So, Now What Do I Do?

Back to our client with "too much data".

Once the data is homogenized, cleaned and segmented, we can look to creative and copy. Since our goal is to make marketing communication as relevant as possible to our recipient (and I would add valuable), we need to begin to tailor our message to our recipient constituencies. In the case of our client, they serve both the residential furniture and the corporate/institutional markets.

Depending on the level of segmentation and the amount of data available, these segments should be further segmented. I would imagine that they serve high middle class individual to the wealthy. From a creative standpoint, images for each of these segments could be different. The same would hold true for male versus female and married versus single- each of these segments will find relevance in different images. Copy should be different as well.

With the institutional recipients, there are additional segments: corporate boardroom and executive as well as libraries and schools. Again, each segment has different needs and budgets, and therefore, creative and copy should reflect that. Further segmentation and tailoring of the messgae could be has around size of the company (Fortune 100 v. Fortune 1000, or public v. private, for example), as again needs and budget may be different.

Setting this kind of communication up, requires a greater investment than just just producing a catalog. However, if we follow the process out, the ROI is there. The need to produce a larger catalog, with many pages that don't apply to me as the recipient is in essence, eliminated. We don't have to buy printing enough for the year and then store the catalog, hoping that they don't become stale. When we fulfill a catalog request, we are mailing a smaller item, providing potentially huge postal savings. We can also fulfill on demand, and so are not tempted to batch fulfillment to save postage and processing fees. We are responding to our prospect at the time of highest interest, and remembering that timing is the largest lever of relevancy, with the content and time of greatest relevance, producing a much higher potential for a sale, and indeed a higher potential for a larger order and repeat orders. We can further increase relevance by personalizing the catalog; perhaps a letter on the inside of the from cover addressed to the recipient or past client endorsements from the recipients gender, marital status or income band. Let me reiterate that using the data should be very subtle. Many people have a visceral reaction to the perception that their privacy has been violated!

When you combine enough data and digital printing, the ability to provide extremely timely, highly relevant marketing communication is limited only by your imagination.

How Do I Handle all this Data?

I work for a small mail house in the U.S. that is rapidly evolving into an integrated marketing company. This involves building new processes and creating new technologies (at least the way we are going about it does).

Yesterday, one of the partners came in an told me that we have a client who is struggling with their data. They apparently have 16 or so different databases, with different structures and different purposes. The question was, how do I handle all this data? Asked a different way, the question is, "How do I compile this data in a way that allows me to generate relevancy for my prospects and clients?"

Obviously, this is a two pronged problem, first is the data itself. Second, is the messaging.

The first step we will take is to homogenize the data. This means that we will change its data structure to allow us to combine it. Simply put, we will name the fields the same, add an id field and a field for "list" and then combine the data. As I said in a recent post, this will be done in a spreadsheet because it is the most flexible tool for this kind of work. We will need to include every field in the source data, because this will ultimately be used in the segmentation process.

The next step is to cleanse the data. This involves making sure that the mailing address is in the mailing address field and that the address is current and up-to-date. It will likely involve NCOAing the list and should include formalizing the names.

Once the data is combined and clean- we need to look at the need to segment. I propose that all the fields in the source data can be consolidated down to maybe four or five. (I haven't seen this data yet, but I understand it is pretty rich.) Before we can segment the list, we need to have an understanding of what the marketer is attempting to do with the data and what segments they think they serve. We also should be aware of the need to augment their list with purchased data, so building data around list readily available data elements. This may require data enhancement services, that is sending the list out to have available information appended to our database.

Depending on how prolific the company is with their mailing activity, we may be able to do some modeling. Neural Network modeling attaches significance to independent variables and allows for very precise profiling of the actual customer. So, the database needs to collect as much information as possible, preferably from every prospect and customer touchpoint. This may require (and probably should be designed as) relational data structures. My recommendation is "When in doubt, keep the info." I like to refer to myself as a kook (keeper of odd knowledge) when it comes to managing a database. Keep it all- you can always set what you don't need aside.

As we go down this road with this company, I will add real world experience to the probablys and shoulds of this post.

I will cover the messaging issue in another post.

Apparently This is Already Happening

I was researching the "Anatomy of a Direct Mail Piece" yesterday and happened upon a post to Ad-Verse. The author is a brand marketer from the Seattle area with some very strong opinions- the ones I paid attention to concerned direct marketing. He seems to despise direct marketing! But after I finished licking my wounds, I realized he actually was not opposed to direct marketing as much as he is against mass direct marketing aka low relevance direct marketing aka spray and pray.

For those with the time and the intestinal fortitude (his blog should be rated R) you can find him in two locations; what is apparently his old site and what is now his new site.

Should I Use Relevancy Marketing - Redux

I just reread my post of a couple of days ago. I didn't answer the question very effectively; it is actually very simple.

If the widget you are selling provides a $10 profit and never has to be replaced, and it is all you sell; don't engage in Relevancy Marketing.

If you don't care about your customer, and feel like there is an unlimited supply of them;' don't engage in Relevancy Marketing.

If you expect immediate returns and have no patience for long-term strategies; Relevancy Marketing is not for you.

Otherwise, Relevancy Marketing makes sense. Do it.

Should I Use Relevancy Marketing?

Relevancy Marketing creates a pattern that allows any business to engage in it. It uses simple steps to create a powerful system of increasingly targeted communications to enhance customer retention and acquisition. So, every business should use it, right? Not necessarily!

To answer this question, let's look at retention and acquisition separately. First let's look at retention and begin by defining it. Customer retention is the act of keeping customers as long as possible because it generally costs significantly less to keep a customer than it does to find a new one. This is a mindset. It involves customer service. It permeates a company. It effects everything from how the phone is answered to how often the trash is emptied. It is about being "easy to do business with". But, it is more than that.

Larger companies calculate something called customer lifetime value to project the results of customer retention efforts. This is a fairly involved calculation. Play with it if you like. I propose the we simplify our approach at this time. (CLV is an important concept to understand, and I encourage you to see where you stand, as more and more data becomes available through Relevancy Marketing.)

Let's look at the value of a customer over time by determining total projected sales for a customer. Of course, I am counting on you business having relationship customers that stay with you over a period of time (relationship marketing). It also assumes that the value of that customer staying with you is high enough that the cost of customer retention activity provides a positive return on that investment. This calculation is a very simple one:

Total Projected Sales = Avg Customer Lifespan * (Avg Frequency of Sale * Avg Sale $)

Customer retention activities can impact each of the above variables. It can result in a longer lifespan, a high frequesncy of sale and a higher average sale value. It also allows you to examine the effect of marketing to current customers with the penetration/saturation mindframe; "Am I selling everything to every prospect the I can?" This is cross sell/upsell thinking. And if you aren't thinking this way about your customers, believe me, someone is.

Customer acquisition on the other hand, is adding customers. More importantly, the right customer. You may engage in some form of advertising now; print advertising, cable, radio or transit bill boards, to name a few. But, unless you have planned properly, you can't tell what the results are of these efforts, and you certainly can't be sure that your highest vale prospect is reacting to your promotional efforts. relevancy marketing will allow you to not only target your audience, but more importantly, to determine the effect of your effort as well as make them more effective as time goes on and the data is compiled.

There are two ways that Relevancy Marketing empowers customer acquisition: the first is using data available about your current customer to develop a profile that will be used to identify the most likely prospects. We'll talk more about this later. The second is by making the communications with these prospects more effective.

Determining the value of customer acquisition activities involves answering the following questions:

• What does it cost me to gain a new customer, presently?
• What is the value of the average new customer over time (Total Projected Sales)?
• Who is my best customer?
• How much more is my best customer worth to me than my average customer?
• How can I get more of my best customer?
• What does it cost to get more of my best customer?

If the value of your best customer is in accordance with conventional wisdom, they are on the order of four times more valuable to your company. (This is just applying the 80/20 rule.) More than likely it pays huge dividends to go get more of them.

Relevancy Marketing will allow you do do just that.

The Relevancy Continuum

The purpose of Relevancy Marketing is to develop a system whereby an organization evolves from a one-message-fits-all (or no message, that is, a state of no marketing) to an ability to communicate one-to-one.

The relevancy continuum is simply the progression for one-message-fits all to individualized communications. At one end of the continuum is static, black and white messages. While there is a certain utility for this kind of communication, its functionality is limited.

The largest single driver of response for direct marketing communication is relevance. Relevance includes the ability to address life cycle events (births, new home, new job, marriages) and seasonal or other trigger events (purchase of a new car for instance). It also includes the ability to deliver specific information that communicates to the recipients unique needs, desires, preferences, attitudes, attributes and perceptions. Additionally, personalized copy and creative increases relevance and therefore, response.

The objective is Relevancy Marketing is to move from a no or little information to enough information to create relevant marketing communications. My contention is the goal of the initial communications is to generate the data that will eventually drive the relevancy. Only after the data has been collected should the investment be made into the high relevancy communications. I don't mean that the data has to be all collected before relevant communications can start, on the contrary, if you have only enough information to communicate relevantly with one recipient, I say do it. Obviously, this will require an on-demand distribution model for these communications. Exactly!

So, begin at one end of the spectrum: low relevancy, high volume, low value and move to the high relevancy, low volume, high value communications. At the lowest end is the non-personalized black and white postcard. This is a great tool for list cleansing, announcements, or for when an offer is strong enough and compelling enough to break-through.

This is followed by non-personalized color, personalized black and white and personalized color communication. Color is useful for communicating information that requires learning- especially quickly. For example, a "we have moved" type of communication. Color increases the speed of learning, attention span, comprehension and recall, as well as improving the possibility of being picked up by a reader. Color also persuades and motivates better than black and white.

Basic personalization, today understood as more that just the name and address, begins to create relevance. With nothing more than an address, it is possible to create interest buy using the city or metro area in copy or headlines. With a bit more sophistication, images can be changed to reflect the city of the recipient. This is nothing special, but it begins create aesthetic changes that will create interest and increase response.

Thus far, our data is simple name and address data, and we are communicating on a one to many basis. Next up in the continuum is one to few communication. Additional data is required as we move into this realm: additional geographical data, census tract data, income, home ownership, children, gender, even lifestyle data are some of the elements of this data. Obviously, we should have a plan for the data we collect. (I'll point out here that one should be careful of the big brother appeal- don't be super obvious and direct with what you know; you'll have respondents questioning your credibility and honesty right off the bat!)

Communicating one to a few is called versioning. Versioning of communication is, as the name suggests, creating different versions of the communication based on affinity groups of recipients. Perhaps, one for women and one for men or one for singles one for married people to name just two. This takes the personalization from just aesthetic relevance to now contextual relevance. More response drivers are being met, and response rates will climb. This requires a higher level of sophistication in terms of programming and more creative and copy and will therefore require a greater commitment to investment and staying the course.

The highest level of Relevancy Marketing might contain additional components: data elements that are specific to the individual and varied enough to create conceptual relevance- that is, communication that includes highly individualized information about the recipient. An example of this is the college that creates course catalogs based on information requested by the recipient. Another component that might be used to create relevance for the recipient is called trigger data. Trigger data includes birthdays, moving, anniversaries, etc. Trigger data creates timing relevancy. Timing relevance is currently the largest relevance lever for high response. This might be used to bring a couple back to your restaurant: "In honor of you anniversary, we have made a reservation for you...", or "in honor of your wife's birthday...".

With a plan, low relevancy communications can generate data to allow higher relevance, generating more data, creating even higher relevance. Read relevance here as response rate, conversion, sales, revenues and profits.

But to make it work, you have to get started: mail something, anything!

Structure of a Mailing List

In my job I deal with mailer created lists and databases on a daily basis. Some are better that others, but almost all could be structured better. Since data is one of the cornerstones of Relevancy Marketing, I thought I'd take a minute to discuss how it should be structured. I am only addressing the mailing list portion of the data here. Any additional data can be structured as required.

An address consists of several elements:

• Addressee or Contact
• Firm (Company or Organization)
• Address
• Suite, Floor or Apartment
• City
• State
• Zip or Postal Code
• Country

There is only one address line. So your data should contain only one address line. You may in fact have two or three on your data, but only one is the mailing address. The problem is that in your data, it may be spread across three fields. This type of data structure greatly increases the chances that the address will be washed out of the list during the zip+4 encoding, or become UAA (undeliverable as addresses) during mailing. Let's look at what an address might look like:

Chuck Lafean
PO Box 665
123 Court St.
Floor 2
Auburn, MA 09210

What I have here is actually two addresses, a mailing address and a physical address. If this is run through CASS sofftware (coding accuracy support system, this is a USPS term, describing confirming an address is deliverable and appending a delivery point barcode), both addresses are probably good. Great, right? No, the physical address 123 Court St. may not have a mail receptacle, and if it does, floor 2 may not. The correct mailing address in this case is:

Chuck Lafean
PO Box 665
Auburn, MA 09210

Here is a basic template for a simple mailings list: http://spreadsheets.google.com/pub?key=pG1NFxLqU5zi5m9gi7U-xZw. Simply copy and paste the header into a blank spreadsheet.

Data Storage

At this point in the process, you are just storing data. The structure of the data is very fluid at this point. That is you really don't know what you are going to end up storing- the number of fields, what the fields are going to contain, etc. For this reason, I recommend using Excel or some other spreadsheet program.

A spreadsheet is ideal at this point because it is very forgiving and adaptable, and adjustments can be made rapidly. With a database, it is necessary to know what you are going to need to store from the very beginning. I'm sure someone will disagree with that statement, and I would admit that yes, database programs can be adapted after creation. However, I have yet to find one that automatically updates across a schema when adjustments are made. If you have one, go ahead and use it, but just remember, you are now the database administrator- do you really have the time for that at this point?

Another primary advantage to using a spreadsheet is that virtually everyone can operate the program. Even if they have never used the program before, they can be quickly taught how to enter data.

While ultimately you might want to create a relational database structure, that is not necessary now. A spreadsheet is easy to break up into parent and child tables at any point in the future and even on an ad-hoc basis.

Additionally, Excel provides easy access to ad-hoc queries and even cross-tabs (I realize this is perhaps overly technical, if you don't know what I mean by these terms, that's ok, don't worry about it.)

In short, Excel provides all the power you need future, and the simplicity and speed you need now.

A note to geeks:

It may seem like you would want a more robust database program to collect this data, and eventually you might. But unless you are already well versed in that program, taking time to learn is a momentum killer at this point. Now is time for action, not steep learning curves. Hiring the development out is not any better. First of all you can't even design the schema of the database yet so paying someone to develop the database is a waste of money; money that should be spent communicating and collecting additional data. Second, database development can be pretty time consuming, and at this point in the process you don't have the time.

Excel allows for rapid collection of data in a very free form way and with very little training. This is exactly what is needed at this point.

Anecdotal Data

As we begin the process of Relevancy Marketing, we begin to add the data that will us to determine what is relevant to a specific client or prospect. While CRM and database marketing rely on historical and modeled data, Relevancy Marketing generates it own data.

At the beginning of that process is anecdotal data. In other words, what the company thinks they know about a particular customer or prospect. For example;

1. Rank this customer, from one to five, in terms of how recently they have done business with us.
2. Rank this customer in terms of how frequently they have done business with us.
3. Rank this customer in terms of how much volume this customer has done with us.

This is an RFM analysis for those paying attention.

other questions might be,

• Rank our account penetration of this account, that is how effective have we been at selling to all the users of our products or services in a particular account.
• Rank our account saturation for this account. This is the number of product/service lines we sell to an account- are they buying everything they can from us?

All ranking should be done on a scale of 1 to 5. This allows for distribution in quintiles or groups of 20%.

Other anecdotal data:

Business to business:

• SIC Code or Industry Type
• Geographical data
• Company sales volume
• Number of employees
  

Consumer:

• Geographical data
• Lifestyle info- what do they do for fun
• Income
• Homeowner Status
• Marital Status
• Birthday
• Anniversary
• # of Children
• Children's name
• Children's birthdays

This data is going to be used in three ways; we will use it to segment our list, and we will use it to develop a profile of our customer and we will use it when we test the effectiveness of our marketing.

It may appear that some of this information may be very private, and it is. As you begin to collect this data you need to become very conscious of security. How you store it is another potential challenge. I recommend Excel or some other spreadsheet at this point.

Collection of this data should involve anyone who might be able to contribute. If your company has sales reps, they should provide info about just their customers, thus breaking up the task. Print out a customer list and have them rate 10 percent of the list daily of 10 days. Try to make this as unobtrusive as possible. But speed is important. This should be completed with a two week period. The data entry should be complete within and additional week. You may need to hire a temp to accomplish this, but at this point speed of execution is critical.

Concurrently, data from the first mailing should be entered as well. Mail pieces with undeliverable addresses need to be purged or corrected. Responses need to be entered as well. Simply add a column in the spreadsheet for First Mailing Response and put an x or a data in it. We'll use it in a bit.

I guess I should mention that if you have access to historical data in your system that you should use it only if you can get at it quickly and without outside help and you can get it into the spreadsheet with the rest of the data. Your relevancy marketing database should include more than just current customers, but prospects, suspects, etc. Again, at this point speed of execution is more important than accuracy- that will come.

Get Started

One of the centerpieces of Relevancy Marketing is to get started. Relevancy Marketing is designed to move a company, who by virtue of their place in the market, a lack of staff or a corporate culture that prevents it, from no, little or ineffective marketing into a long-term strategy of increasingly customer-centric relevant marketing communication.

Since these organizations lack the historical focus on this type of marketing, we can expect very little in the way of reliable customer data. This consequently leads to a lack of well developed knowledge of who the customer is for the purposes "database marketing".

As a result, the first step of any relevancy marketing campaign is to do a mailing. This may seem contradictory, but I submit that continued complacency is worse. The concept goes like this:

Quickly collect a list of all current and past customers, prospects, chamber of commerce lists, vertical market members, etc. This list may seem next very low quality and it is, but we are going to use it to collect data. Have the list NCOA'd, that is updated to reflect change of address orders, cleaned, and duplicate entries eliminated.

The mail-piece should be a 4-1/4 x 6 postcard, with very generic information on it (but including a phone number). This card is the mail equivalent to a 12-second "elevator speech": we are, we do. This card should be mailed at a first class postage rate, which, when presorted equals or is less than mailing the same piece at bulk rates. It also acomplishes something else that is very important. You will also get the cards with any remaining bad addresses back at no charge. You have begun to collect data!

The trick is not to get hung up on anything. If the list universe is over-thought, or the message over-thought, you will stay right where you are- doing nothing. Just get something out.

This mailing accomplishes four critical things:

1. we are starting,
2. we should generate some leads,
3. we will end up with a very clean mailing list,
4. we have begun to develop momentum.

While that mailing is being prepared and mailed, the next step begins: the collection of anecdotal data.