[:en]Discussion of best practices for online marketers.[:]

Facebook points to the future of personalization

Facebook logo

Looking to Facebook to get a peek into the future of the head-spinning evolution of the digital world is hardly sheer genius.

But it is at least wise — not so much because Facebook is the most-cutting edge enterprise when it comes to digital innovation, but more because Facebook is brilliant in understanding where consumer desires and expectations are headed.

Like any hugely successful market leader, Facebook is not always first on a trend, but it is sometimes first on understanding how to combine already identified trends. Its MO of late has been to buy the technology it needs to build its vision of the future.

That’s why I was so intrigued with the stories coming out of Facebook’s F8 conference in San Jose, Calif., this week. Take the story that appeared in the Mercury News of San Jose concerning changes to the company’s Messenger app.

There were a lot of little pieces mentioned in the story — adding mobile payment capabilities to games, making Messenger more like Snapchat (if you can’t buy ‘em, join ‘em), accommodating those who want new ways to access commerce bots or find businesses or order food.

What the story didn’t expressly say, but what the story illustrated, is that Facebook is charting the course of the future of commerce. And at the center of it all is personalization. Facebook is tackling that with the vast trove of information it has on its members.

Facebook has the advantage of being able to connect every online action with a known individual. Users sign on to Facebook. It has no need to segment or target. It can develop one-to-one personalization because it knows exactly who the “one” is.

The Mercury explained that Stan Chudnovsky, Facebook’s head of product, believes that Messenger will remain a go-to by using artificial intelligence to improve the personalized experience it gives users.

Personalization applies to content, products and even friends

That kind of personalization, of course, can apply to content or products or even finding compatible friends who users don’t even know. But take that kind of personalization and connect some of the other dots in the Merc’s piece and the story becomes all that more interesting.

For instance, Chudnovsky, speaking at F8, also said that FB was unveiling a feature that would make it easier to have group chats with bots, the story said. The app will allow groups to summon bots to share music from Spotify and to make restaurant reservations.

IDC analyst John Jackson told the Merc: “You want to meet the customers where the customer is, and that’s where we are in messaging apps. For businesses, it’s a very logical way to engage in that way, and they’re taking very significant strides in making bots available.”

Meet the customers where the customer is. It’s the very mantra of the omnichannel retail crowd. But for Facebook it might be a more about keeping the customer where the customer is — namely on Facebook.

So far, so good for the Menlo Park company. According to the Mercury News about 20 million businesses are already connecting with users on Messenger. And the company says Facebook expects that number to grow, which makes sense, especially if the latest moves are in line with what consumers are looking for.

The improved app make Messenger more like China’s WeChat, a messaging app that has become a huge commerce platform because it allows such flexibility and because China has long been a mobile-first digital culture.

Facebook, meantime, has been pushing into e-commerce for some time. And so it’s no small thing that the Silicon Valley social network company is trumpeting personalization as one key to its future. There is little doubt that personalization is also a key to the latest era of digital commerce and communication.

Those who offer content and products on the web will no doubt be staking their futures on finding the right recipe for personalization.

The reported announcements at F8 touch on another intriguing puzzle for both digital commerce and digital content providers: the seemingly contradictory chore of figuring out how to personalize for groups.

The features enlisting bots for creating playlists for groups or allowing group ordering, for instance, might well need to offer personalized recommendations for those groups, as oxymoronic as that sounds.

Chris Martin, chief technical officer of Pandora raised the idea of coming up with group personalization at the Structure Data Conference last year. What do you do, he asked, if you have an algorithmically aided music recommendation engine that is playing to the crowd, so to speak? Martin suggested the machine might add context clues to the mix to come up with the answer.

For instance, I could see a machine that gathers signals that would help it consider whether a number of people had gathered for a birthday party, a sports tailgate party or a funeral.

It’s all a reminder that personalization is likely to be one of those things that marketers never quite master. Just as the technology to achieve true one-to-one personalization becomes available, it could be that it’s time to work on one-to-many personalization.

None of which is reason to be discouraged. In fact, it’s reason to be energized, as we race to get ahead of the future. All the while, looking to Facebook to show us the way.

Facebook logo courtesy of Facebook.

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.


What site merchandisers can learn from Buster Posey and the San Francisco Giants


Wouldn’t you love to click on a story with that kind of headline and find a one-word answer: Nothing.

Yeah, the, “What ‘X’ can learn from ‘Y,’” construct is a pretty tired concept. But hear me out. The San Francisco Giants take on the New York Mets in a winner-takes-all, loser-goes-home wild-card playoff game tonight that will propel one team, one step closer to baseball’s World Series.

And veteran Mercury News sports columnist Mark Purdy likes the Giants’ chances. What’s the source of this confidence? All-star catcher Buster Posey. Well, and, hall-of-fame-bound manager Bruce Bochy. And the way the two use easily accessible and actionable data.

But not just data. The secret behind Bochy’s, Posey’s and the Giants’ success is the way they combine data and gut; science and art; the machine and the human.

It’s what modern merchandising is all about, or should be all about. No doubt, successful site merchandisers know their art — their brand’s story, their customers’ likes and dislikes, the emotional connection consumers have with brands, products and even shopping itself.

But they’ve struggled getting their hands on the data they need. Only 37 percent of merchandisers surveyed by Forrester on BloomReach’s behalf said they strongly agreed that they had access to the data they need to make key decisions.


And, not surprisingly, that lack of vision leaves merchandisers’ teammates — namely marketers — with little confidence in merchandisers’ ability to provide useful insights into customers.

Baseball, that timeless game, offers lessons and a strong argument for better outfitting merchandisers with the tools and data they need.

From Purdy’s column about the Giants’ 2016 foray into the postseason:

“Every team has piles of numbers and statistics to ponder before a series begins — or in a single-game winner-take-all situation, such as Wednesday. The Giants leave a printout of their opposition scouting report at every locker. There is also a players-only website they can consult to call up granular stats on every pitcher or batter. Posey can look at the site to see which pitches an opposing batter most frequently swings at and misses. He can dial up video of every at-bat an opponent has had against Madison Bumgarner or any other Giants’ pitcher.”

And what does Posey do with these buckets and buckets of data? Again, from Purdy:

“But in 2014, I learned from Posey and other Giants that while he consults all of the information, he takes it to the field and selectively ignores it. Posey will respond to what he’s feeling in the moment — from his own pitcher, opposing batters, umpires, even the weather conditions.”

Art and science. You hear it again and again from top merchandisers. And while the art, like baseball, is timeless, the science moves forward at an increasingly dizzying pace.

Laura Freedman, who worked as a merchant before launching the e-tailing group in 1993, talked to me about the changes for an earlier story.

“When I started out, there was no data,” she said. “You just had to understand it. But I think having the data helps: What gets positioned where; What gets shown first. And then you can use it in your favor.”

Most merchandisers would no doubt agree that data is invaluable, but it works best when paired with their instinct, knowledge and understanding of human nature.

It’s a winning combination — and no doubt a combination that retailers will want to have at the ready as they enter the crucial holiday shopping season. In fact, you could argue that the holiday period is retailers’ post-season: a winner-take-all, loser-go-home contest that can make or break a franchise.

Photos by Mike Cassidy

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.


Digital merchandisers can break down data silos: Part two of four


If you follow retail trends, you’ve heard over and over how e-commerce, and mobile commerce in particular, has completely revolutionized the way consumers shop and buy.

What you haven’t heard as much about is how that same revolution has upended retailers and those who work for them, dramatically transforming roles and in some cases devastating businesses that were too slow to adapt.

For a look at the heart of the disruption, consider the job of site merchandisers. They were once called upon primarily for their sense of style and fashion — and for knowing what the season’s big seller would be. Now, they are expected to be all that, plus take on an analytic role, using data to determine winning and losing products.

It sounds like a fascinating job, but there is a catch: Recent research, commissioned by BloomReach and conducted by Forrester Consulting, indicates that merchandisers often don’t have the tools they need to master their new data-driven duties. In fact, the survey shows, modern merchandisers are being underutilized and, arguably, underappreciated.

Caitlin Molinari, who built a career in retail merchandising and marketing, says she’s seen the pigeon holes that site merchandisers can find themselves in when they are overwhelmed by data — and all but unarmed in terms of analytical tools.

See the entire site merchandising series

“They need to be almost let loose a little bit and have more trust put in their roles,” Molinari, who’s worked for Nordstrom, Pacific Sunwear and Teefury, says of digital merchandisers. “They’re making these daily decisions in merchandising and creating those unique experiences. I really don’t think they’re set up for success.”

In this, part two of a four-part series on the future of site merchandising, we look at the potential for properly armed merchandisers to break through the various silos of data and thinking that rise up in e-commerce organizations.

Molinari, who recently founded Caitlin Molinari Marketing, says merchandising has too often been seen as a “plug-and-play” position. Executives, buyers and others turn to merchandising with a list of products, brands, styles that need to be promoted based on corporate goals or past buying decisions.

“But really, they have so much data at their disposal,” she says. “They need to be given the freedom to implement what the data is telling them based on the user behavior.”

Molinari says she’s seen merchandisers buried in so much data that they don’t know where to turn first. But new tools that cut through the clutter and provide instant insights are giving merchandisers the opportunity to take a larger and more collaborative role in e-commerce organizations.

Merchandisers are now able to quickly understand how the categories and products they’re responsible for are performing. More importantly, such cutting-edge merchandising tools illustrate why a product is doing well or falling short. They let merchandisers see their customers’ shopping journeys — how they arrive at a page, where their next stop is when they leave, what products they view during the same shopping session.

It’s the kind of information that puts merchandisers in an ideal position to break through the silos that arise in any organization. It gives them the broad view that is essential for building the kind of comprehensive customer experience that many retailers seem to be searching for.

Finding ways to break through silos of data is about more than making merchandisers’ and marketers’ lives better. Opening up organizations also means customers can have a better shopping experience.

And isn’t the customer what retail is all about? It’s certainly what everyone says it’s about.

Phil Stocker, senior manager of site merchandising for home-goods seller Wayfair, which recently adopted a new set of merchandising tools, says his team is well on its way to understanding customers in a whole new way.

“Long term, I see this tool getting us closer to where we are the voice of the customer internally, where other parts of the company, all the way from acquisition to retention, is leaning on us for insight on who are customer is, how they shop and what’s important for them.”

(Full disclosure: Wayfair is a BloomReach Compass Merchandising customer.)

The potential shift is not insignificant. Forrester recently surveyed 100 digital merchandisers for BloomReach and found that, in general, merchandisers don’t believe they have the access to the data and tools that they need to make the best decisions for their companies.

In fact, Forrester’s research also indicated that the lack of data hurt merchandisers’ standing with their marketing colleagues. Only 36 percent of the marketers that Forrester polled said they strongly agreed that they should integrate merchandisers’ product knowledge with marketers’ own customer expertise. The marketing side simply didn’t believe that merchandising had the data needed to help attract new customers or keep existing ones.

Talk about a silo.


Breaking those silos apart takes more than technology and tools, Molinari says. Better integrating different retail functions has been a focus of her career and something that she works on with her retail clients today.

“The best thing I can tell you,” she says, “is break down those cubicles and break down those office walls and just go sit with everyone.”

Learn from the other teams. Keep each other apprised of goals and initiatives. If marketing is launching a campaign focused on a landing page, hold a meeting of marketers and merchandisers before the campaign even begins. Coordinate on the relevant products.

If the campaign will feature jean shorts, for instance, merchandising might have ideas about an onsite feature about the fashion or maybe a notion about a spring-break tie-in with the shorts or a plan to build a lookbook.

“One without the other is not good enough any more,” Molinari says about merchandising and marketing. “It really has to come down to the mentality that you need to collaborate.”

Merchandising and marketing need to examine and discuss the results of their efforts together, too, she says. How many product page views were there? Did conversions increase? Did the campaign help sell similar items?

But to be a part of the bigger discussion, merchandisers need to have ready access to meaningful data. And there, Forrester’s research shows, organizations have a lot of work to do.

analytics dashboard

Only 44 and 42 percent respectively of the 100 merchandisers Forrester surveyed strongly believed they had adequate tools to gain actionable insights into shopping behavior and inventory management. Only 35 percent strongly believed they had what they needed to analyze their customers’ journeys. The numbers were even worse for revenue management, merchandising analytics, pricing adjustment, purchasing management and assortment planning.

But the right tools can make all the difference.

“What helps, as a manager, is to see the team get really excited about having a really easy way to measure every decision they’re making about the site,” Stocker says.

And why not get excited? Armed with easily accessible data, merchandisers find themselves in a position to provide the marketing team with the insights it craves. They can help buyers understand what is selling from what category pages and what is not.

Having the right real-time data helps build an atmosphere in which various teams can start to see the bigger picture — the view beyond their individual silos.

“There are so many different departments. There are so many agendas,” Molinari says. “It really comes down to, is this going to improve the business. If you win, we all win.”

It’s a different way of thinking. But clearly, with the accelerating pace of change in retail, thinking differently is called for more than ever before.

Photo of silos by Counse  and data dashboard by Intel Free Press published under Creative Commons license. Photo of Molinari courtesy of Caitlin Molinari

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.


Are you learning from your customers who don’t convert?

So, how’s this for an idea: Maybe your best customers are the customers who don’t buy from you.

Wait. Wait. Don’t laugh so hard that coffee shoots out your nose. I’ve thought this out, really. (Better yet, I talked to some people who actually know what they’re talking about.)

First, I’m not saying your best customers are the customers who never buy from you. Those would be your worst customers. (You’re welcome.) What I mean is, say you run an e-commerce operation. What’s your conversion rate? Two percent? Maybe three? Do you ignore the 98 percent of visitors who don’t convert?


Of course not. Those 98 percent (for argument’s sake) are a gold mine of information, a virtual focus group, telling you things about themselves and your sites that you’d never figure out without them.

“You can follow user paths and look at analytics and see where customers are, where they’re engaging, how they’re engaging, how long they’re on your site, how many pages they visit, where they tend to leave,” says Brian Beck, senior vice president of strategy at retail consultancy Guidance. “Studying those things closely, can lead to actionable steps to improve engagement with them, which ultimately can lead to better conversion.”

And the key to building a better conversion rate in the long-term? A better customer experience, says Jay McCarthy, of UK-based Qubit, which uses data to build customer experiences for online enterprises.

“The idea, over time, is we learn about your affinities, your preferences, your behavior and then we can tailor the experience to match this data that we’ve collected over time about you. So they’ll be more relevant to you.” McCarthy, Qubit’s vice president of product marketing, says. “A lot of these customers that you’re bringing on (to your site) don’t convert immediately. They are nurtured over time and then they become, hopefully, loyal customers who have lifetime customer satisfaction and lifetime value to you.”

The truth is that the non-converters — the 98 percent who visit a site, but don’t convert — are simply not converting on that particular visit. Some of them never intended to convert on that particular visit.

Maybe they’re keeping up with the latest trends so they’re ready to buy when, well, they’re ready to buy. Maybe they like your brand or a brand that you carry and they’re checking in. Maybe they’re comparing prices or seeking inspiration for a piece of furniture or a special outfit or something fun to wear on a vacation coming up in a couple of months. (Consider the recent Wall Street Journal story about the amount of time some women research purses before buying, which prompted this radio interview with the story’s author.) Or maybe they’re shopping online and buying in a store.

None of which is to say that conversions are not important. That’s what it’s all about. It’s what retail executives and managers measure themselves and subordinates on. It’s how people make bonuses. It’s the R in ROI. And certainly what a customer actually does buy and how he or she came about buying it is valuable information.

But McCarthy and other experts urge retailers to take the long view.

You spend a lot of money trying to get people to come to your site, or to one of your marketing campaigns, and yet, the vast majority, maybe 90 percent of your money is spent on acquisition and only 10 percent on retention and lifetime-value-kind-of-work,” he says.

Beck says that, of course, there is nothing wrong with paying close attention to conversions, but in the current digital world, it’s not enough.

“I would say online conversion, as a singular metric, is useful, but if you are anyone except a pure-play online retailer, you have to think differently about conversion,” he says. Conversion is “part of it, but it’s got to be about your overall conversion, particularly if you’ve got 300, 400, 500 stores.”

In fact, Rick Kenney, Demandware’s head of consumer insights, has gone so far as to suggest that conversion rate be scrapped in favor of an orders per shopper metric.

Perhaps not surprisingly, the mobile revolution has fed into this need for retailers to be more thoughtful, more nuanced, when they measure success. Mobile has changed everything. Today, a consumer might visit a site four or five times in a day, while researching a product, never intending to buy on his or her phone.

And then, boom, he or she buys on the laptop at home. Or, boom, he or she shows up in your store to buy the product they’ve been browsing on the small screen.

The trick in all this for a digital retailer is not only knowing that not every visitor is looking to convert, but also knowing what each individual is after on any given visit and then offering the content that meets that need.

It’s a matter of creating human-aided, data-driven systems that know when to send an email, when to offer a promotion, what particular products to show, which products to group together on a page, what product recommendations or related searches to show to each individual visitor.

“That’s the holy grail that everyone talks about — the whole personalization thing,” Beck says. “It’s about addressing the customer need at each touchpoint at the right time. And that’s hard to do.”

Many retailers have a ways to go in learning from their non-converters, Beck and McCarthy agree. But both say there is good news for retailers, too, as big data technology continues its gallop into the future.

“For retailers, web merchants, there are tools that are becoming more and more available, commonly,” Beck says. “It’s becoming democratized in some ways. It’s becoming more accessible from a cost standpoint.”

For instance, Beck mentioned BloomReach’s technology, which helps site merchandisers understand how customers discover certain products and shows whether merchandisers have properly displayed, described and presented those products.

Water Shoes

Say, for instance, a retailer saw a significant number of shoppers who searched for “water shoes,” but did not convert on a page that showed them all kinds of boots and footwear designed to keep feet dry in wet weather.

And say a significant subset of those frustrated searchers stuck with it and ended up buying or adding to their carts waterproof sandals and outdoor shoes that were made to wear into water to protect a swimmer or wader’s feet from rocks and such. A retailer would know to expand its website’s thesaurus so that “water shoes” would be associated with “sports sandals” and “river sandals” and more shoppers would find what they wanted.

More retailers are concluding that those tools, like Compass Merchandising, can help them leverage the information they receive from visitors who don’t initially convert — leverage it to ensure that they eventually do.

You can create lifetime value to your company that is five-fold or 10-fold what maybe that fleeting kind of transaction might be,” McCarthy says.

Which stands to reason, given that those non-converters are your best customers. You may now resume drinking coffee.

Photo of window shoppers by Audio-Technica published under Creative Commons license. Photo of water shoes by Mike Cassidy.

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.



E-commerce answers are the main course at Scott Silverman dinners

Think of a Silverman dinner as an advanced course on retail, complete with fancy appetizers, impressive entrees, paired wine, a few laughs, some civil disagreements and a better than fair chance of ending the night with a full stomach and a full brain.

At the center of the events is Scott Silverman, something of a retail whisperer, who hosts as many as three dozen dinners a year, on average, that draw generally no more than a dozen e-commerce executives from some of the country’s best-known retail companies.

Scott Silverman

“It’s an industry that moves really fast. It’s easy to make mistakes,” Silverman says of retail, which he’s been watching since his decade as executive director of Shop.org, starting in the late 1990s. Shop.org is the digital division of the National Retail Federation (NRF). “My experience at Shop.org was a foundation for the value of people getting together and also seeing that in this industry, in particular, there is a higher level of willingness to speak and to share with each other — with the knowledge that people are going to share back with you.”

Also at the table is a retail technology vendor, usually a software-as-a-service provider, that pays for dinner and Silverman’s networking services and joins the conversation. But these dinners are not sales sessions. They maintain a salon vibe —  a room of experts learning from each other.

Howard Blumenthal, who oversees e-commerce growth and omnichannel experience at Advance Auto Parts, says the Silverman dinners are a chance to supplement the knowledge that a retailer can pick up at the massive national trade shows aimed at the industry — but on a much smaller, more intimate scale.

“These are local essentially,” says Blumenthal, who works in San Francisco and attended a March dinner at the Absinthe Brasserie & Bar in that city. “It’s a low-key event that allows people who live near each other to actually mingle and talk about topics with their peers.”

Indeed, Silverman’s dinner circuit recently has included San Francisco, Los Angeles, New York, Boston, Minneapolis, Columbus, Seattle and Dallas. The get-togethers have given him a unique vantage point for understanding the challenges that internet retailers face and the opportunities they are chasing.

Silverman leads the conversation, starting out with an ice-breaker and then moving into a few key topics that retailers are focused on. And the retailers talk, generally warming up to each other fairly quickly and showing a willingness to dispense wisdom even among those who might be competing for consumer’s finite retail dollars.

So, what are the big topics?

“It’s evolved over time as our industry faces new and different challenges,” Silverman says of the dinner conversation. “Today, interest in mobile is very high as the percentage of customers that are interacting with retailers on a mobile device keeps increasing.”

Which Silverman says is ratcheting up the pressure on retailers, because it’s expensive to build a mobile experience, while also building a desktop experience. Omnichannel, or linking together different shopping and sales channels, is a big issue for store-based retailers, he says.

“One topic that has an impact on everybody is customer acquisition —  and also competing with Amazon.”

Talk about an appetite suppressant. Silverman says that Amazon could indeed pose an existential threat, depending on the retailer.

“It’s a really big question,” Silverman says regarding survival in the age of Amazon. “They have so much scale.”

He said there are brands that sell directly — Lululemon and Under Armour, for example. They will keep selling to consumers, whether Amazon also sells their products or not, Silverman says.

“The retailers that are selling the same products as Amazon is where it’s hardest,” he says. “And there you’re seeing them find ways to differentiate themselves by using their deep product expertise.”

Think of Williams-Sonoma, he says, and the sort of expertise a store salesperson can bring to the subject of cookware or kitchen gadgets. At that point, it comes down to how loyal a customer is once he or she has taken advantage of the in-store knowledge.

“OK, they get all their information from the specialist and then they go, ‘Well, here is Amazon. And they can get it to me in one day and they’re not going to charge me for shipping. And their prices are about the same. It’s so easy. So how much market share can you prevent from going to Amazon?”

Gourmet tables

Silverman is passionate about retail. It’s not one thing, but it starts with the people.

“I always found it to be an industry where there were more, kind of, down-to-earth people,” he says.

And then there is the strategic thinking required in a world where a retailer might have millions of customers and one of his or her goals is to develop a relationship with those customers. And there is the growing role of the technology that’s involved and the way it’s evolved.

Yes, Silverman has had his chance to put his mark on the industry that he loves. When he was executive director of Shop.org in 2005, he and Ellen Davis, a public relations person for the National Retail Federation, came up with the term “Cyber Monday.”

The label was the online corollary to “Black Friday,” the day after Thanksgiving and the biggest brick-and-mortar shopping day of the year. Cyber Monday falls on the Monday after Thanksgiving and ranks among the biggest holiday shopping days online.

“This is one of my favorite things to talk about,” Silverman says.

He explained that Shop.org did a lot of research about holiday spending, both to provide insights to members and to get the attention of media outlets, which love producing stories about holiday shopping trends.

“We had been doing this study for a number of years leading up to the holiday season,” he says, “and we saw data showing a big bump in online shopping the Monday after Thanksgiving. At the time, one of the big reasons for it was that the high-speed internet connections were at people’s offices and not necessarily in their homes.”  

The phrase led to an online deals site, that led to a revenue stream, but not one that the Shop.org team wanted to keep for themselves. And so, they established a scholarship meant to honor a late colleague while providing financial help for students studying in fields associated with digital retail.

The idea for Silverman dinners also grew out of ideas building on ideas. The gatherings, which Silverman calls “E-commerce Leaders Dinners,” were inspired by another NRF colleague, Cathy Hotka, who brings together retail CIOs for similar dinners, but with different topics and retail titles.

At Shop.org, Silverman managed dinners for its members that invited 40 to 50 e-commerce professionals. The events involved four or five tables of eight, or so, each — each table engaged in its own conversation. Silverman wondered how it would work if he better curated the crowd, going for a smaller group of general manager and executive-level professionals, all sitting at one table.

The dinners were born.

“I think Scott really relishes that role in trying to bring conversations together and connect people,” Blumenthal says.

And why not? Good food. Lively conversation. And the chance to help people in a dynamic industry help each other succeed.

Photo of Scott Silverman by BloomReach, restaurant by PortoBay Hotels & Resorts published under Creative Commons License.

Mike Cassidy is BloomReach storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.

Tinkering with automated product rankings can lead to a really bad day


by Mike Cassidy and Romil Shah

As the retail industry is upended by changing consumer habits and rapidly evolving technology, the life of a digital merchandiser has changed in ways that were practically unimaginable even a decade ago.  

And that’s good news, right?

Of course it is. Think about it: Never before have site merchandisers had the kinds of powerful, data-fueled tools that allow them to apply science to their vast knowledge of the art of picking products, building brands and telegraphing trends.

But with great tools comes great responsibility.

The era of big data and the tools to immediately analyze and act on it has created the need to master the balance of human experience, intuition and inspiration with machines’ ability to analyze information at an incredible scale in order to amplify human expertise.

What are we talking about? Think about real life. When things are going well, it’s human nature to want to double-down on that and make things even better — and merchandisers, of course, are human.

Under Armour shoes

So, say one particular pair of Under Armour shoes abruptly begins to sell at a dizzying rate. The temptation, of course, is to pop the top on your data-driven recommendation engine and tinker with a way to boost more Under Armour products to the top of your page. Hoodies, shorts, backpacks.

After all, the brand is on fire.

Boosting the brand is a natural impulse, but in some cases it’s the wrong one.

The truth is, that when it comes to a commerce website, like the human body, there are symptoms and there is the disease. There are the outward signs and the internal cause. Those Under Armour shoes? The UA Curry Two? They happen to be the shoes worn by Golden State Warriors point guard Stephen Curry, the superstar team leader of an NBA franchise that recently completed an unprecedented winning regular season and fell one game short of a second consecutive NBA championship.

The increased sales have very little to do with brand and everything to do with who the shoe fits. It may not be the sort of backstory that spills over to Under Armour backpacks or hoodies for that matter. So, a heavy handed boost of all Under Armour products may simply annoy a Nike sneakerhead who is looking for a pair of KDs (which your customer data should have told you since he’s been gobbling up every pair of KDs he can for years).

The hypothetical scenario is not only an illustration of the power of celebrity. It is also a cautionary tale of what can happen when digital merchandisers impulsively give in to the temptation to fiddle with machine-learning systems that were carefully crafted to optimize performance.

And boosting brands is just the beginning. There are plenty of other ways the impulse to manually “fix” an e-commerce site can lead to unintended consequences.

Say you sell a watch from a well-known, top brand and therefore you make the decision to boost it to a top spot on the category page. After all, last year’s similar model was a hot seller.

But the surprise is that it is simply not selling. It turns out early customers are not pleased with changes the watchmaker made to the latest iteration of its product. The reviews on the site are brutal. The aggregate customer rating is dismal.

That leads to a situation where the watch is doing well in terms of customer engagement. After all, it’s a new watch from a top brand. There might even have been publicity before its release. It’s possible fans have been waiting for the watch to hit the market. And so, they click. But then they see the reviews and they don’t buy.

You might be tempted to bury the watch and replace it with a product that is converting at a higher rate. More conversions equal more revenue. But rather than manually taking charge, which would mean constantly monitoring and moving the products on that page, would you be better off letting algorithms do more of the work?

After all, your carefully crafted algorithm will note the watch’s poor performance, which will cause it to fall lower in the site’s product rankings.

That is not to say that overruling an algorithm is never called for. It is just to say that such changes should be carefully deliberated and considered in the broader context of the whole site and your overall goals.

Site merchandisers today, for instance, rightfully see themselves as responsible for managing the digital experience — the entire experience. They know that the way in which consumers experience their brands can be as important, or even more important, than increasing the sales of a product.

As one who manages the digital experience, you pride yourself on creating a feel, an emotion, an aesthetic. It’s what attracts new customers and keeps other customers coming back. And it’s not something algorithms are good at. Consider a recommendation engine: All the data in the world might tell you that a significant percentage of men, say 15 percent, who buy a particular designer-label shirt also tend to buy a plush bathrobe.

bathrobe brigade

But to say to a customer (through your digital site), “Hey, you just bought that nice shirt, you might also want to buy this comfy bathrobe,” is simply a discombobulating experience. There is no art there. That’s the time when a skilled merchandiser might want to overrule an algorithm to offer a high-performing sport coat or tie as a complement to the shirt.

The key is to know when to make a move, so that the best-laid plan doesn’t backfire. It starts with taking a holistic view and appreciating the relationship between human and machine.

As a site merchandiser, you and your team decide how to optimize your site. You consider goals and come up with your priorities. Maybe it’s conversion or user engagement. Whatever your priorities, you design your algorithm to optimize for them.

Every change you want to make should correlate with those underlying goals. Left to its own devices, a well-crafted algorithm will learn and consistently work toward achieving your priorities. If you tinker with it in pursuit of some goal beyond your core priorities, you will undoubtedly create unintended consequences and your automated system will suddenly be working at cross-purposes.

In the end, it is counterproductive to isolate one data point or one search query and, based on that, make changes to the work that algorithms are carrying out automatically. A retailer might have more than a million products on its site. Shoppers searching for items have a nearly boundless variety of ways of describing them. How many queries can you reasonably improve by tinkering with the algorithm?

Again, that’s not to say that merchandisers and site optimizers don’t need tools and visibility into the algorithms that drive their site. They do! But those tools should be designed to optimize or strategically tweak a product or category. Those changes should drive specific outcomes in the customer experience with no unintended consequences. Unless those product and site experts both fully understand the mathematics in an algorithm and have the ability to test their algorithm tweaks, the risk is just too great.

And speaking of testing, whether your algorithms are built in-house or by a trusted vendor, testing any changes to that algorithm is critical. Your quants or the vendors likely spend a great deal of time measuring the impact of every tweak they make (or at least, they damn well better be!). You can’t tweak and test properly unless you have the right data, testing platform and statistically significant sample to pull it off with confidence. If you don’t have those things, yet still want to fiddle with weights and goals, it’s a “be careful what you wish for” situation.

The impulse to jump in and make a change based on insufficient data points is understandable, but dangerous. It’s the kind of thing that leads to the sort of logic at the core of that old joke:  My company is losing 10 cents on every widget we sell. Don’t worry, though, we’ll make it up on volume.

When it comes to tinkering with with your automated ranking weights, don’t be one of those trying to make it up on volume.

Photo of Under Armour shoes by Sandy Dover Creative and men in bathrobes by Paul Downey published under Creative Commons license.

Mike Cassidy is BloomReach’s storyteller. Romil Shah is an engineering manager at BloomReach.


Video: How to future proof your e-commerce operation

It’s not news that successful retailers can no longer keep separate systems for online data and in-store data.  But, says Sean Cook, formerly of Aptos, just how to combine that data successfully is easier said then done.

In our latest installment of BloomReach University, Video Campus, Cook talks about turning to the cloud to “future proof” your retail business; to be ready for the next big thing in retail data — even if nobody knows exactly what that’ s going to be.

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.

St. Louis Cathedral, New Orleans

BloomReach Relevance Report: Collision 2016 edition

The BRRR just flew back from the Collision tech conference in New Orleans and boy are our arms tired. Hey, it’s the BloomReach Relevance Report. What did you expect in the way of humor?

St. Louis Cathedral in the French Quarter

St. Louis Cathedral in New Orleans’ French Quarter

It was no accident that the Collision folks chose New Orleans for the third Collision. The show goes for eclectic and there are few more eclectic cities in the the United States than New Orleans, with it’s elegant southern charm juxtaposed with its off-the-hook Bourbon Street party scene. This was a conference that had impresarios of the NFL (Brett Favre) and SEO (Rand Fishkin). It included presentations on stage by PJ Morton, keyboardist of Maroon 5 (seen below with Chris Kaskie of Pitchfork) and Brad Smith, CEO of Intuit.

PJ Morton of Maroon 5 talks with Chris Kaskie of Pitchfork

No, it’s not Shop.org or NRF’s Big Show and it isn’t laden with retail-specific panels or discussions. But there was plenty for retailers to learn at Collision and we’ll run some of those down in words and pictures. There was plenty of wisdom dispensed from the marketing stage.

Rand Fishkin of Moz

Rand Fishkin of Moz

Neil Vogel, CEO of About.com, talked about the pioneering internet site’s recent radical reinvention.  The site, a mishmash of advice and answers to questions, launched nearly 20 years ago, when the internet was a very different place. Vogel said that in 2016, the idea of a one-stop site is not nearly as appealing as a strong vertical — for instance a site that is all about health.

“You can’t be everything to everyone,” Vogel said. “You don’t want diabetes advice from the same place you learn to stain floors,” he said. And so he announced that About.com was launching a stand-alone health site, pretty much at that very moment.

Neil Vogel of About.com

Neil Vogel

Of course, he got wild applause from those at packed into the Market Stage area.

“Don’t clap,” he said. “It’s wasting energy. Take out your phones and start clicking on stuff.”

He was joking. Kind of. Not really.

Traffic is king and IAC, About.com’s parent, is hoping this new strategy gets more. It wasn’t an easy move for an outfit that had been doing the same thing, pretty much the same way, for a long time. Vogel described it as, “forcing a huge cultural change, something that has been essentially the same thing for 20 years.”

The lesson for retaliers? Don’t be held back by legacy. The BRRR had a chance to moderate a couple of panels at Collision, including one with Vincent Yang of EverString and Matthew O’Grady of Nielsen Catalina Solutions and another with Dan Wagner of Civis Anayltics.

Among the many things that came up in both discussions about data with the analytics experts is how emotion can get in the way of embracing data. Nobody was suggesting that anyone should simply turn blindly to data, but there is a danger in clinging to the past. It can be hard, emotionally, if an executive perceives that he or she is giving up control of decisions to a machine run by data scientists. That observation applies to retail executives, too, who at times resist the power of machines, because they relied on gut for so long. The answer is to balance the two  — the art and the science.

Panelists at Collision left little doubt that the winners of the of the coming decade will be those who best manage, analyze and act on the right data. Vitaly Gordon, a data scientist with Salesforce, told attendees that data science is important.

More Collision coverage

It’s the electricity of our time,” he said, explaining that soon data science will be so ubiquitous that people won’t give it a second thought. It will just flow — and it will contribute to great things.

“Every time  you create a product like GPS, an intelligent tool that gives advice that is personalized, you make the world a better place,” he said.

Another lesson came not from any stage, but from the overall vibe and mission of Collision. Sure, the organizers are looking to make a profit, but they are also looking to provide a chance for startup entrepreneurs to network with each other, as well as investors, while working to get the attention of a journalist corps that also attends.

Collision speaker mingle before dinner at the Board of Trade

Collision speaker mingle before dinner at the Board of Trade

There are no doubt somber, sober even, networking encounters, but the organizers seem intent on lubricating the socializing with nightly pub crawls and dinners featuring open bars. The gatherings and meet-ups are a reminder that there is wisdom in the crowd. Like leaders in any industry, retail executives can be hesitant to share ideas around solutions to common problems.

A relatively calm Bourbon Street before dark

A relatively calm Bourbon Street before dark

But with Amazon’s dominance it is time to share best practices without giving away the secret sauce that one retailer might believe provides an edge over the competition.

Startups symbolize the future, which came come at you in a blur

Startups symbolize the future, which came come at you in a blur

And maybe the most significant lesson could be seen in the more than 600 startups that attended Collision as exhibitors. The sheer number was a symbol of the future; a sign that change is coming, that those who will come out on the other side are often those who are ready to take risks and try something new.

Canal Street streetcar

Streetcars move down the Canal Street Line.

The past can be comfortable, easy, beautiful even. And sometimes the old ways will suffice. But it is much more likely that something you don’t even see coming will threaten your business. Unless, of course, you’re the one who nobody seems coming.

Photos by Mike Cassidy

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.

+Aziz, a marketer/musician born for Collision

The Collision tech conference was made for a guy like Aziz Ali.

Aziz is a marketer, who devises strategies around culture. And he’s a musician known as +Aziz, who plays guitar. He is a man who appreciates both the joy of creating art and the discipline of measuring just how his music is resonating.


“It’s actually very different brains,” he says of his musician side and his strategist side. “You kind of need to be your own manager and that’s when you’re sinking into the world of finance and analytics and where you need to be. And then you have this other portion of you that is the side of you that just wants to play.”

It’s only a bit of a stretch to say Ali personifies Collision, the upstart conference with the hip vibe, that expects to draw 11,000 to New Orleans this week. The conference is focused on technology, but more, with parallel focuses on music, sports and marketing. As it did in Las Vegas its first two years, it works to connect with the community in which it’s held.

And here’s Ali, a proud New Orleans resident, who plays music and tracks metrics and knows a thing or two about data-driven marketing: A Collision trifecta. Naturally, he’s scheduled to talk about the band as a brand at the conference on Thursday.

More Collision coverage

Ali lives the tension that Collision embraces  — the balance between art and science, gut and data, when it comes to running an enterprise. In Ali’s case, he’s an indie rock/world music guitarist, who’s launching a band called Kuwaisiana, a blend of Kuwait and Louisiana.

Some musicians he talks to aren’t crazy about data driving art, Ali says, and he worries himself that the fixation on metrics could detract from the creative process.

“Not a lot of musicians that I talk to are comfortable with that kind of stuff, when I start to talk to them about KPIs (key performance indicators) and measuring and having defined audiences,” he says.

But Ali says he’s decided it’s a necessary distraction in today’s music world. And so, he turns to Google Analytics and services like Reverb. At least once a week, he looks in on his social media platforms, which include SoundCloud, Twitter, Facebook and Instagram.

“Usually what I’m looking for is that there is attention,” he said. “For me it’s about awareness and exposure to more eyeballs. I’m usually focused on expanding to new potential communities.”

Ali has identified his core fans and potential fans as those interested in Middle Eastern music and indie music, world music enthusiasts and people in the Middle East or of Middle Eastern descent.

“Those audiences haven’t embraced me to the same degree that they’ve embraced other people that they’re fanatical about,” he said of the last group.

He’s hoping to appeal to “music snobs” who might write a blog post about his music and expand his market. In the meantime, he’ll keep tracking key metrics.

And he’ll continue to increase his visibility with novel projects, like his mashup with a computer programmer who translated hashtags at one of his concerts into bits of music and his partnership with Lululemon that brought his music to yoga enthusiasts.

And, of course, there are the speaking engagements, like the one at Collision, which is right where he belongs.

Photo by Rushing Haise, courtesy of Aziz Ali.

Mike Cassidy is BloomReach’s storyteller. Contact him at mike.cassidy@bloomreach.com; follow him on Twitter at @mikecassidy.