Assessing the Success of an Adoption Program Using Return Rates

One of the cornerstones of the animal welfare industry is its goal to place unwanted animals in forever homes.  Permanent placement of the unwanteds is what draws the majority of traffic to a shelter—potential adopters, donors, supporters, etc.  At the very least, this high profile function of an animal shelter should be in working order, if not near perfection.  In that spirit, what metrics should be monitored to understand progress of the program? 

While it is important to know absolute numbers of adoptions for any given period of time, that metric alone should not define adoption program success.  Absolute numbers are appropriate to use in communications with supporters and marketing materials, but should not solely be used to assess the health of an adoption program.   Internally, one should also consider an adoption return rate.  This will allow you to gauge how well your screening and counseling is working and even your follow-up care to adopters.  Obviously, the higher the return rate, the more of an indication something is not right.  There are a few methods to calculating an adoption return rate; in each case the simple computation is the same, but the numbers supplying the division are different. 

First, let’s use an annual faux data set from Shelter XYZ:

	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sept	Oct	Nov	Dec
Total Adoptions	97	88	99	100	78	84	94	99	102	87	85	105
Total Returns	12	13	8	5	11	10	12	9	4	11	5	6
Total Intake	250	249	234	224	276	270	265	245	235	241	223	256

Scenario #1:  Calculating adoption returns as a percentage of intakes:

March = (8/234) = 3.4%

April = (5/224) = 2.2%

May = (11/276) = 3.9%

Scenario #2: Calculating adoption returns as a percentage of total adoptions:

March = (8/99) = 8.0%

April = (5/100) = 5.0%

May = (11/78) = 14.1%

Scenario #3: Calculating adoption returns as a percentage of last month’s total adoptions:

March = (8/88) = 9.0%

April = (5/99) = 5.0%

May = (11/100) = 11.0%

In each scenario the variance of the resulting rates follows the same pattern—May has the highest rate, followed by March and then April, but my preferred method is Scenario #3.  This allows us to consider returns as a function of adoptions and at the same time permits a consideration of time.  While not 100%, calculating return rates in this way will capture more of the actual animals moving through time.  For example, if a dog is adopted on March 15 and returned 20 days later on April 5, it will be tracked in Scenario #3 in total (the dog’s adoption will appear in the denominator and the return as part of the subsequent month’s return).  In other words, Scenario 3 allows for (sort-of) tracking an actual cohort of animals. 

While I chose to represent the stats in monthly format, the calculations will work for any time period selected (weekly, monthly, annually, etc).  I know there is wide variation in the field about what is considered a return (some 30 days, some two weeks, while others is the lifetime of the pet), I would advise any shelter to remain consistent within their calculations so their own data are comparable month to month, year to year, or any other such defined time period.   This is not to say that you can’t calculate it differently depending on your needs, but be cognizant to only compare data from within the same calculation type.   For example, you cannot compare March in Scenario #1 to March calculated in Scenario #3.  

To put your date into perspective, the national average for return rates according to ASPCA Partnership Communities is as follows (calculated as in Scenario #2 above):

·      Total = 6%

·      Dogs = 8%

·      Cats = 4%

How does your data stack up?