Best Practices: Improving units per transaction (UPT)

Ask any ecommerce director to tell you what they’re working on and they’ll likely rattle off a few KPIs. Driving conversion, particularly on mobile devices, is typically at the top of the list, but making real strides in mobile conversion is particularly difficult for mature e-tailers – there are just too many variables at play. Instead, many choose to focus on more targeted metrics that are still worth millions, but easier to influence and measure like cart abandonment, customer acquisition cost (CAC), and units per transaction (UPT).

Of these three, improving UPT in perhaps the most challenging. Why? Most ecommerce teams are great at suggesting alternatives. If you don’t like this, you might like this instead. This is the ‘or’ of ecommerce. Product detail pages and pre-cart interstitials spend so much time trying to sell you something similar, it seems they lose sight of the item you’re considering in the first place.

What’s more difficult is securing the ‘and’ – shifting focus from selling you an alternative, to selling you another item. It’s a bold move, but when done well can pay huge dividends.

Another reason driving UPT is difficult is customer momentum. When customers are ready to buy, particularly on mobile devices, they are focused on the task at hand. Getting customer attention, to say nothing of clicks, is difficult. Get too aggressive and you might risk the initial sale. Stay too passive and you might not be noticed at all. It’s gutsy to push for a second sale when the first isn’t in the bag.

In this post I’ll look at three different e-tailers and how they approach UPT. Each has very different products: one is selling a custom luxury product with required components, another a cleaning system designed to be used in a sequence, and yet another an outfit created through a combination of items. While their products and customer segments are unique, their merchandising tactics have a common thread.

Selling a multi-SKU product: Blue Nile’s Ring Builder

Ecommerce innovator Blue Nile was one of the pioneers in selling luxury items online. They’ve been selling diamond engagement rings for nearly twenty years. In that time, they’ve redesigned their build your own ring experience again and again. It’s easier to understand why when you look at their challenges. Engagement ring purchases are complex (most guys don’t know the Four Cs of diamonds), expensive (the average purchase is several thousand dollars), and emotionally fraught (will she like it, will she say yes?).

Complicating matters, Blue Nile won’t sell you a setting without a center stone. They will sell you a center stone without a setting, but they would rather sell both of course. Also, they have over 146k diamonds to choose from. So, choices.


Blue Nile could have built a wizard-like experience locking users into a controlled environment, but they likely know most customers want to dive in and browse first and ask questions later. They wisely allow customers to start with a diamond, or a setting, but they merchandise the finished product (the ring) via a dynamic status bar throughout an otherwise reasonably standard search and detail page-based experience.

The status bar not only gives you a sense for your active step, it also coaches you on what’s next. As you complete your steps it updates allowing you to revise your previous choice. Throughout, the chevron laden status bar gives you a sense of momentum.


However, Blue Nile’s not content just selling an engagement ring. While the Average Selling Price (ASP) might be high, diamond margins are somewhat low. That’s where wedding bands come in. Throughout the ring building experience they also call out which setting has a matching band. While most guys wait to buy the bands (wanting to secure a “yes” first) Blue Nile is seeding the idea early.

From an experience perspective, Blue Nile is threading a needle; taking care to leverage existing shopping patterns, while weaving in custom interface elements, messaging, and a little cross-merchandising to suit their needs. We’ll see some similar moves in our next example.

Selling systems: Griot’s Garage’s Polishes & Compounds

Selling car polishing systems might not sound as sexy as diamond engagement rings, but Tacoma-based etailer Griot’s Garage has legions of passionate fans in the valuable car enthusiast segment. With tens, if not hundreds, of thousands invested in their cars, a simple wax and polish solution just won’t do. The Griot’s system includes polishes for use by hand or machine and different options depending upon your need. Sure, you might be here for the correcting cream, but what’s the point it you aren’t going to wrap up with the finishing sealant?

The Griot’s team wisely features an inline educational panel in their browse experience informing customers how aggressive their solutions are on a scale. This inline guidance covers most questions without taking customers off the buying path. For those who still have questions a comparison grid is a click away which wisely sets you back onto the buying path when your comparison needs are met.



Go a click deeper and Griot’s efforts to sell more don’t stop. A simple ‘frequently bought together’ module wisely pairs an item that likely doesn’t need further scrutiny with a single call-to-action to buy both. Have a quick question? A quick view provides more details while keeping you on the path.



Selling outfits: Michael Kors’ Shop the Look

Selling outfits is no small task. Fit issues alone pose huge operational challenges (thanks Zappos for free returns) and fashion, as they say, is fickle. Merchandising alternatives over accessories is just the safer play. But Michael Kors isn’t without ambition having acquired Jimmy Choo and Versace in the past two years.

The team at Michael Kors borrows from the same page (literally) as Blue Nile and Griot’s Garage – it merchandizes multi-item purchasing during the browsing experience. Interspersed with the traditional product grid are double-height modules featuring an entire look. It wisely integrates the shop the look into the browse experience as that’s where the traffic is.



From the browse page, customers can still shop for single items (which wisely cross promote the look from a reasonably standard product detail page), but they also can shop the look on a special hybrid detail page that smartly weaves the essential detail page selection items (size, color, quantity) onto a single page.

Customers can easily buy them all in a single click, or individually select the items they want. Being a multi-channel e-tailer they’ve also woven in a pick-up in store call-to-action as well.



By now the common tactic is clear – inline merchandising in the middle, and not the bottom, of the funnel.  It’s not enough to weave accompanying items into a traditional product detail page. By then the users have too much momentum toward their choice. Experienced retailers know you need to plant the seed while the ground is still fertile, while the customer is still in consideration.

Depending on the constraints of your brand, and your margins, some retailers offer financial incentives to sweeten the detail. This typically comes in the form of bundle-and-save messaging or free shipping thresholds. Experiment with these avenues carefully. Once you start offering discounts, it’s hard to stop.

Even if you don’t convert the customer on additional units in the initial sale, the database relationships between complimentary items can be useful down the road. Re-targeting ads, email campaigns, and on-site personalization all give you another swing at the plate, but be sure to put the customer in charge; letting them shape their preferences.

Done right you can drive UPT numbers without sacrificing conversion of the item at hand. It takes consideration, patience, and plenty of testing to find the execution that works for you and for your customers.


Like what you see?

We’re always looking for interesting challenges and ways to provide value.
Contact us today and let’s talk about what we can do for you