It was a bone-chillingly cold Friday afternoon in February when I first met Melissa Middleton and Fritz Lanman, founders of what may perhaps be the future of the fashion magazine. Their startup, called JNSQ, is the first digital style magazine that runs on the iPad as a native application. The online (desktop) site for the magazine is JENESEQUA. For those who wonder how the magazine format can thrive in a digital era, JNSQ may provide that solution as a highly integrated digital experience that provides premium content enhanced through interactive features only possible within a mobile platform. 

Melissa and I had been chatting the week before through phone, but Fashion Week provided the ideal time for us to meet in person so I could gain a better understanding of her startup.  From a quick visit to Pamela Love’s Fall/Winter 2012 presentation at Milk Studios, where Native American tribal motifs met the iconic steampunk ambience of Miyazaki’s films, to Sarabeth’s of Chelsea Market where Melissa and Fritz were waiting, fashion week is seldom kind to schedules.

Much like any community, there are strong performative aspects to both the tech and fashion world. They both have informally enforced dress-codes that stand at strong contrast to each other. The fashion uniform of mono-chromatic tones, sleek cuts, and modern silhouettes would probably have a heart-attack if it had to don the stereotypical tech garb of free printed tee’s filled with corporate logos or witty one-liners in bright RGB color-scale hues. And let’s not even get started on the ill-fitted jeans and sneakers. So at the confluence of both tech and fashion, it stands to wonder what one should wear at a startup like JNSQ. Needless to say, this kind of question has little room in a serious consideration of what JNSQ has to offer, which is definitely a lot.

Why Mobile & Fashion?

Source: FastCompany

Mobile applications for smartphones and tablets are a highly underserved market in the fashion industry considering how fast mobile device penetration and usage is growing. For your consideration:

• Mobile is one of the fastest growing sub-sectors of the digital era, with 42% of consumers already owning smartphones and 84% of those users actively using their phone while shopping in-store (Deloitte’s 2011 Holiday Survey)
• Smart phone sales are projected to exceed global PC sales by 2014 (The Mobile World Congress, 2011)
• 27% of the mobile phone population in the United States has a household income that’s $100,000 or more This is an audience of 64 million Americans (L2, 2011)
• 91% of mobile users use their device to socialize in some manner compared to 79% for desktop users (Digital Buzz, 2011)

With tech titans Google, Microsoft, and Apple increasingly focusing their efforts on mobile devices, it should be clear to anyone that mobile represents a huge opportunity (or threat). The mobile app represents an exciting new frontier in the digitalization of fashion, but fashion brands have proven slow and ineffective at capturing this space. Only 20% of brands with an app offer differentiated content on it.

In the tech world, this kind of industry stagnation is a signifier of a lack of innovation or an inability to see the opportunity therein – which is when start ups come to disrupt the industry with new products and services. Now, it’s happening in fashion too, with tech-oriented start ups promising exciting new possibilities in the brave new world of fashion digitalization.  And without a doubt, JNSQ is part of this new cohort of fashion/tech startups.

Why Current Fashion Magazine Apps Fail

What JNSQ offers is more than just an editorial point-of-view and great content, they can potentially create a best-in-class platform for magazines to effectively digitize their content. Just think about how coveted JNSQ would be to Condé Nast or Hearst if it could crack the digital magazine puzzle, or how deadly it could become to the business models of the incumbents? Most fashion magazine apps simply re-hash the same content from their main site and print magazine. The Cut, Marie Claire, and Elle are trying to penetrate this market, but aren’t reaching their full potential for several reasons.

Screenshots of JNSQ

First, the underlying technology isn’t strong enough. My guess is that most fashion magazine applications are not being created in-house* but rather with agencies or white-labelled solutions, which in my experience never creates a stellar product (adequate is never good enough when it comes to innovation). For example, The Cut for the iPad constantly crashes, and content is pulled incredibly slow at times. The UI is heavy and the UX is not optimized for browsing on the tablet. Reading in a print magazine is largely a linear task, we turn from page to page. But in the digital world content, it is multi-directional and dynamic. We open multiple tabs, bookmark pages, jump from link to link and find content from friend’s recommendations. Of course, it’s understandable that agencies or white-labeled solutions are being used to develop these applications – it’s incredibly resource-intensive to develop an application. And magazines just don’t have the latent development capability to effectively develop novel platforms for content.

*[Note: My friend at Condé Nast informed me after publishing the article via Twitter that magazine apps at Condé Nast are done in-house using white-labeled platforms and not through agencies: "Condé Nast uses the Adobe platform. CS5 has a lot of interactivity built in already. We add in HTML + Javascript too."]

This is where a smaller start-up like JNSQ, with a strong tech background and knack for beautiful UI/UX design can create a platform for mobile fashion magazines.

Second, current fashion magazine apps do not take full advantage of the interactive capabilities of tablet devices. Just look at the most creative and successful type of mobile applications and you’ll understand where there’s work to be done: mobile games. Angry Birds and other mobile games take advantage of the iPad’s accelerometer, multi-touch enabled screen, and other functions to create truly interactive and immersive experiences. The fashion magazine app must also follow suit and take it a step further. Integrate with the API’s of other platforms like YouTube and create new ways to engage with users.

Here’s an idea: what if I could literally drag items I liked in Marie Claire’s iPad edition with my finger to a special cart. It doesn’t matter if it’s from an editorial spread or an advertisement, all items would be able to be saved into the cart. At the end, I could then buy everything through the application. All Marie Claire has to do is tie into the ShopStyle API (which is what startup Inporia is currently doing) to get product feed information without building an e-commerce platform from scratch. Marie Claire can take a cut of the sale from ShopStyle or form direct relationships with the brands themselves. Think of all the new revenue streams it could create.

JNSQ is taking advantage of the iPad’s capabilities by doing just this. Integration with YouTube for seamless video watching or being able to swipe your fingers on the screen to switch between pages are just a few of the interactive features in their app. As the app continues to develop, I am expecting an even stronger platform with interactivity that engages and excites users.

Third, there is a lack of integration with social media platforms to drive viral growth and foster a sense of community. Some of the greatest tech startups are gaining viral traction because of their smart integration with social media platforms. Most fashion brands think adding a social media toolbar to the end of their content will be enough to encourage users to share and like content back to their social media profile, but that’s not enough. Facebook apps like Washington Post Reader and fashion/tech startups like StyleMusée are taking advantage of social data to encourage friends of current users to download the app and be part of the community.

A Certain Je Ne Sais Quoi About Fashion & Mobile

While JNSQ was publicly released in January 2012, the app has been available to download since October 2011. They have been quietly building up a dedicated reader base as a stealth-startup, a move that has enabled them to build up their application over time so that its underlying value prop is solid.

Great concept aside, don’t think of JNSQ as being an underdog in the fashion community – the founders can hold their own weight. Melissa comes from a background in beauty (Beauty.com, Clarisonic) and telecomm (T-Mobile) where her experience in both fields allow her to draw upon different insights to drive a strong vision. Among other roles, which includes being the former Senior Director in Corporate Strategy and Acquisitions at Microsoft, Frtiz is an angel investor with an impressive portfolio – does Square and Pinterest sound familiar? The two already have an editorial team that’s pumping out content for their magazine. Of course, the devil lies in the technology that powers the app. They have it covered as well with a seasoned developer with an equally impressive background: an Ex-Amazon engineer with a PhD in Comp Sci from Princeton. (Seriously, can someone make a fashion/tech dating website? I would sign up.)

So what does JNSQ stand for? In playful reference to the phonetics of  JNSQ’s website, “Je Ne Sais Quoi”  is a French phrase that literally means “I don’t know what.” It refers to a hard-to-describe quality that makes something distinctive or attractive. If you were to ask what the exact components of a successful fashion magazine application were to be, no one would have a comprehensive list. This is the challenge tasked with JNSQ as they brave the startup world to create a digital fashion magazine that runs natively on mobile devices. But already, they have an app with a certain flair, a certain charm, a certain je ne sais quoi.

You can follow JNSQ on Twitter or Facebook. To download the app for your iPad, click here. 

Even if you missed the livestream, the show will be replayed here.

You finally got this far, and you’re ready to make your idea a reality. But, you want to make sure that this idea will improve your site, not inadvertently do more harm than good. The last thing you want is expect one thing is to happen only to find that the exact opposite result occurs!

In Part 4, we will discuss how to test your ideas to make sure they work.

Validating An Idea

Prabal Gurung SS 2012 (Source: Style.com)

No matter how much work and analysis goes into driving insights and turning them into actionable plans, even the best idea can go awry. That’s why testing is so important. Testing shouldn’t have to be a difficult process, but should be vigorous enough so that you can validate the performance increase of your idea.

For example, if you believe that posting more on weekdays around 10AM will drive more traffic, then you can test this hypothesis with some sample posts during that time frame and compare performance with posts that are posted during a later time. Here are some tips to help you test an idea:

1. Set a benchmark to compare to. You will need to see if implementing your idea improves your site’s performance in some way. To do that, you will need to track your site’s performance before you implement your idea (the baseline) to after you execute your plan.
2. Set a time frame, such as a week or a month depending on the nature of your test, and then do a before and after comparison. The ideal time to test completely depends on your site and the type of action you are taking. For example, something as simple as posting more about a certain topic or posting at a certain time is a more immediate change and results can be visible within days. However, something like changing how you post titles, do SEO optimization, or tag items to make your site appear higher on search results can take a month or longer to have a noticeable impact.
3. Don’t forget that something else might have happened. Even if you get a negative result, don’t worry. This is why you’re doing a test! If it’s not working out, you can go back to the way you used to do things. But don’t be too hasty in dismissing an idea. Be sure to make sure other things aren’t impacting your stats. Maybe February is a great month for your blog because of Fashion Week but March is a lull month for you. Seasonal changes in traffic can easily mess up your test, so be sure to take that into account!

These are only the basics; brands and technically-inclined bloggers should definitely read more detailed guides on site testing because there are so many amazing ways to do tests. For example, crawl tests allow you to see how users are actually using the site down to the mouse level. AB testing lets you actually create a control group and test group, that way they can have one group of users on the old site and another set on the new site to do a much better comparison of performance. This guide won’t cover any of these techniques, but I’d recommend you at least start with articles like:

• Online Journalism Review’s Guide to Usability Testing (link)
• Tech Radar’s Gurilla Site Testing Guide (link)
• Kiss Metric’s Intro to A/B Testing (link)

In the third part of the ”Analytics for Brands And Bloggers 101” series, we’ll discuss how to take your insights and learn how to drive change with it. You’ve connected the dots and have found a trend that’s of interest. But how can you take that information and turn it into an action? How do insights actually result in change?

---

From Insight to Action

---

Source: Rad Hourani

You’ve found something interesting and you’re asking – “now what?”. Good point.

Let’s take an example: you notice that your posts with backstage photos during Fashion Week have higher-than-average engagement metrics – users spend more time on those posts, they click on more links, they come directly to those posts from referral links. You know that backstage photos do better than other posts because the metrics tell you so. But this is an abstract insight. What can you use with this information to take a concrete action?

One course of action is that you increase the amount of backstage posts you do if readers like it more. This is an example of a positive action to grow your blog by emulating successful elements of your site. 

In another example you might notice that readers tend to visit your site on Wednesdays from 9AM to 10PM with readership tapering off during the week until Saturday when traffic drops off from Saturday to Tuesday. You might then conclude you should not post as much on weekends because readers tend to not visit your site during that time. This is an example of a negative action to grow your blog by divesting from unsuccessful elements of your site.

But not all insights lead to such obvious actions. For example, you might notice that you typically have a ratio of 40% returning visitors to 60% new visitors for your site. And over time, you are increasing that ratio so that it’s now 50% returning to 50% new. Now, let’s say you also note that returning visitors also tend to view several more pages than new users but that returning visitors tend to only visit once a week. New users, while viewing less pages, tend to also access your site from Tumblr and Twitter whereas returning visitors come directly from their bookmarks. Now the situation is more complicated. While returning visitors might visit more pages, they come back less. New visitors tend to access your site from social media sites, where they might be more willing to share your link with other new users. Furthermore, as you gain more returning users, this might mean that while you’re doing a good job at getting a loyal audience, it might also mean you’re not attracting as many new readers to your site.

This example is the kind of scenario you are more likely to encounter. In these situations where actions have both positives and negatives, you must be careful to weigh them and choose the course of action that best suits your needs.

Weighing Options

When you are confronted with multiple options, you need to be able to weigh your options and choose the best one. Here are some key tips on evaluating your options:

1. Pick an evaluation criterion: You can’t compare apples to oranges. You cannot compare Option A, which would increase number of visitors to your site, to Option B, which would increase average amount of time spent on your site, because they are different types of metrics. Is getting 1,000 more visitors better than having all visitors spend an additional 30 seconds on your site? One way to get around it is to create your own conversion criterion – based on your own personal preferences. Maybe you think 100 visitors is worth an average time on site increase by 10 seconds. Now you can compare the two options. Another option is to focus on only one evaluation criterion (ie – time on site, visitors, number of hits from a certain region) and then only consider options that would directly impact that criterion.
2. Focus on evaluating outcomes, not single options: Maybe Option A and Option B can both be done because they are not mutually exclusive. When you’re evaluating options, group options into bundles. For example, you might have 10 actions you could take, and while Option A might yield the best result as a single action, doing it precludes you from doing anything else. On the other hand, while Options B, C, and E might be individually lower impact than Option A, they don’t preclude each other and combined their effect is greater than just Option A.
3. Take into account ease of implementation. You also want to make sure that it’s feasible to implement your ideas in a timely manner. Don’t forget that what sounds good in theory is not easy to do in real life!

In the Part 4 (Testing), we will discuss how to test your ideas to make sure they will really make a positive impact on your site.

In the second part of the “Analytics for Brands And Bloggers 101” series, we’ll talk about how to make insights from data from a hypothesis driven mindset. Most guides on analytics focus on teaching you statistical technicalities – that’s both cumbersome and ineffective.  Instead of telling you about all the different kinds of metrics you can use, this guide will teach you how to identify situations in which using data metrics can be useful and how to make insights that will let you make smart decisions.

The goal of this part is for you to know when to use data analytics and how you can make insights without any rocket science. 

Finding Opportunities

Source: Luo Hao

If you have the time and interest, then by no means do I intend to dissuade you from conducting extensive due diligence of your site’s analytics. But for the digital brand manager and fashion blogger with limited time and other important tasks to attend to, data analytics shouldn’t prevent us from doing our job.

Personally, I recommend a weekly basic diagnostic check at the end of the week or during the weekend (if you’re inclined to work on a Saturday/Sunday). A diagnostic check should only look at top level metrics (see Part 1).

Every month though, you should do a more in-depth look at your site’s performance. To do a deeper check, you should get on your analytics platform (eg: Google Analytics), export your data into a .CSV file (you can look at your analytics platform FAQ if you need help doing this), and use a program like Excel to dig deeper into your data.

Of course, if this is daunting for you, then don’t worry. The monthly check is usually overkill for bloggers, unless you a lot of large traffic or make a significant amount of revenue from advertisements, affiliate marketing, and other sources. However for brands, this detailed check is important and if you don’t have an analyst in your team to help you do a more comprehensive study, then I would also recommend you supplementing this guide with a more technical one to learn how to.

Regardless of how often you analyze your site analytics (but make sure you do it regularly because otherwise you’ll never do it), the key principal for finding opportunities remains the same.

Find noteworthy comparisons between two or more metrics. The theory of relativity doesn’t just apply to physics, it is the key to understanding your site. If I told you that I had 100,000 visits last month, it does not tell you much about my site’s performance. While 100,000 visits might seem great… if I spent $100 million acquiring those users then its a pretty lousy number ($1,000 per user acquired). Similarly, if I had 999,000 visitors last month, then growing only by 1,000 visitors in the course of a month is even more lousy. But, if I typically only gain 10 additional visits a month, then an increase of 1,000 visitors is a lot better.

See? It’s all based on a comparison of a baseline metric. Don’t just look at your stats in a vacuum, compare them to other data points. Common ways to compare data are:

• Time-based (eg: this month versus last month)
• Content-based (eg: content from one category versus another category, like street style posts versus self-style posts)
• User-based (eg: engagement from returning users versus new users, or visits from one geographic area versus another)

If you notice one metric being much higher or lower than another metric, this usually warrants some further inspection. For example, if in one month you gain a lot more traffic than usual (you usually increase traffic by 5% every month but this time you grew by 15%) – you will want to find out why and try to understand what you did right. That’s how you’ll find an opportunity for growth. (If you want to be really statistical, you can even try to find if the difference between your comparisons are statistically significant).

Source: Dream Sequins

So you’ve looked at your data metrics and noticed changes in your site performance along two or more variables. Maybe you realize that returning users from New York view 20 pages per visit whereas the national average for returning users and new users is only 5 pages per visit. Now you need to connect the dots and try to understand the implications of this finding. This is where you make insights. 

Form a hypothesis to explain your discoveries. If New York returning users are viewing more pages than other users, you should try to understand why. Is it because NY readers tend to read more than other readers, is it because returning users engage with your content more, or is it something else? Form hypothesis – they can come from your own intuition, from patterns in your data, from feedback from readers, or any other number of sources.

Validate your hypotheses.If you believe NY returning readers tend to read more because the content is relevant to people in the NY area, then make sure you test this hypothesis by seeing if any recent posts that might have been targeted to other readers result in lower readership from people in NY. This is because your causal mechanism is the relevance of the content, not an underlying behavior difference in your NY readers. You don’t want to accidentally overlook the real underlying reason. For example, it might be the case that NY returning readers tend to only come on once a month and thus read more during their visit because they are catching up on content they haven’t read before. In this case, you need to validate if NY readers tend to visit less often than other users on a monthly basis. As a caveat, you can never really validate a hypothesis, just disprove it. Since you don’t have that kind of time, I would recommend you focus on positive validation over negative refutation, even if you might overlook something. As long as you are close, even if you do not fully understand the underlying reason for a trend, at least you will be on the right track.

Source: COTTDS

Don’t infer reasons for a trend from the data. Remember in Part 1 where I discussed how data tells us what is going on, not how or why. Use data to validate or refute hypotheses (see above). It can be easy to accidentally infer something from the data that’s not there. For example, you might notice that you’re getting more traffic to your site from Google and make the inference that users are therefore increasingly finding your site through search engine traffic. Inference can get the best of us, we see something and assume something that the data isn’t telling us. In this case, all the data tells us if that more traffic is coming from Google (let’s say 10% of your traffic last month was from Search and this month it’s 20%). This doesn’t mean that next month this percent will rise again. This could be a one-time increase, an anomaly, or a temporary trend before reverting back to normal. For example, say you did a series of posts on a topic that drives all of your Google Search traffic because its the first result on the search. But if a year later these links slowly fall under the first five sites linked to on Google, then your search traffic will drop.

Once again, data will tell you what is going on, but it’s up to you to connect the dots.

As a fun, nerdy aside – this is how your brain works too, it’s a giant inference machine really. It takes in information constantly, but can’t actively process most of it so it needs to infer it. This is great usually, but can be exploited. For example, we can only actively process a small percentage of what we actually see (some psychologists say that the brain only processes 10% of what your eye sees, the other 90% is a visual construct from your brain that literally fills in the blanks – see Richard Gregory on Visual Perception). So, a majority of what you “see” is what your brain infers based on relative information.

In the image above, you can see how exactly what I mean by inference. Close your left eye and stare with your right eye at the cross. Then slowly move closer to the computer screen while always shutting your left eye and only staring at the cross. Eventually the black dot will disappear completely. All that’s left is white space. The dot doesn’t actually go away, but your brain infers white space because it’s a blind spot. Cool huh?

In Part 3 (Driving Change), we will talk about how to drive changes on the site from your insights. We’ll focus on how to interpret your insights and how to easily translate them into actions that will allow you to make changes on the site. We’ll discuss the limitations of your insights as well so that you don’t let data control you.

In the Part 4 (Testing), we will discuss how to test your ideas to make sure they will really make a positive impact on your site.

When it comes to running and building a site, behind-the-scenes work is usually approached with a sigh and a groan. Let’s face it, between designing a beautiful layout, posting interesting content, and engaging with users on social media platform, the front-end work on a site is the sexy part of running a site. But of course, without your SQL databases, PHP coding, and CSS styling your site would be no where. In a similar vein, analytics is an oftentimes glossed over but critical component of running a successful site. Both brands and blogs rely on analytics to drive insights and turn them into actionable digital strategy.

But despite how important analytics are, it’s a science that is oftentimes underutilized. Analytics shouldn’t be a dreaded task or  buried deep within your site’s code and then forgotten. Driving actionable insights from analytics data is a key component to brands and bloggers. 

Top Level Metrics

First – let’s look at the top level types of metrics you’ll want to pay attention to. It doesn’t matter what analytics platform you’re using (I use Google Analytics primarily) as long as it tracks most of these data points.

Source: Givenchy

1. Traffic: How are people coming to your site?

• Number of visitors
• Number of views
• Unique vs repeat
• Sources of traffic (How are readers finding your site? Search, referrals, direct linking, etc)
• Device type

2. Audience: Who are the people coming to your site?

• Location
• Language
• Frequency (How often do repeat users come back to site)

3. Content: What are people reading on your site?

• Number of pageviews
• Average time spent on site
• Views by post
• Site Search (What are users searching in your site)
• Bounce rate (Percent of users who leave after just viewing one page on site)

---

How Metrics Can Help You

---

Source: Think PR

Just using the basic metrics provided to you is not enough to drive deep analysis of your site. Being able to manipulate and create additional metrics is critical to being able to understanding site performance. For example, if you were to divide the number of visits over page views, you can get the average number of views per visit, which is a good metric to understanding how engaged your users are. But this still too vague for us.

Now, let’s take that figure and find the average page view per visit for repeat visitors and then the average page view for first time visitors and you  can find the value of converting readers to subscribers is. Let’s say they are the same, this could mean that you should not focus your energy on getting users to subscribe to your blog, because repeat users don’t engage with your content more than first time users. This could be because your content is so good that even a first time user will be engaged right away – this would mean you need to focus on acquiring more unique traffic from new sources. Conversely, if you have a higher page views per visit for repeat readers, then you should focus your energy on creating ways to have subscribers come back to your site and focus more on maintaining your current reader base than expanding it.

This simple example is a powerful reminder of how useful data analysis can be for brands and bloggers. Using only the top-level metrics found on your Google Analytics dashboard, you would only be able to understand to know how many visitors and page views you have in a given time. We then combined the two to get another metric, average page views per visit – but this can only tell us how engaged users are in general. But not all users are created equal. Just like how brands must segment their market to understand how to target different groups of people, we must also become more granular in how we use metrics. By comparing engagement between returning users versus first time users, we can understand where we should focus on energies.

Another example – we can easily find where our readers are located when they view our site. This is a top level metric that is extremely helpful (and it also comes in a nice map of the world on Google Analytics, which eases those of us who dislike numbers and statistics). Suppose we have 100 monthly visitors from New York and only 30 from San Francisco. If we wanted to spend money promoting your blog locally, most of us would say NY is the obvious answer. But think about how much more information we could gain if we combined this information with another metric such as average time spent on site. If your NY readers spend 1 minutes on average on your site, while SF readers spend on average 5 minutes on your site, then the picture changes. While you have more readers from NY, SF readers spent a total of 150 minutes on your site monthly whereas NY readers spend only 100 minutes on your site. It might make sense to focus on acquiring more SF readers if they tend to spend more time on your site.

Metrics can help you make decisions about how to grow your site effectively from marketing to content to site layout. Empirical evidence can be an incredibly powerful tool to help you understand how readers interact with your site, and it can also act as validation or refutation of ideas you may have…

---

What Metrics Can’t Help You Solve

---

Source: Morgan O’Donovan

But metrics is not a golden ticket to success. A successful site, whether as a brand or a blogger is not just a science. It’s also an art. Analytics is often called a decision support tool because it helps us make decisions. But it cannot tell us what to do. The greatest strength of data is also its greatest weakness – it tells you what is happening without bias or an agenda. It does not usually tell you why or how it is happening.

Let’s go back to the example of our 100 NY monthly readers and 30 monthly SF readers. We must take care about making decisions based on the fact that SF readers spent a total of 150 minutes on our site while NY readers only spent 100 minutes on our site in a given month. These figures do not tell us the underlying reason why users in SF spent more time on your site for that month. Maybe in the month you are tracking, you wrote a piece about SF style bloggers and therefore wrote something more relevant to SF readers. But if you don’t continue to write content geared towards SF readers, then you will lose that level of engagement. If in other months your NY readers tend to spend more time on your site, then it might be smarter to focus your marketing dollars on NY-based advertising.

Even if SF readers consistently spend more time on your site than NY readers, new SF users you acquire from advertising might not be as engaged as your current SF reader base, therefore lowering the average amount of time SF readers spend on your site.

Data cannot tell you why users act the way they do, only what they do. This is why we use data as a support tool, but ultimately our understanding of underlying dynamics is key to successfully using analytics to make changes. In order to test our hypothesis about such matters and to validate the effectiveness of changes we make to our sites, tests are critical (we’ll cover this in Part 4).

---

How We Should Use Analytics

---

 The problem for some people is that they are unsure of what metrics to use and how to calculate them. This guide does not aim to be a comprehensive guide to learning about every type of metric you can use – nor would that be very helpful. Most of us are not data analysts by training, and trying to teach you to be one would not be an effective use of our time. But furthermore…

… for other users, the problem is the exact opposite – they are so immersed in data and metrics that they become inundated with numbers and figures. Buried in so many metrics, they become unable to see the bigger changes and drive actionable insights (this was my problem when I started my career doing business intelligence analysis work for Google).

One problem is not having enough analysis to make an informed decisions and the other is having so much analysis that you are paralyzed by data overload.

Both of these problems can be solved with a simple philosophy I have about using data analytics for digital brand managers and bloggers.

1. Minimalism is beautiful, both in a Helmut Lang dress and in data analysis. Less is more. Only do what’s needed to solve the problem. If you don’t need to run another piece of analysis, then don’t do it. Excess data only complicate things.
2. Always know what you’re trying to solve before you dig deep. Top level data helps us diagnose problems and see the big picture, use it help you flag potential problems or identify opportunities. If you try to collect data before you have a problem you’re trying to solve, you’ll never stop collecting data because you’ll never know when you have enough. (Recall the first rule) 
3. Data doesn’t know your site visitors, you do. Don’t forget to always rely on your intuition and experiences, because they help you understand your metrics and drive insights.
4. But don’t become so biased by your beliefs that you forget that you may only be seeing a small part of the picture, and data analysis is helping you see the whole picture.
5. Everyone makes mistakes – you might have an amazing idea to grow your site and the data supports it. It still might not work. Testing is a critical component to validating your ideas before you commit to it.

In Part 2 (Making Insights), we’ll talk about how to make insights from data from a hypothesis driven mindset. Instead of telling you about all the different kinds of metrics you can use, this guide will teach you how to identify situations in which using data metrics can be useful and how to make insights that will let you make smart decisions.

In Part 3 (Driving Change), we will talk about how to drive changes on the site from your insights. We’ll focus on how to interpret your insights and how to easily translate them into actions that will allow you to make changes on the site. We’ll discuss the limitations of your insights as well so that you don’t let data control you.

Finally, in Part 4 (Testing), we’ll go over the importance of testing your changes as well well how to do some basic tests.

Winning “Best Application Award” at the Hackathon

Some of the greatest products are made from the unlikely combination of two seemingly disparate fields. Velcro and dynamite, though they might be considered more “accidental discoveries” than anything, highlight how innovation can come about when cross-pollinating ideas. When I was at IDEO, considered one of the world’s most innovative design companies in the world, I came to acribe to their “human-centered design” method to solving problems. Applying this philosophy to how I approach business, design, and even academics has been one of my most empowering ideologies.

Last week I had the opportunity to be part of an amazing experience – the 2012 PennApps Hackathon, self-proclaimed as the largest student-run hackathon in the nation, and with over 140 entrants and news outlets like TechCrunch and Wired covering the event, just being able to witness the event unfold was really inspiring (though mostly filled with sleep-deprived face-palming and non-stop working).

What is a hackathon? If you’re not a programmer, you might be inclined to recall that scene from The Social Network where a bunch of nerds engineers drink booze and code all night long for glory and honor! Well, not totally off the mark. At its most basic level, a hackathon is an event where programmers collaboratively build programs or application within a specific period of time to encourage innovation, fun, and networking. The PennApps Hackathon is a 48-hour hackathon (meaning we only had a paltry two days to make something from scratch) done in teams of up to four people to make an application (not a program).

Ok but first, pause a bit. You might be thinking why I would be involved in a hackathon. Sure, I can do minimal front-end work with PHP, CSS, and HTML. But start asking me to incorporate more advanced coding like jScript and jQuery and you’ve lost me. Admittedly, when I was approached by a friend to be involved in the process I was wondering why ask me to join. In a 48-hour time frame, in which every second counts (and trust me, every second did count, we submitted our code 5 minutes before the deadline), you’d want as much coding resource and expertise as possible.

There is a danger to that mindset though, as my friend reminded me. From his experience, when four programmers get together to make an application, they think the same way, their strengths are multiplied but so are their weaknesses. And oftentimes they just mash functions and APIs to make what they think is interesting from a technical standpoint, but from a user point of view is a Frankenstein-like product with poor user experience and design. Why I was asked to participate in a hackathon was not to be a programmer, but to give a unique design direction and shape product development from a user experience perspective. (Which harkens us back to the philosophy of “human-centered design” and the need for cross-pollination of fields.)

Ultimately, we created a drink discovery application for the Android phone that lets you discover mixed drinks through fun experimentation. You can check out the site we made (it’s on Tumblr since Tumblr was a sponsor for the competition) here and hopefully it will be available to download one day! The application won the Best Mobile Application Award, which was amazingly humbling and exciting for our team. Everything from the user interface (UI), graphics, mock ads, site, and even wording are part of a cohesive aesthetic direction (I opted for a more playful 1960′s cocktail era feel). This is important because the application’s front-end design is critical to user perception of the application and even a small inconsistency in this can absolutely ruin everything (branding operates in a similar way). The user experience was designed around optimizing ease of use and intuitive functionality that doesn’t require any explanation to use while still offering a robust and powerful experience for users. In other words, we took a cue from Apple and tried to make the application as simple to use as possible – minimal user pathways while still providing powerful functionality. For us, this meant the user being able to consistently discover new drinks and being able to share new recipes and ideas with friends.

Fashion & Digital: A Rocky Marriage

Creating a mobile application for this hackathon made me wonder why we didn’t see more mobile applications in the fashion arena. It isn’t all too surprising though, fashion has been notoriously bad at integrating technology. Social media and fashion makes so much sense from a content distribution perspective and even from a sales perspective, yet brands have been woefully slow at capitalizing on the trend. Even worse were brands trying to fight against the digital transformation and magazines seeing social media as an “us vs. them” situation when in reality it should have been a question of how to integrate the digital into the analog world of print.

Mobile (smartphone) is one of the fastest growing sub-sectors of the digital era. The mobile market is one of the fastest in the world, with its growth even surprising Google, with 42% of consumers already owning smartphones and 84% of those users actively use their phone while shopping in-store (Deloitte’s 2011 Holiday Survey). The Mobile World Congress provides these additional statistics that should give us all pause:

1. Smart phone sales are projected to exceed global PC sales by 2014
2. Mobile search and web use is growing 8x faster than the equivalent point 10 yrs ago for computers
3. Half of new online connections are from mobile devices

Questioning the recent L2 report published on Mobile IQ for prestige brands, Marketing VOX noted that despite the celebration of successful case examples in the industry, there is an overall underinvestment in the mobile sphere from luxury/fashion brands:

The IQ distribution reveals a significant, widespread underinvestment in mobile, as nearly half the brands were designated as feeble. The report found that only two-thirds of prestige brands maintain a mobile-optimized site, and one-third of these mobile development efforts do not yet support commerce. Also, while 70% of brands have mobile apps, fewer than 20% of brands have created unique app content for tablet devices such as the iPad, which register high usage among affluent consumers.

There is a huge market opportunity in digital for fashion, and within digital, the lowest hanging fruit is clearly mobile. Especially with the boom in tablet devices, fashion brands should be strongly considering how to integrate their marketing and even sales operations with their mobile strategy. But of course, you need a mobile strategy to begin with (something I fear some brands don’t even have yet).

Remember my philosophy of cross-pollination innovation? Fashion and mobile is an area ripe for this kind of innovation. Human-centered design in fashion might seem strange, given that fashion has historically been a form of art-centered design; but in the right context, an anthropological approach to fashion lets us think critically about how people can interact with fashion through the mobile and digital medium. This mindset will enable designers and retailers to create powerful new ways for fashion brands to create a new revenue stream, gain brand exposure, and engage with consumers in a fundamentally new and powerful way.

1. High-End Designers Move Downmarket

Designer collaborations with the likes of Uniqlo, H&M, and Macy’s made big news last year and it seems over the past couple of seasons more and more designers are trying to get a piece of the collaboration scene. Last year’s big names were Missoni for Target and Versace for H&M collections but let’s not forget Karl Lagerfeld for Macy’s and Giambattista Valli for Macy’s. Up and coming is Marni at H&M, another Versace collaboration at H&M, and Jason Wu for Target.

From a business perspective, a collaboration makes sense for smaller designers who may not have the manufacturing and distribution resources of a larger retailer like H&M or Target. As designers increasingly favor the use of collaboration lines and as more and more retailers offer to work with designers (would love to see AllSaints come in here), I’m both excited and scared to hear about what new lines await us as matchmaking intensifies between retailers and designers.

3. Entertainment Tries To Move Somewhere with Fashion

Well, while fashion is moving all over the place this year, entertainment is trying to also cash in on fashion. From Project Runway’s All Stars  to Joe Zee’s All on the Line, Americans are loving the rise of “fashion entertainment”. This is a lovely trend indeed for most fashion icons, because trying to get more popular exposure seems to be second nature (present company not necessarily excluded). Designers and fashion icons seeking ways to get involved in Hollywood fret not. Already we have or will soon be watching Lifetime’s 24 Hour Catwalk (hosted by Derek Blasberg and Cynthia Rowley), NBC’s Fashion Star, ABC’s Jane by Design, and Bravo’s It’s a Brad, Brad World. 

Maybe it will be the evolution of fashion’s street style. Could you imagine reality TV around the lives of fashion bloggers?

4. Stealth Wealth & Subdued Product Design

Occupy Wall Street reminded us of issues facing our society vis-a-vis income inequality and wealth distribution. In politically sensitive times, consumers are more conscientious of how they present themselves in the face of their social circles and the public. In an era of online shopping where many consumers already indicate that they prefer shopping from the privacy of their homes and as American markets become increasingly sophisticated in their fashion taste (moving away from branded replenishment accessories to actual fashion pieces) the fashion world will need to pay increasingly attention to shopper preferences for product design. Consumers will turn away from ostentatious display of wealth, instead opting for more unbranded items and pieces that to the untrained eye will not appear to come from any designer.

5. Continuing Expansion in China

China continues to represent one of the fastest growing markets in world across industries but especially for fashion. As China’s wealth grows, so does demand for access to luxury. And when Chinese cultural norms make it acceptable and even important for both men and women to carry bags as a status symbol, fashion labels are completely enamored with the China market. Of course, rapid expansion has its own problems and as brands venture to a new country, poor planing, miscommunication, and cultural misunderstandings are common obstacles.

Brands are still trying to perfect their China strategy but in the meantime we can learn from the mistakes of other brands – some more obvious than others. 

6. While Thinking More about Brazil

Brazil, while not getting as much attention in the global market as China, is also part of the Big Four. While fashion media has focused its attention primarily on China, we should not forget the huge market opportunities in Brazil as well. Indeed, over 30 brands are already set to open news stores in Brazil including Prada, Dolce & Gabbana, and Fendi.

Of course, just like in China, expanding to a new country has great upside but also has its own pitfalls. Among other potential problems, brands must be cognizant of the dangers of licensing as a means to expand to new markets, localization issues, and cultural differences.

7. Private Sales Model Evolution

While Gilt Groupe’s success continues to enthrall business professional across industries as an amazing success story, they are still not profitable. And with more and more start ups going the private sales route, the increasing saturation of Gilt Groupe copycats does not necessarily bode well for anyone with a similar business model. Even more problematic is that brands are tightening inventory thanks to better supply chain planning. As the economy improves and overstock decreases even more, private sales sites will need to evolve their model to accomodate changing market conditions.

8. Fashion Bloggers & SOPA

SOPA, the bill that will definitely censor and restrict the way you can access the internet, will have a very real impact on fashion bloggers and social media if it passes. Hopefully, we will be able to update this list in the future with a giant cross-out on this item, but for now SOPA threatens to force fashion bloggers to be more cognizant of where their images and content comes from as they now risk an entire site shutdown if a copyright holder complains.

Even barring passing of the bill, SOPA has reminded the industry of the increasingly difficult nature to track copyright and protect it. Fashion digital content is no different. As the blogging phenomenon is made more accessible with sites like Pinterest and Tubmlr, fashion copyright seems like an extinct creature in the 21st Century. However, bills like SOPA are attempts to address copyright violations. How will the fashion industry follow suit? Will sites that make money by publishing fashion images like Style.com and spend money buying rights to the images try to restrict content being taken from its site?
9. Traditional Media Integrating With Social & Digital Media

As traditional print media continues to decline in both readers and ad revenues, fashion magazines must continue to innovate their business model in order to stay relevant. Fashion magazines reluctantly inserted themselves into the digital world, but now appear to be embracing the trend. Last year, we saw more and more magazines establishing a presence in social media, creating a digital version of their magazine (with online-only content), and even introducing iPad applications for their publication. More creative ideas include 3-D content, fashion films, and moving fashion images.

This year will be no exception to the trend, as more magazines strengthen their digital position, but also be on the lookout for continued business innovation. Magazines are looking to collaborate with the entertainment industry as well as incorporate an element of e-commerce to boost their revenue through new streams.

10. Will Dior Ever Make a Choice?

But seriously… will they?

Vera currently has a three-fold market distribution strategy that focuses on the high-end market, the upper-middle tier of aspirational consumers, and the mass-market. Vera Wang and Vera Wang Bridal focus on the upper-end of the market, while Lavender Label targets an aspirational segment that lies between affordable and luxury prices. At the mass-market level, Vera has Simply Vera, several licensing deals, and several other diffusion lines like White by Vera. (Note: As much as I generally dislike how diffusion lines dilute brand equity, Simply Vera has worked wonders for Kohl’s.)

This distribution strategy seemed to work fine until the global recession came around and shook up the entire market. Now, Vera is opting to shift her focus from high-end downmarket to where Lavender is currently positioned. The upper-middle tier is an increasingly popular niche for fashion brands to target with prices ranging from the hundreds to low thousands as the economy is slowly recovering and consumers regain an appetite for fashion. To reinforce this change, Lavender is now going to be called “Vera Wang” (the current name for her high-end label). Her upper-end line will be renamed “Vera Wang Collection”.

When talking to the WSJ, Vera Wang Group’s president, Mario Grauso, noted that Vera Wang’s upper end line wasn’t producing the profit margin he’d like to see:

“There’s much less of a customer for Collection…The cost is enormous. I don’t make money doing it. I lose a lot.”

This is a common scenario for fashion brands. While brand equity is built from designer collections, revenue is driven downmarket from lower end lines. In fact, most high end labels lose money on the collection, but make up for it via perfumes, bags, and accessories (most of which they don’t even produce and design in-house).

Of course, there is a careful balance that needs to be made between aggressive sales through lower end lines and maintaining an exclusive, high-end image. While Grauso is correct in asking the rhetorical question:

“How many women are really interested in clothes off the runway that are a little challenging and weird-looking?”

The point is not that women are really interested in buying the clothing from the runway, but that they fantasize about being able to afford a gorgeous and almost theatrical Vera Wang dress then buy something from Lavender to satiate that dream.

The Vera Wang Group is charting through an exciting time as it considers how to best grow. The brand is expected to reach $1 billion in revenues this 2012 and this new market strategy should help them reach this point. I fundamentally agree with Vera Wang’s strategy, but the devil is in the details. Can she grow her RTW label without alienating her traditional consumer base upmarket? Can she grow sales down market without deteriorating her strong brand image? Will Vera ever agree to pretend to be my mother for a day? Questions, questions.

Image: Vera Wang Resort 2012 (source: Style.com)

I loved hearing about their strategic initiatives to boost their brand reputation by cutting down assortments and changing the store facade away from their Tom Ford era. Their logic in translating aesthetic and creative issues into tactical business actions was well defined. The overall strategy was articulated in a way that convinced me that Gucci is heading in the right direction. I am excited for the brand as it continues to innovate its product offerings and brand messaging.

Among Gucci’s major initiatives include the following:

• A major facelift to existing stores to update the feel of the brand past the Ford era
• Retail expansion into Asia, especially focusing on China - 21 percent of the company’s revenue comes from China at the moment according to Patrizio di Marco, Gucci’s president and CEO
• Stepping up their social media efforts, such as (1) finally acquiring the @Gucci handle in early 2011, (2) launching a China Gucci blog in late 2011, and (3) working with an amazing street style blogger to help snap Gucci across the world
• Refocusing messaging across advertising channels to focus on the quality and heritage of the brand & collaborating with iconic, Italian products and figures such as Fiat 500 and Martin Scorsese
• And a handful of other interesting tactics like selling Gucci bags in 7-11′s in Korea (yes, it’s true)

PPR’s Gucci Group is second to only LVMH in size and profit (think eternal rivalry like Ash and Gary from Pokemon, or  pirates and ninjas, or just Red Sox vs Yankees). While Gucci Group might (2010 annual profit $1.3 bn) not be overtaking LVMH (2010 annual profit $4.1 bn) in the immediate future, market signs from the past year have indicated that Gucci’s new strategy is having a positive effect.

From the NYT (Feb 2011):

PPR’s annual profit rose 1.4 percent last year to 965 million euros ($1.3 billion) from the previous year as revenue climbed 7.5 percent, to 14.6 billion euros ($19.7 billion), the group said Thursday.

I especially want to applaud the digital and social media team at Gucci – be sure to congratulate them on Twitter (@gucci) for their outstanding efforts. From what was literally a fragmented and minimal presence in social media to now one of the fastest rising in the industry, Gucci’s online efforts have resulted in significant online brand equity, joining the ranks of Burberry and other digitally savvy brands. From the WSJ (Nov 2011):

For the second year in a row, Gucci — the Italian luxury brand known for its “kissing G’s” as well as its life-in-the-fast-lane image — is the most searched fashion brand online.  That’s according to Microsoft’s Bing search engine, whose research, released today, is part of Bing.com’s annual search-trends list.

It’s a new year and between the 10th attempt to renew my resolution that “thinking about the gym burns between 0 to 0 calories” and the prophesied end of the world, it will be exciting time to continue to follow Gucci’s digital and retail expansion strategies.

In fact, when trying to recruit volunteers for a fake study within the study, wearing a branded label helps improve response rates:

“To see if this perception had an effect on actual behaviour, the researchers did a number of other experiments. For instance, one of their female assistants asked people in a shopping mall to stop and answer survey questions. One day she wore a sweater with a designer logo; the next, an identical sweater with no logo. Some 52% of people agreed to take the survey when faced with the Tommy Hilfiger label, compared with only 13% who saw no logo.”

When playing a classic negotiation game in which one person gets to decide how much money to transfer to a partner, those who wore a designer shirt got 36% more money than someone wearing a plain shirt. This result was replicated in varying degrees to other aspects of human interaction from charity to job performance. For example, wearing a Lacoste shirt can command you a 9% higher starting salary at a job interview.

This study strengthens the assertion that brand equity posses real symbolical power over others.  Time Magazine suggests that carrying around an Hermes Birkin could help get a higher salary and better job recommendation. I think there’s a lot of interesting potential research here that could really help improve both our understanding of how fashion functions in society as well as help brands improve their business operations.