Let’s get a little more hands-on this week and walk through something practical that will immediately impact the conversion potential of any web site and increase the performance of both paid and natural search campaigns. The examples I’ll use today are specific to ecommerce, but the same ideas can be translated to other industries, and I’ll close by talking about how to make those translations happen.
Cross-selling is one of the most powerful ways to improve your conversion rate (especially for search traffic), but surprisingly few web sites execute it well—or at all. It works because people think they know what they want, but oftentimes are willing to upgrade or move sideways into another brand if the features are right and the sale is done well. This impacts conversion rate in three ways:
- Upsell: Place higher-margin items on the page to increase profits
- Add: Sell additional items in the same cart
- Catch: Provide substitutes for when the customer is underwhelmed and may abandon
Best Buy does a very decent job—does data say it’s the best job?
The first two of these are there to boost performance and the third is there to prevent failure. That third one is often overlooked. Sites get so focused on trying to sell at a higher margin or add quantity that they forget that some customers want to go in the opposite direction. In essence, these sites are saying to the customer, “If you don’t want this or something more expensive, we don’t want any of your money at all!” This unidirectional approach ultimately appears to succeed, but total sales volume is lost in the process.
You might say that the word “underwhelmed” disagrees with a downgrade. Remember, though, that your customers are making value-based decisions. They are not there to review the features of your product; they are there to strike a balance between benefits and price, and if that balance is off, they will be underwhelmed. Thus, they may very well be much happier with a technically inferior product.
When to cross-sell?
- In the middle!
- At the end!
But do it well. Every page presents a cross-sell opportunity, but I’m not talking about in-your-face stuff. Look at it this way: if someone searched for something specific, give that to them. But put yourself in their shoes and say, “Where would I look if this didn’t feel quite right?” Put your recommendations there. You can do this both at the category level (Sony TVs—have a link somewhere to Panasonic, another good brand) and at the product level (Sony XBR 52″—what’s a comparable product our customer should know about?). Once they are in the cart and a product cross-sell is no longer available, start talking about the accessories that this person will need. A lot of sites do this part very well. But it’s okay to be subtle—people will find it.
A real world case study
I’m going to use an example of luxury retailers, because I believe this is one of the biggest cross-sale failures on the internet. The customer base of these sites is rich, indecisive, rich, impressionable and rich. And they’re also rich. And when I say indecisive and impressionable, it’s not because they aren’t intelligent; it’s that they have the resources to make indecision a financial option—just get it all! And it’s not that other sites like Best Buy or CDW don’t share this same affluent audience, but those sites definitely don’t have the homogenous population the luxury retailers do. And sites like Staples.com have to compete with the fact that personal interests and emotion don’t affect sales: people have paper and stapler budgets to work within. Yet all three of these examples do a relatively good job of cross-selling, certainly better than sites like Saks, Neiman Marcus, Bergdorf-Goodman and Barneys (who is showing purses for a shoes search).
First, let’s take a step backward. To me, the most alarming part of this whole comparison is that this situation is completely reversed in the real world. At Staples, I get a college sophomore who knows as much about an LCD display as a trained parrot. No effective cross-sell at all, really. When I get to the register, some manager who is having a miserable day says, “You know this won’t work without a USB cable, right, [dumbass]?” Not a great sales situation. But the site does a good job of these things that aren’t executed well in real life.
In Neiman Marcus, on the other hand, you’ll have a personal shopper walk around with you, grabbing things off of shelves and racks, putting things together to show what matches what. You walk in for a pair of socks and walk out with three suits and four shirts that go with the socks. Brilliant. Their web site does none of this, though, and that’s what has me scratching my head.
I suppose the biggest complaint you might have as a reader would be about resources. That something like what I’m suggesting involves investing in an expensive tool or burning a lot of programming hours, and that might be true, but don’t you think those hours will pay for themselves? I would also say that there are a few things you can just do manually, starting small. Carve out 5% of your product pages and change them over to a template that allows you to add complementary or substitute products manually. Pick that 5% using your best judgment, but I’d pick pages that generate significant traffic from search and have middling conversion rates. Throw in a few category-level pages while you’re at it (rather than focusing entirely on individual products).
Using a mix of web analytics data and common sense, pick 3 or 4 links and thumbnails you’d like to put on these pages for other products or categories you think the user would find useful. See what happens. Make adjustments. See what happens.
If, on the other hand, you’re already doing some of this on your site, ask yourself whether you know for a fact that the mix of cross-sells you offer today is factually the best it can be. I don’t mean you’ve come up with some sort of algorithm that is good and automatically chooses things. I mean you have actual, hard data to show that it can’t be done better. When people land on category level and product pages, do the substitute products they choose frequently differ from the ones you’ve chosen? There’s opportunity there.
Whether you have to start small or start big, you need to find a way to start. This is particularly important from an SEO standpoint as well because you’re giving users more control over early stages of their path. While we can control the landing pages in a paid search campaign, we do not have this degree of control for natural search, so you need to find ways to get people to the content they intended to consume as quickly as possible. Researching the commonalities in their paths (that’s the only time I’d ever use paths, which are analytics treadmills) should give you some ideas for where they’d like to go. Take them there!
Translations to other verticals
Content / Media: I’d say this is actually a bit easier than retail, because on content sites, people are primed to ask, “What’s next?” Once you’ve clearly stated where the user is in headers and the first few lines of your content, offer additional topics and stories at the top of the page if this isn’t what they’re looking for. And, obviously, if someone reads the whole article, don’t leave them stranded at the bottom of the page. Put some follow-up or “we also recommend” links there.
Travel: Destinations, destinations, destinations! Travel sites still do a pretty terrible job of helping people assess the landscape of options they should be considering, particularly with last-minute travel. It feels very random to me; it’s too wide. People want the ability to narrow it and not make it self-serve.
Telecom/Cable. This is the only vertical where I’d say people need to dial it back. Search for “cable television” on Google and click through the paid links to landing pages. Puke, puke and more puke. I want cable— not phones, internet, calling circles, 4G wireless, and the lowest rates to Mexico. Control yourselves, people. I can’t even find why I was there in the first place. Maybe if I could, I would consider your other services.
Hopefully that’s enough to get you started. Nothing here isn’t obvious, but sometimes it seems too leviathan to take on. Just start small, get a few small wins, and watch what profits can pay for.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.