling or on vacation. Other times, it is because of a lack of knowledge or desire by the other party sharing the email address. In any case, if there is more than one person on an email address, the likelihood of actually reaching the intended party can differ. Your email may need to be tailored differently based on an understanding of the relationship between the parties.

Survivorship at the contact type level is done to identify the likely primary contact or decision maker for a specific good or service at a contact point. We term it flagging the "Head-Of-Household". In many cases, name prefixes and suffixes or titles and gender are used to identify the classical decision maker that should survive a household level match as the best individual contact.

Depending on the good or service and the culture of the buyer different rules may apply for determining the "Head-Of-Household". There can be many "Head-Of-Household"s.

For example, in the US , the Head-Of-Household has classically been identified as the father, however, in households with no father, it would be the mother. This can vary depending on what product or service you want to transact. When it comes to certain household purchases, mothers tend to take the lead as they more commonly assume the role of home-maker. When it comes to purchases that involve long term debt, fathers tend to take the lead as they more commonly assume the role of chief bread-winner. In these cases it is in their best interest to take the lead in the transaction as if it goes poorly they would be left to make do with a less than optimal solution.

Culture also has an impact, as in the Netherlands it is common for parents to not specialize in the care they provide to the household. Men will sometimes assume home-maker responsibilities and women do the bread-winning.

In many countries, the decision maker is an absolute figure. Once the decision is made, good or bad, it is never overturned because the cultural norm says to do so would embarrass the "boss" and it is worse to embarrass the "boss" than to fix a mistake. The USA and Canada are notorious exceptions to this rule. They are meritocracies. In the US and Canada , decisions can be changed if the situation merits such and a reasonable argument can be made. Most other countries appear to be slowly moving towards a hybrid of an autocracy and meritocracy.

Identifying and establishing a relationship with the decision maker is a critical show of respect that should be shown for business to proceed in many cultures outside the US and in some countries it is mandatory .

Address Types

The term "address" is so common between cultures that many people think it is a fairly simple concept. It is very common but it is not so simple. There are many address types. Among these I list at least 3 with very different characteristics: Mailing, Shipping, and Legal.

A mailing address is an address that the postal service will attempt to deliver to.

A shipping address is a location where a shipper will deliver goods.

A legal address is an address that defines a piece of land/property.

Not all addresses are mailable.

Not all addresses can be shipped to.

Not all addresses are legal definitions.

There are different authorities for each type of address.

"Mailing addresses" are essentially owned by the postal service. These must identify a fixed and current location along a route that a mail carrier will deliver mail to.

"Shipping addresses" (such as a loading doc) are an agreed upon description between the shipping company and its customer. For example, a shipping address may include directions like "The Corner of Castro and Duboce Streets" or "Building 2, Loading Dock A".

"Legal addresses" are owned by the taxing authority. They do not change until a transaction occurs that requires conversion of the address to current conditions. For example, if you buy a piece of property and the post office changes your street name, your deed remains the same until you sell the property and only if you also change the deed, which is a likely requirement when you record the deed with the taxing authority.

There are addresses that can fall into none of the above three types for various reasons including 1) The postal service and/or shipper refuse to deliver there for safety reasons, 2) The property may have an alternate entrance that differs from the postal, shipping, and legal address, 3) The address is obsolete for all three types, or 4) The address is planned.

And there are additional address forms, e.g. Internal Routing, Contingency, et. al.

Internal routing is when addresses include information meant for an internal mail or delivery handler, such as a company mailroom. It could include such elements as "Mail Stop", "PMB" (Private Mail Box), "Dept", "Section", "Bldg", etc.

Contingency addresses are those serviced addresses that provide for alternate delivery when the recipient is not available. These include compound addresses such as a multi-line address block with "PO Box" and a street address on the next line. The assumption is that if the street address is not deliverable, the PO Box address may be. The US Postal Service discourages this practice. The Canada Post format does permit a more intricate form of this. I believe they will also eventually come to discourage it due to the implicit costs.