Amazon’s Trying a Fascinating Price Display Change

Earlier today I was over on Amazon, doing my daily obsessive check to see if anyone’s written any new reviews of any of my short stories or novels (I’m a writer on the side) and I noticed something pretty fascinating. The change is subtle, but it’s definitely there. Also I’m not sure if it’s a test or an a full-fledged rollout … but as minor as it may appear, Amazon has made a major change to how they are displaying product prices in search results.
Read more

Totals and Subtotals in Baskets May Just Waste Space

A few weeks ago I tweeted that I was going to try what I considered to be possibly a “crazy” site design / usability test. In fact, the idea at the time seemed so crazy that at first I limited it to a very small amount of customers, as it was such a departure from norms that I didn’t know what to expect.

I’ve long argued that the whole idea of our classic concept of digital shopping carts on web sites is a very backward way of shopping – especially in the art of increasing average order value. The whole concept of giving people a running total of how much they are spending just seems to run counter to encouraging continued shopping.

Imagine that if when you were shopping at Target you got a continual update on your cart showing you how much you were spending. You’d likely keep a much closer eye on it, and work harder to stay within your budget, and let that continually increasing total ticker start to limit your buying behavior. If every time you threw a new thing in the your cart you received an update on just how much all that stuff was going to cost, you’d probably end up with less in the cart. Instead, since our physical carts don’t have digital price displays on them, we tend to shop in physical stores with the intent to buy what we want to buy, maybe keeping a rough running total in our heads, but mostly we just get what we want to get – then maybe at checkout we’ll put an item or two back once we see that subtotal start to go higher and higher.

Could we increase average order value by not continually showcasing an increasing $ total every time someone adds an item to their cart?

So, the theory behind my test was that by employing the same tactics we could increase average order value by simply not continually updating and showcasing the order total as people shop, but instead leave all calculations of totals until late in checkout – where people have already become so invested in the purchase that cart abandonment will be minimally affected (and still allow people to edit their cart at that time, without having to restart checkout).

The test for this was simple, in our shopping cart I just hid all sections that display totals and subtotals and only left the line item prices in place. Yes, people could still add up their totals in their heads, but there wasn’t a continually growing order total value to scare people into stopping their shopping. We ran the test for a few weeks, and the results … the results showed that it made absolutely no difference to buying behavior.

That’s right, the big area on the page devoted to displaying all the calculations for totals made absolutely no difference in buying behavior. One possibility for this could be that my theory isn’t valid, and that order totals (especially for our low ticket items) really don’t affect how people purchase – they just buy what they want when they want it and aren’t turned off by increased basket sizes. What was clear though was that for the most part this section of the basket mostly just took up extra space – but even with the side-effect of moving the “proceed to checkout” button higher on the page (it’s normally below our subtotals), there was no effect.

In the end we stayed with things how they are, as we hadn’t tested what happens when there is an active promotion that requires a minimum purchase threshold to get a discount – that’s the one instance where we could see it actually having importance.

We’re considering additional tests to see what happens when you pull focus away from price and put it much more on the product. A new test is being thrown around of not even displaying price on product display pages (all of our products cost about the same, and we aren’t competitive on price since we sell a premium product) – but that one is going to be a bit harder to sell internally. Another followup being considered is a version of the basket without totals and subtotals, as we previously tested, but also hiding the individual line item prices.

… think of all the time and resources that could have been potentially put toward projects that actually do matter in the customer experience …

In the end the most surprising part of this test though was that it really had no impact at all. It’s in instances like this that I’m most often reminded that just because something is always done a certain way doesn’t necessarily mean that it’s doing any good. Why do we show subtotals in our baskets? Mostly because it’s standard to do it and we don’t really think about it. But if it really doesn’t matter at all, think of all the time and resources that could have been potentially put toward projects that actually do matter in the customer experience, instead of building out potentially complex calculations for a place where they serve little to no value.

Balancing Personalization and Consistency

I’m down in Philadelphia at the annual Monetate Summit for a few days, and while I’m here I’m obviously immersed in the world of personalization. There are quite a few great conversations going on here, with lots of great ideas being tossed about – and it’s doing a great job of getting my neurons firing about how I can better work towards a personalized experience with Musicnotes. Not anything particularly new per se, but still it’s already gotten me to reinvigorate a few projects that had stalled.

One item that has come up several times here that I still haven’t been able to pull the trigger on, even though it sounds great in theory, is the idea of manipulating site navigation so that it adapts to a user’s preferences or behavior. The example I give for this would be if you are a sporting goods site, if you usually have “baseball” as a main category in your navigation, but based on customer preferences or shopping behavior, you can assume they play hockey, should you change out that link from “baseball” to “hockey” in the main navigation?

I’ve heard it argued that yes, you definitely should, as it’s a service to the customer to give them quick and easy access to the category they have shown a propensity toward – plus you’re removing “noise” of a category the customer may not be interested in. I can definitely see this being useful, and can understand the argument (and in fact have made the same argument myself), but I still keep coming back to the final decision that it is not a smart move.

By keeping navigation elements static we can bring about  sense of ease and familiarity with the site that helps the customer remain comfortable shopping there.

My reasoning for why I still haven’t pulled the trigger here is this: site navigation is one of the few areas of a site that remains throughout a shopping experience. It’s part of the header (or sidebar) and frames the rest of the experience. Yes, it is an important tool in helping the customer to find where they want to go and to take them there, but it also acts as a kind of “landmark” for the customer as well. By keeping navigation elements static we can bring about  sense of ease and familiarity with the site that helps the customer remain comfortable shopping there.

… the “Who Moved My Cheese?” effect.

One thing you might have noticed if you do any site experience tests is that oftentimes when you make a change on your website you’ll see an immediate impact which will fade out and level off in a short amount of time. What you’ll often see, if you dig deeper, is that this impact can be more noticeable for existing, regular customers than for new customers. Mostly this is due to the fact that regular customers already know your site and the experience and how everything works, and when you change things, they can get lost as they regain their bearings (think of this as the “who moved my cheese?” effect). New or infrequent customers will have a lesser impact spike as there isn’t anything to compare against other than your other experiences – there is no learned behavior that you are asking them to adjust or unlearn.

This is where I run into the problems with changing navigation. Even if your navigation isn’t particularly well-suited for some of your smaller segments of customers, forcing them to dive deeper into the site to get where they want to go, at least they know how to get there. (For new customers, you probably don’t even know enough about them to even make a good decision as to how to change their navigation, since they are new and unknown). Everything is in relation to somewhere else on the site, and as customers use the site they learn where it is. Kind of like how you know exactly where on your desk that certain letter is, even if no one else can find it – since you put it there, you’ve taught yourself where it is by your own previous knowledge.

Then again, maybe people can adapt to changes, and by having a constantly adjusting site navigation experience that always does its best to put relevant navigational links in front of customers you end up with a better overall experience. The question is, where do people go to look for what you hid on them? What happens if I normally shop for my son, and he loves baseball, but I was just last week shopping for a new set of hockey skates for my niece … and now I’m back on the site looking for more baseball? If my behavior has caused the site to replace my usual quick access to baseball with hockey, will I be frustrated that I can’t easily find my baseball area any more? If the category is buried too deep and the navigation can’t get me back to where I feel comfortable, will I abandon?

As I said, it’s an area where I can see the arguments both ways – and so far I’ve stuck with the more conservative approach. I can see how a constantly morphing experience can always treat the customer with the right content at the right time, but  does this attempt at personalization end up disorienting the customer?

Sure there are baby steps you can take, like “quick links” to category pages, etc. – but I’m more curious as to what happens if you go all-in and adjust the navigation completely depending on who your customer is. Will they be better off? Or just confused?

… what happens when they call customer service and the support team can’t guide them through the site …

Even worse, what happens when they call customer service and the support team can’t guide them through the site because the customer’s experience is completely different from what the support staff sees? If there are no anchors or landmarks, how does one get his or her bearings?

After this week, I’m thinking of giving it a go on my own – as I’ve definitely heard some good arguments for it, so much so that they might outweigh the possible negatives. I guess the only way to find out for sure is to take the safe route and test.

If you’ve done anything in this area, I’d love to hear from you. How have you modified your site experience to adapt to customer behavior on-the-fly?