In its article, Dealing With the Dreaded “A” Word, Pamela Barden writes about a topic that really needs to be addressed more.   Lots more.

Attrition is a reality in our lives every day.  For example, my bank account is going to deal with attrition because my car’s Check Engine Light came on today.  NFL teams deal with attrition every time they are assessed with a penalty or the quarterback get sacked.  Then again, not all attrition is bad.  I’m hoping that by spending a few extra minutes on the treadmill each day, the couple of pounds I put on over the holidays will attrit right off my hips.

While we address these issues on a regular basis and come up with ways to slow <most> attrition, too many non-profit organizations ignore the problem.

While you can’t always compare one organization’s attrition rate to another, here are a few statistics that are fairly standard:

• Of the New Donors that you fought to acquire last year, 65%-70% will not give again this year.
• Of your core Multi-Year Donor segment, 35%-40% will not give again this year.
• Of the donors you reactivated last year, 60%-65% will not give again this year.
• Of all the donors who gave to your organization last year, around 50% will not give again this year.

But these don’t have to be standard metrics.  By embracing your inner Sherlock Holmes and doing a little sleuthing, you can get a betting understanding of WHO is attriting, WHY they are attriting, and WHEN attrition rates are highest.

What was the campaign topic on which they were acquired?  Some of your donors may have been joined your organization because of a certain campaign topic.  If you are no longer talking to them about that topic, they may leave.  What channel drove them to making their first gift?  Just because their first gift was an online donation, doesn’t mean they only want to be communicated with online. 

Do you see months with peaks in the numbers of donors lapsing?  You will almost always see a peak in the same month in which tend to bring in the most donors.  Be proactive!  Identify at-risk donors several months before they lapsed and contact them. 

There are many ways in which you can lower your organization’s attrition rate, but it involves developing a separate attrition strategy. 

As the saying goes:

To have what you have not, you must do what you do not.

It was a warm summer day in 1974 when I found myself at my very first Major League Baseball game.  It was Shea Stadium in Queens, NY and I was with my best friend and her family.  I was 9 years old and life couldn’t be better.  Well, I guess it could have been better had the Mets not lost to the Cubs that day.  And so began my love of baseball and the torture of being a Mets fan.

Now baseball, more than any other sport that I’m aware of, is inherently connected to statistical analysis.  If there’s a baseball metric that you can think of, I guarantee there is someone out there tracking it.  In 2002, the Oakland A’s turned the baseball world on its head when they applied statistics in a very non-traditional way.  The A’s were faced with the reality that they were a $41 million team that needed to compete against teams with much larger payrolls (the Yankees had a $125 million payroll that year).  Through the use of sabermetrics (derived from
the Society for American Baseball Research), the A’s discovered that slugging percentage and on-base percentage were more correlated to scoring runs than the traditional metrics of batting average, RBIs and stolen bases.  This allowed them to find and sign talented, but under-valued players.

The book, MoneyBall, by Michael Lewis tells the whole story.

Did it work? Brilliantly.  The A’s won their division in both 2002 and 2003.  The analyst in me loves this story, but the baseball fan in me quickly saw a problem.  A big one.  While the A’s did make it to the playoffs, they never made it past the first round. One of the reasons is that I believe they had simply become too reliant on statistics and were not risk-takers. Playoffs are all about taking risks and if you don’t do that from time to time during the regular season, you’re going to be at a disadvantage.

Interestingly, non-profit organizations seem to trend in the opposite direction.  Instead of applying statistics and analytics for long-term gains, each campaign is more or less treated as a play-off game:    going for the one-time win each month.  These organizations are usually end up running in place.  However, the organizations that study their retention rates and metrics related to getting a second gift and develope strategies around those metrics will see steady and
consistent growth.

The inspiration for this post comes from excellent blog  Cool Data Blog which reminded me that MoneyBall is being made into a movie.  Baseball and analysis, I ask you, could there be a better script for a movie?  OK, the fact that it stars Brad Pitt doesn’t hurt either!   Here’s the trailer:  MoneyBall Trailer.

Oh, somewhere in this favored land the sun is shining bright;
The band is playing somewhere, and somewhere hearts are light,
And somewhere men are laughing, and somewhere children shout;
But there is no joy in Mudville – mighty Casey has struck out.*

Now, despite what I said earlier, statistics are still important in the post-season.  This became painfully clear to Boston fans during game 6 of the 1986 World Series.  Going into the bottom of the 10^th inning, the Red Sox were up by 2 runs and only 3 outs away from taking home a title that had eluded them for decades.  Despite the fact that he was only batting .143, and despite the fact that he was playing on two bad ankles, and despite the fact that there was a better defensive player on the bench, the Sox manager made the fateful decision to leave the aging Bill Buckner at first base.  As the Mets’ Mookie Wilson swung the bat, I’d be lying if I told you that I knew who was more shocked – Red Sox or Mets fans-as the ground ball rolled slowly down the 1^st base line, passed the bag…and through Bill Buckner’s legs… allowing the winning Mets run to score.  Two nights later, (unlike that summer day in 1974) there was much rejoicing in Queens.

*From Casey at the Bat by Ernest Lawrence Thayer

• The world’s first parking meter was installed in Oklahoma City on July 16, 1935.
• In Bristow, Oklahoma it is against the law to serve water to a customer in a restaurant unless one peanut in a shell is also served.
• The first shopping cart was unveiled in Oklahoma in 1937.
• Whaling is illegal in the state of Oklahoma.
•  In Tulsa you may not open a soda bottle without the supervision of a licensed engineer.
• The town of Beaver is the Cow Chip Throwing Capital of the World. It’s here that the World Championship Cow Chip Throw is held each April.

And….according to a recent article in Parade Magazine, Oklahomans are the most charitable in the US with an average annual giving close to $1,600.

Now before you run off to your list brokers and re-route all your prospect mailings through Oklahoma, I want to share a quote from 19^th century British Prime Minister Benjamin Disraeli: 

There are three kinds of lies: lies, damned lies, and statistics.

To illustrate this point, I looked at the 2010 revenue by state for one of our clients.  When I calculated the annual MEAN giving by donor, it wasn’t Oklahoma sitting on top, but Utah.  Seems that the average Utah donor gave $510,000 to this charity in 2010. 

You can imagine my disappointment when I calculated the MEDIAN donor value which was a much less fabulous $25 – pretty much in line with the other states.

This is a common phenomenon when looking at states where you may not have a lot of donors, but that also have one or two mega-donors.  This is why when looking at measures of central tendency, it’s supremely important not to overlook the MEDIAN (the point at which 50% of the donors fall above and 50% fall below).  Journalists tend to “overlook” the median because the results are rarely as exciting as the mean.

So for some final words of wisdom: don’t ever take statistics from journalists or politicians at face value.

Oh, and avoid the town of Beaver in the month of April!

Now that you’re set, we’re going to have a paradigm-shift in which I’m going to facilitate your empowerment to diversify a culture-change to a donor-centric platform through mailing-smarter 2.0.

Unfortunately, the term “mailing-smarter” has become little more than an over-used buzzword.  And a frequently mis-used one at that.  For example, many people use the term to mean selecting donors that will maximize the ROI on a particular campaign.  That is not “mailing-smarter”, that is “maximizing the ROI on a campaign”.

Mailing-Smarter is a big-picture project and involves, not just the upcoming campaign, but at least an entire year’s worth of campaigns.

Below are five steps to Mailing-Smarter for the long-term:

I.  Identify Your Organization’s Key Performance Indices
KPI’s vary from organization to organization and depend on the group’s donor structure and long-term goals.  A good way to get started with identifying your KPI’s is with a file audit.  Some common KPI’s are:  # of New Donors acquired annually, retention rates (2^nd Year donors, Multi-year donors, overall), # of active donors on the file (for example, you may want to always have at least 1 million active donors), Upgrade/downgrade %, Annual revenue, Average recency

 II.   What Are Your Goals for Each KPI?
Do you want to maintain the current level or increase them?  

III.  What is Your Strategy to Achieve each Goal?
Do you need to mail certain segments more?  Test messaging to certain target segments?  Add trigger mailings? 

IV.  Select Who to Mail
When it comes time to create a mail file, the first step should be to address your KPI goals.  If your goal is to increase Multi-Year Donor retention and your strategy is to contact all Multi-Year Donors with a recency of 8-12 months each month, then identify these donors and put them aside.  Repeat this process for each KPI goal and put each targeted segment aside.

Once that process is done, then take the remaining universe and do your normal selection process (RFM, scoring, etc.)

V.   Due Diligence on Non-Selected Donors
It’s a good idea to do one last sweep through the donors that did not get selected in Step IV.  An important metric to look for is Last Contact Date.  If you’ve got donors who haven’t been contacted in 60 or more days, you may want to consider adding them back into the mailing.  Bottom line is that they won’t give if they aren’t asked.

The main question is whether a demographic overlay is actionable or does it just provide some interesting pie charts.  A little of both I think. 

House File Mailings
Trying to select donors for an appeal mailing using demographics is tricky.  RFM still reigns supreme in that aspect.  But that doesn’t mean that you can’t still use demographics to your advantage.  What I’m referring to is messaging.  For example, if your file is diverse in age ranges, you may want to test different communications to your 70+ donors.  Or, you may want to do some premium testing basted on donors’ interest codes. 

A word of caution.  Try to keep any demographically based messaging as subtle as possible.  Letters mentioning a donor’s age, income and the fact that they like cats will only serve to creep them out.

Acquisitions
Another way to apply demographics is to first break your active donors into quartiles or deciles based on a metric or your choosing such as lifetime donations or donations in the last 12 months.  Once that’s done, compare the demographic breakouts of the top performing group to the other quartiles or deciles.  If you discover any demographic values that are more present in that top group than the others, you can try to target the sources of those types of donors in your acquisition mailings. 

This information is particularly useful for those marginal lists.  If you knew one  of those lists contained look-a-likes to your top donors, it might be worth the investment to include it in your mail plans.

If any of you have had luck using demographics in other ways, I’d love to hear from you.

Just like analysts, archaeologists have to rely on data so that they have an idea as to where to start digging.  And once they start digging, they have to sift through what’s important to hold on to, and what can be put aside.  Here’s a perfect example:

In November 1922, a relatively unknown archaeologist discovered the relatively small tomb of a relatively minor pharaoh.  Doesn’t sound very promising, does it?  Especially when you consider that all the previously discovered tombs had been robbed in antiquity and a well know archaeologist had already declared that there was nothing else to be found in the Valley of the Kings.  But this archeologist had done his homework and his data showed that there was another tomb to be found.  On November 26^th, that archaeologist was able to make a hole in the tomb, just big enough to pass a lit candle through.  He was asked if he could see anything.  His response?  “Yes.  Wonderful things.”  That archaeologist was Howard Carter and the tomb belonged to King Tutankhamun.

Archaeologists dig for valuable tombs and ruins.  Non-profit data analysts dig for valuable donor segments. 

Using data, archeologists can find new sites.  Those sites then provide a wealth of data that can be overlaid onto existing sites.  For example, since King Tut’s tomb is the only (mostly) undisturbed royal Egyptian tomb ever to be found, it provides archaeologists with data that can be applied to other royal tombs which had been robbed.

In October, I was able to cross off a personal bucket –list item by visiting the ancient site of Baalbeck while vacationing in Lebanon.  Baalbeck is the site of the largest Roman temple complex in the world.  As you can see by the picture below, it’s quite an impressive site.

From Ancient Lebanon: Monuments Past and Present by M.J. Strazzulla

But archaeologists can only do so much to put these sites back together.  Every subsequent civilization uses pieces from the prior civilization, so putting these ancient ruins back to their full glory is simply not possible.  This is where data from other finds can be applied and blanks can be filled.  Below is the same temple, but with a data overlay.

From Ancient Lebanon: Monuments Past and Present by M.J. Strazzulla

 It’s exactly the same concept for direct mail.  Transactional data can give you a strong foundation as to who your donors are, but by overlaying that data with other information such as demographic or psychographic data you will have a much more complete picture.  Transactional data will tell you who your best donors or customers are, but demographics will tell you where to find them.

With that, I hope 2011 brings happiness, good health, and great data archaeology!

See you next year!

Here’s a really helpful short video on the different types of KPIs and how to set them:

As American pastimes go, calculating averages is right up there with baseball and apple pie.  Why?  Well, I think it’s a result of our innate need to classify things (including ourselves) as either above-average, average, or below average.  I mean, who doesn’t enjoy being told they are above average in something.

In the non-profit world, the average that is always calculated and studied is gift amount.  Did it go up or down?  Who is giving above average? 

Calculating an average is very simple and everyone understands what it means, but is it the only metric that should be examined?

The statistical mean (aka “average”) belongs to a category of metrics called Measures of Central Tendency.  This category also includes metrics such as median, mode, and trimmed mean.  While the mean is by far the most commonly used measure, it’s also the most misleading.

Take a look at the following groups of numbers:

Group 1:  1000, 75, 25, 10, 5

Group 2:  250 75, 25, 10, 5

Notice that only one number differs between the two groups, yet the Mean for Group 1 = 223 and the Mean for Group 2 = 73.  Imagine if you were comparing the average gifts from a campaign year over year.  This would be quite a concern.  But that concern would quickly disappear if you looked at some additional Measures of Central Tendency such as the median or the trimmed mean.

The median is simply the value in which 50% of the cases fall above and 50% fall below.  If you sorted your gifts by amount, he’s the guy smack-dab in the middle.  In the example above, the medians for BOTH groups are 25.

With a trimmed mean, you lop off a certain percentage (usually 5%) of cases at both the top and bottom of your data set.  If we remove the top and bottom numbers of both groups, we now have a trimmed mean of 36.67 for BOTH groups.

The other Measure of Central Tendency that I mentioned above is the mode.  This is the value that appears the most frequently.  Since no value appears more than once in the above example, there is no mode.

So, while the mean is easy and understandable, it can also be greatly affected by a few extreme values and thereby potentially leading you to mistaken conclusions.  This is why it’s so important to look at additional metrics such as the median, trimmed mean, and mode.

Finally, here’s an anonymous quote I recently came upon:

Don’t be content with being average. Average is as close to the bottom as it is to the top.

I’m just going to tweak this to make it a bit more accurate:

Don’t be content with being median. Median is as close to the bottom as it is to the top.

Have an above central-tendency weekend everyone!

While I agree that putting your contact strategy down on paper is crucial to any fundraising program, it only makes up the left side of the equation.  Here’s an important maxim to remember:  Just because it’s written down does not mean that it’s happening.

I once went through the exercise of mapping out donor contacts at a fairly large non-profit where I once worked.  They had a very sophisticated multi-channel marketing plan and we literally took over an entire conference room in order to map it out.  It was a quite impressive sight when we finished.  It had “swim lanes” to separate channels and months and the all-important multi-colored sticky-notes.  We confidently concluded that most donors were getting at least X-number of contacts over a 12-month period.  The only problem was that it wasn’t reality.

When I went back to my computer to run reports to verify our beliefs, I found that while some of our donors were getting the X-number of contacts, most were not getting nearly that number and in fact, a large number were getting only 1 or 2 contacts over the 12 month period.  How could we be that off?  Easy.

Donors get suppressed from mailings for any number of reasons.  Maybe they are a new donor that was acquired during a follow-up appeal month in which they don’t qualify.  Maybe they requested not to receive a certain type of mailing.  Maybe they simply got Nth’d out of the mailing.  When this happens, their recency-months will automatically increase.  When that happens, they are less likely to get pulled for the next mailing.  It happens more than you know.

This is why it’s essential to run periodic contact-audits on your file.  Run a frequency distribution on the number of contacts over the last 12 months and then take a look at the donors in the low-contact buckets.  It’s also a good idea to look at how long it takes for a new donor to get their first house mailing.  For this second part, make sure to break the report out by month of acquire.

Designing a robust communications schedule is challenging so it only makes sense to make sure that it’s actually reaching your donors.

We’ve all got our assumptions about direct mail donors and online donors.  We want to be able to neatly put each donor in one channel segment and then treat them accordingly.  Unfortunately, donors have a habit of not always playing along with what makes life easier for us.  I know.  I’m one of those people who muck up the system by saving all my direct mail pieces for a month, pick which ones I want to respond to and promptly go online to made a donation.  But is this normal behaviour or am I just weird? 

A recent article in Fundraising Success magazine sheds some light on this issue.  Here are some of the interesting takeaways:

• 32% of online gifts originate from direct mail promotions (apparently I’m not an oddball after all!)
• Online donors switch to direct mail 30% of the time.  Direct mail donors switch to online 3% of the time.
• Multi-Channel donors retained at an average rate of 16% compared to 7% for single-channel donors.

To read the entire article click here:  http://www.fundraisingsuccessmag.com/article/trends-fundraising-sector/1