Direct to consumer online-only retail is not disruptive because its online. Do not mistake good marketing for a new and innovative business model. Warby Parker was a special case driven by very interesting market dynamics that don’t apply to the other companies in the space. Let’s dive in and take a look.
Over the last few years there has been a surge in new direct-to-consumer online-only companies, particularly in the apparel space, companies like Everlane, Pickwick & Weller, Beckett Simonon, and Pistol Lake. These new models have been framed as disruptive innovations by investors, the media, and of course the startups themselves, but the fact is that they are not offering something that is fundamentally different from the market incumbents from a business model perspective. These verticalized retailers are mostly trying to replicate what Warby Parker has accomplished with eyewear, but unfortunately that category had very specific advantages that don’t apply to other vertical apparel segments in general. Warby Parker used online distribution to put the first chink in the armor of a huge monopoly, Luxottica.
It owns Ray Bans and Oakley. It makes glasses for other brands it does not own, including Ralph Lauren, Chanel, Prada, Burbury, Prada, Tiffany, and many more.
It owns glasses stores Lenscrafters, Pearle Vision, Oliver Peoples, Sunglass Hut and others.
It owns Eyemed, the nation’s most popular vision insurance company.
Breaking this monopoly is the reason Warby Parker has been so successful. They were able to cut the cost of designer glasses in half while managing 60%+ operating margins and doing so while still donating a free pair for each one they sell.
Online Only Apparel Retailers
Unlike Warby Parker, who was offering something fundamentally new in designer quality glasses at half the price, the online only apparel retailers fundamentals are already very similar to the incumbents in terms of price, quality, and underlying costs.
These new online-only apparel brands claim to bring you designer retail without the middle-man. I bet you can’t name one direct to consumer designer retailer that sells glasses that existed before, but you probably have heard of some designer direct-to-consumer retailers that sell apparel or footwear: J. Crew, Banana Republic, American Apparel, Gap, Aldo, and many more.
These new brands often claim their advantage is eliminating the middleman and the markups that go with them, but these are straw-man middlemen. These brands compete with name-brand vertical retailers, not the designer brands they claim to compete with. Their economics are strikingly similar to incumbent retailers and their online only nature helps margins a lot less than they claim.
Looking at the numbers, J. Crew, Banana Republic, and Gap have COGS (Cost of Goods Sold) of around 60% of total revenues. Online-only retailer Everlane sells shirts they claim to make for $6.70 for $15. Their cost of goods is already at 45% of revenues before even factoring in merchandising, warehousing, and distribution costs. (which big brands are required to include in their COGS calculations). Beckett Simonon says they have similar COGS of 60%. This isn’t any different than the COGS of the big direct-to-consumer brands. The difference isn’t made up in selling, general and administrative expenses either as data points indicate that online only brands have seen that those costs approximate the cost of retail stores themselves (as we will see in a minute).
Growing Up Offline
As these new online-only brands continue to seek growth it seems likely they will pursue distribution across all available avenues. If there is a positive tradeoff between lower margins and higher profits by increasing sales volume, it seems very likely that these brands will make those tradeoffs. The obvious application of this strategy is opening stores. We’re already starting to see this with formerly online-only brands launching forays into bricks and mortar retail.
Bonobos has launched a number of Guide Shops. Bonobos CEO, Andy Dunn, cited a couple reasons for this. First, they “found that about half of would-be customers would not order apparel online because they wanted to feel the merchandise.” and second, that “the cost of marketing a Web site and the cost of free shipping both ways was approximating a store expense.” We’re already seeing younger brands trend in this direction as well. Frank & Oak dipped their toes into the water of brick and mortar retail with a holiday pop up shop this year. In general, the trend is for these online-only stores to move offline as they grow.
Why Now?
So, if there isn’t really any underlying business model advantage why are we seeing so many of these new brands show up now? My theory is that there is now a critical mass of consumers that are now comfortable with shopping online has given brands two things they’ve never had access to before: low startup costs and broad early reach.
Until recently, starting a vertical retail brand meant opening a physical store and being limited in audience to customers in the nearby surrouding area. This meant paying for expensive retail space before having any idea if your concept would even work. Now brands can test their ideas and acquire audiences before they even launch. Flint and Tinder, a men’s underwear brand, raised almost $300k from over 5,500 backers before they even shipped their first product. Everlane had tens of thousands of signups before they even announced what they were building. Because of the proliferation of social media, new advertising technologies, and better analytics it is easier than ever to reach more consumers more often in ways that are more targeted for less money. It is easier than ever to build a brand from scratch.
These trends are enabling new brands to quickly discover where there are gaps in the single-brand retail marketplace and scale up quickly when they hit on something that works. Frank & Oak is filling the gap for J. Crew style men’s clothing at a sub-$50 price point. Bonobos filled the gap of a better fitting preppy brand, or maybe just an alternative to guys who are tired of Brooks Brothers. With some of these new brands its unclear what gaps they’re filling.
Who Will Win?
The brands that will do well are the one’s that understand that they are not going to win by leveraging better economics due to their "online only” nature, because these economics don’t exist. The brands that succeed will be the ones that are good at finding gaps left in the single-brand retail marketplace that previously have not been exploited due to the difficult pre-internet economics of starting a new single brand retailer, and agressively pursuing them.
The future of analytics is event based, with events tracked back to individuals. People not pageviews. There are a couple obvious trends that are driving this:
Tablet and mobile are now growing at a much faster rate then the web, as mobile moves to take over web in terms of market share we’ll see apps driving more usage than websites.
Websites are moving from thin-client to thick-client and single page applications and other similar Javascript heavy sites are becoming the norm. Client-side frameworks like backbone.js, ember.js or new full-stack frameworks like Meteor are further enabling this.
This shift has led to the introduction of event based tracking tools like Mixpanel, KISSmetrics, etc. and the beginning of the end for any traditional pageview-based analytics solutions (Google Analytics, Chartbeat, etc) that don’t figure something out quickly.
As the data becomes easier to capture people tend to capture more of it and more data leads to more interesting applications of that data. New event-based analytics data has many applications other than just static data analysis. We’re now seeing the usage of what would previously been considered analytics data to do very sophisticated analysis or affect product behavior in real-time.
The Present
To date, most of the startups playing in this market offer the ability to do one thing that is important enough to the business that they’re willing to pay the high price of integration. The problem right now is that there are far too many companies out there with limited feature sets tailored towards consumer startups for a variety of different applications where costly (in terms of engineer time) integration is required. Many are undoubtably useful when implemented but the cost of integration may lead to them not being used:
Mixpanel
KissMetrics
UserFox
Customer.io
Intercom.io
Klaviyo
Vero
… and many more.
Each one of these requires data collection at the event level aggregated to people. In each case the hardest part about the sale is getting someone over the data integration hump. Is it worth it?
The Future
Similar to how we see salesforce as a ‘data platform’ for managing CRM/lead type data something will emerge that will be the ‘Salesforce’ of application level ‘event’ data aggregated to the ‘person’ level and their platform will provide access to 3rd party developers to build the long tail of tools for data analysis.
Their key competencies will be with dealing with large amounts of structured data efficiently and making integration to collect that data extremely easy.
Pro: If they become the obvious, simple integration choice for developers (while also storing the data themselves) they will have an extremely strong market position to build into this vision.
Con: They don’t actually have a metrics product and if they became threatening the leading provider could actively try to disable such an integration.
Pro: They seem to be winning event-centric metrics to date and are in a good position to open up a data platform.
Con: They may not have enough critical mass to make 3rd party integration interesting for anyone serious. Entire companies are built on Salesforce’s platform, this may not yet be possible with a Mixpanel platform.
If you liked this post you should follow me on Twitter here
If you are a guy in need of new clothes you should buy them here
Since Fred Wilson wrote the original post about “mobile first” companies nearly two years ago, there has been a large uptake in this philosophy. In general, this makes sense as internet usage is moving rapidly off the desktop and onto mobile devices. Every day more new services are being developed that are truly “Mobile Only” where their web presence will always be relatively minimal (Uber, Hotel Tonight, Instagram, etc.).
What concerns me are recent articles encouraging companies for which a web presence is part of their near term strategy to launch “Mobile First, Web Second”. I disagree that this is the right model unless there is some core part of your company’s value proposition that you can only prove on mobile. At Lookcraft we believe that the future of shopping is primarily mobile (particularly tablets), but we’ve consciously chosen a “web first” strategy. The reasoning behind this is simple, mobile development is still slower than web development by nearly an order of magnitude.
When you’re working on an early-stage startup your number one goal is to find product-market fit. If mobile is not a core part of your company’s value proposition and you are spending precious development cycles working on a mobile app instead of a website you’re severely limiting your product iteration cycle and its going to take you a lot longer to find product-market fit. Sometimes “mobile first” is the only strategy that makes sense, but until mobile development speed catches up to web development (Companies like Parse are doing a lot towards making this easier, but the industry still has a long ways to go to catch web-dev tools/frameworks) you should think twice about if its the right strategy for you.
If you liked this article, follow me on twitter here
Almost every large successful internet company right now has a team dedicated to growth, but it isn’t some sacred thing limited to companies with millions of users. Since there seems to be a lot of interest (http://andrewchenblog.com/2012/05/11/how-do-i-learn-to-be-a-growth-hacker-wor… around what the process is like, hopefully this post can shed light on the situation.
So what is growth hacking? I think of it as the practice of gathering data, exploring that data, and exploiting knowledge uncovered from that data in a systematized way to directly further the business goals of your company. Its more than just a role fulfilled by one person or a team of people, its a way of thinking you can use to make sure you’re getting the most out of the work you are putting into your company. It works for small startups (we prioritize development using this framework at Lookcraft all the way up to huge companies (Facebook’s growth team has massive influence across the whole organization).
I like to think of it as a simple 5 step process that you repeat on a regular basis:
1. Determine Your Company’s Primary Goal
Because your job as a growth hacker is to directly further the business goals of your company you must first understand what the primary goal of the company is and why. This is deliberately singular because its hard to do a good job serving multiple higher level contexts. A growth hacker with multiple goals falls prey in a more subtle way to similar forces that cause most corporate homepages to be an amalgamation of the top few things each department wanted on the homepage. You must determine what the primary goal of the company is. This goal of varies widely across markets and business stages, in a startup they can often change very quickly. Example goals for a more established company are generally more around growth and might be stated like; grow top-line revenue, grow profits, or gain market share. For a startup, goals are usually based more around discovery and might be stated like; determine the value proposition and product experience that resonates best with our target market or figure out which channels allow us to scalably acquire users in a cost effective way.
2. Identify Potential Actions
The next thing you should do after you determine your goals is to identify actions that will help you accomplish them. These might be things like growing your homepage conversion rate, getting your users to log in more often, or getting a higher percentage freemium users to upgrade. These actions are specific, and their implementation should lead you closer to your goal. You should write as many of these down as you can think of.
3. Prioritize Actions by Highest Leverage
Its important to look at the actions you’ve defined and determine which ones give you the highest leverage towards achieving your goal. Generally the actions you’ve defined will effect change in some kind of conversion funnel you care about that is directly related to the goal the company is trying to achieve. Consider which of these actions give you the greatest benefit with the least effort? In other words, think about where you have the ability to improve most on a percentage basis. If your goal is to increase number of paid users using your SaaS offering and your conversion from homepage visit to email address is already at 10% you’re probably only going to be able to achieve a 1.5x-2x improvement at the most. If your homepage visit to email conversion is at 10% and your ratio of paid users to unpaid users is 1:1000, you can probably improve that ratio by 10-50x so you’d be much better focusing your effort there. For this strategy to be effective you obviously have to have a good idea for what achievable metrics are. The best way to do this tends to be just asking people who have done similar things. There is also a wealth of public data out there for things like email open rates, freemium conversion rates, and more which you would be well served to educate yourself on.
4. Identify and Prioritize Tactics
Once you prioritize the actions you want to take you should make a list of tactics for accomplishing those actions. If the action you want to take is to increase your ratio of paid:unpaid users some tactics you might try would be things like; implementing an automated email campaign for free users at certain product trigger points, running promotions for free users to try the premium version of the product, or improving messaging for premium users with expiring credit cards (reducing churn also improves your ratio). You should then go through the same process here as you did for determining actions. Look at each tactic and estimate how much effort it might take from an engineering and product perspective versus the estimated perceived benefit, cut and/or implement features accordingly. There is also a large educational component here. Its hard to prioritize effectively if you don’t have a good understanding of the tactics that are available to you and their estimated effects. Again, I would suggest asking around here (Andy Johns from Quora and previously Twitter and Facebook gives an excellent starting point on this Quora thread), but this is really something that you tend to learn through a lot of trial and error, and tactic effectiveness can also vary widely from product to product.
5. Execute
Its a trial and error process, so launch and iterate. Good luck!