# Ask string amounts: to round or not to round

Rounding can be controversial.  On the one hand, round numbers could potentially help with fluency, which is crucial in persuasion.  Rounding out ask strings can help you get out of weird numbers that consistent upgrading can create (e.g., if you donate \$30, then upgrade by 50% each time, that’s \$45, \$67.50, \$101.25, then \$151.875.  And if you haven’t got a ha’penny, then God bless you).

On the other hand, there is a potential draw in using an odd amount specifically to stop someone short (e.g., your \$17 will feed X people).  In fact, in face-to-face settings, panhandlers found that when they asked for change, 44% of people contributed.  When they asked for a quarter, 64% contributed.  When they asked for 17 cents or 37 cents, 75% of people made a contribution.  This is called the pique technique: the idea being that the odd request breaks people out of their normal mental structures, forcing them to think about what you are saying.

However, this may or may not be as applicable in non-face-to-face environments.  Burger et al took a look at the mechanism by which this worked.  They found that contributions only increased among people who came over to ask a question; there was no difference in giving between people who were given a specific answer (e.g., “I need to buy a stamp”) versus those who were given a generic answer (e.g., “I need to buy some stuff”).  Since there isn’t a mechanism for someone to stop and ask you a question in the mail or online (yet!), this technique may not work on asynchronous platforms.  And, in fact, a study of the pique technique when applied to causes, rather than donating directly to the person face-to-face, found no significant difference with this technique.

There is, however, evidence that rounded numbers can increase giving.  One study of donations found that rounded values increase giving by seven percent.  Specifically, they found that people were more likely to choose things that were on the ask string (what they called an appeal scale) than rounded numbers not on the ask string, but that a good number of people wrote rounded numbers in as the other when not on the ask string.  Additionally, they found when a round number was on the ask string, there was a particularly strong pull of that number on donations.

In addition to articulating that round numbers have a pull that is independent of the pull of a person’s internal reference point and the ask string itself, they also helped define a conundrum.  To wit: what is a round number?  As the authors put it:

“While the notion of a round quantity is seemingly intuitive, it is nonetheless difficult to make precise. Round numbers are operationalized in the present paper as the face values of commonly used French currency notes or small integral multiples thereof. Based on the data used here, this functional definition accounts for all but a negligible proportion of off-scale donation amounts, in the sense that adding or removing additional rounded values does not substantially alter any aspect of the analysis.”

As we noted in the all about the Benjamins piece, these banknote values are especially strong — \$100 in particular.

In summary, fluency seems to trump creativity in the case for round numbers, especially at higher levels.  If you are looking to increase average gift, it should be to something that is highly fluent like \$20, \$50, or \$100; if you have to be creative, do it with lower ask amounts so that they don’t have as much draw.

However, there is some evidence that having one oddball, very high donation amount can be helpful, as we noted with the Make-a-Wish Canada example here.

The idea of providing an outlier as a potential anchor and/or moneymaker is an interesting one.  While I could find no data on this from the nonprofit world, there are studies from the for-profit world that indicate that the existence of one very high product can increase what someone is willing to spend on a medium-priced product.  This explains the existence of those news stories you hear from time to time about how someone has created an even more expensive hamburger, made with wagyu beef, caviar, gold leaf, and unicorn tears.

Medium rare, unicorn tears on the side, please.

Having this burger on a menu, in addition to the free publicity, could also increase average spending.

I, like many, am currently obsessed with Hamilton despite the minor handicap of not having seen it or likely being able to see it in the near future.  (Not shockingly, the idea of someone writing like they’re running out of time appeals to me.)  But the Founding Father most relevant to our work is still likely Ben Franklin.

Do not reproduce.

In nonprofit work, we tend to view the \$100 level as a bit magical.  Once someone makes the psychological jump to three figures, they are likely weighing the effectiveness of your nonprofit and not just the emotion.  Additionally, when someone is giving in the same number of digits, it’s easier to upgrade.  That is, the jump from \$100 to \$200 seems easier for donors to make than \$50 to \$100, even though it’s more money to increase.  And \$100 donors in acquisition are instant candidates for high-touch treatment, as they are showing significant support of your organization from the start.

There’s a lesser known place where \$100 is magic: in your ask string.

Reiley and Samek did a study in which they tried two different ask strings). The first was \$35, \$50, \$75, \$95, \$250, and other.  The second was the same as the first, but with \$100 swapped in for the \$95.

Revenue per solicitation went up 29%.  Average gift went up almost 20% and response rate went up 7.6%.  Let me repeat that: the ask string with the higher gift amount in it also had a higher response rate.  This is a rare thing.

The major reason for this, they hypothesize (and I agree), is fluency.  As I wrote in the piece on word choice:

People tend to prefer things, people, objects, etc. that are easy for them to understand. (study here). This is known as the fluency bias. There’s a reason that only eight names cover more than half of our presidents (James, John, William, George, Thomas, Andrew, and Franklin (which used to be a lot more popular than it is now)). Names that are more common help people rise faster in occupations. Believe it or not, stocks that have ticker symbols that can be pronounced as words outperform stocks that can’t be.

So, you would think logically that \$100 isn’t the only fluent number.  And you would be right.  Another study looked at whether people were more likely to give \$20 or a strange amount like \$20.03 (if they graduated from the college in question in 2003).  People were less likely to give the strange amount (although this was not statistically significant at the .05 level).

So remember to create round numbers in your ask strings unless you have a really good reason to: the ease of recognizing and using these amounts will benefit your organization and get you more of those elusive Benjamins.

When you can, see if you can get the \$100 into your ask string.  It increases your response rate with no loss in average gifts.  Wins like that don’t come along every day.

I’m working on a book on ask strings.  My goal is to make it free to subscribers of my newsletter here.  So if a round-up of the science and psychology of donation amounts sounds interesting, please sign up today.

PPS. Tomorrow would also be fine, as I won’t be done with the book tomorrow, but that lacks urgency, don’t you think?