“Lions cannot afford to hunt mice because they will literally starve to death, even if they catch them.
Lions and all large carnivores have to hunt game large enough to justify the investment, so they have to hunt antelope and zebra.
Why is this important?
Because most senior executives are really big on chipmunks.”
What are you hunting these days?
I know. I know. Hundreds of people are going to write me back and say that they’re plant-based (known to carnivores as “a missed steak” – ha!) but I am serious.
Are you targeting rats or gazelles?
I ask because these days I see far too many dinky tests and far too few big swings.
Sure, many mega companies can get away with doing one-base tests because they do hundreds, sometimes even thousands, of tests each day. The majority of companies though? To grow their businesses, they need homeruns.
One of the biggest homeruns I saw last year?
Donuts and donut holes.
What’s a donut? A donut is a series of special pages. Sometimes the donut is 3 pages, others it’s 30. Whatever it takes to do the job. (Some folks refer to one page as a donut. I prefer to think of one-pagers as landing pages.)
You may know donuts as wrappers, wraparounds, minis, stringers, micros, hubs, pacemakers or any of a dozen other names. (Most of those names have already been used for other things though so I find that calling donuts things like “minis”, for example, is very confusing.)
You can create a donut for different customers, channels, devices, problems, and so on.
The purpose of a donut is to take a challenge you face on your sites (phone, tablets, and/or desktop) and address it with a very specific pathing experience for the user.
In other words, you figure out where the users are struggling and you come up with a better way to serve them. Granted, it doesn’t always need to be a struggle – it could just be a fantastic growth opportunity.
What’s a donut hole?
These days, folks refer to donut holes as something that happens inside your site — for example, a special checkout for Shopping customers. (Checkouts are considered inside because people typically don’t land on them as the first page of their user path.)
To me, the whole donut/hole thing can get convoluted pretty quickly so my general rule of thumb is that if a user can’t start from it, it’s a donut hole and not a donut.
Donuts and holes originally came into vogue because many companies were disappointed when their responsive sites didn’t automagically work. In other cases, marketers wanted to do things like feature different payment terms on their mobile sites with more grace and less of a performance hit than just truncating PayPal off the bottom of the screen. Or it was because they didn’t want to deal with Catalog Quick Orders or extend credit.
They got even more popular when lead gen companies found how lucrative they can be when used appropriately. For example, if you’re B2B, you may want a totally different experience for your Thomasnet customers than the ones you use for DSAs (dynamic search ads.) You might know that your .gov customers aren’t going to finish their $10,000+ transactions online so capturing the lead info is far more valuable to you. Or, that one subset of your prospects needs a free sample pack whereas the other one needs a free webinar.
Why are donuts so damn tasty, er, successful?
Frankly, most donuts fail. However, many marketers find the whole donut concept rather delicious.
Because it pushes them out of their comfort zones to try “risky” things at the same time as giving them an opportunity to supercharge proven winners.
Marketers tend to get caught up in landing page tests – call-to-action above the fold vs call-to-action below the fold; PayPal vs Visa Checkout. In my experience, donuts tend to take things up a few levels higher. What would we design for Customer X if we were only selling to her? What do we do for Customer Y if we want to get more bang from our buck from him?
Donuts also tend to work in old-school worlds like catalog marketing. Most catalogers are dusting the seats of the Titanic. Their never-ending yapping of how the Co-ops have ruined their business is exhausting. Yet, when you tell them that their online customers have different tendencies than their offline customers they are so busy discussing their imminent demise that they refuse to do anything about it. Introduce a donut test to them – wherein the offline folks get a completely different set of navigation than the online folks – and their ears perk up. The thought of keeping things the way they’ve had it for the past 30 years gives them a lot of comfort. The promise of the new users getting a phenomenal experience as if you designed it for 2017 (instead of 1997) is exciting and often insanely profitable.
I have a good friend who serves as the VP of Marketing for a multi-billion dollar (and prehistoric) brand. He insisted their one-page checkout was THE only checkout that would work for them because they tested it extensively in 2012. He tested donuts to prove me wrong (we bet on whether or not they’d work.) Today, he still has his infamous one pager but he also has over 100 versions of his checkout because of donuts (technically holes, I know.) They have a checkout for Shopping customers, one for Retail Me Not customers, one for people who come from Food Network, one with instant credit, one with a sole alternative payment method, one with fifteen (yikes!) alternative payment methods, and so on.
Donuts seem great and all but how are they a big swing? They seem like old school direct marketing.
Donuts are indeed old school marketing. ON STEROIDS.
You’re right though, donuts themselves aren’t big swings. In fact, you probably did some version of them years ago.
With that said, creating them now in 2017, is a whole different experience.
One of my favorite clients has a very complicated product line with over 800 variations of one product. They tested a few donuts in the Summer of 2015 and now over 75% of their landing pages representing over 90% of sales on that product show three or fewer items. 800 down to 3.
Only one person in the company really believed in the test. Everyone else thought it was nuts and that it would most definitely fail. They had gotten to 800 variations over 20 years and each variation had a story. They were committed to each and every single one of those stories. Now they’re looking at whittling their line of that particular product to a dozen which will shutter millions of dollars in inventory, a building in their warehouse complex and so on. It was a swing for the fences and it resulted in several homeruns. And for the record, they’re conducting this as a two-year test with so many back tests, they need a special database just to keep track of them. But even if they end up with 50 or 100 products instead of 800? It’s a huge savings.
A ferocious lion can kill whatever it wants (even though lionesses do most of the hunting) but can you imagine if he had to feast on just mice? He’d starve to death. Same as you. Move on from the mice.
All our best content is published in our weekly email 2-16 weeks before it’s posted on this blog. (The super-juicy-and-often-controversial stuff never even makes it here.) Sign up for FREE now.