RIP WebTV – The Internet For The Rest of Us

Today, we mourn the passing of a truly amazing product. After 17 years of faithful service, WebTV (later known as MSNTV) has officially shut down for good. WebTV had a huge impact on my life, both personally and professionally, so I thought I'd take a moment to talk about what I loved about WebTV, what I'll miss, and what I'll never forget.


I also recommend reading Brad Hill's farewell to WebTV article. He experienced a different side of the company's story than I did, and his article is a great read.
An Inside View of the WebTV Revolution That Didn't Happen
- By Brad Hill (no relation)


My Introduction to WebTV

I first discovered WebTV in 1997. I had just gotten my AA, couldn't afford to go on to a four-year university, and wasn't really sure where my career was headed. So when I met a cute redhead who told me about this great customer service gig in the bay area, I called in a favor to couch-surf with a friend for a while, moved down from Sacramento, and took the plunge.

[Aimee and Tash, I will never be able to properly repay you. Thank you!]

I joined the company a few weeks before it was acquired by Microsoft, but the negative impact from that didn't really start to manifest until a year later. That first year, WebTV was the quintessential startup environment. The CS team was housed in the same ratty garage on Alma St in Palo Alto that the company had been founded in. The building was crammed with young, idealistic 20-somethings who had their eyes set on changing the world, one user at a time.

It was a magical time for me. I was surrounded by great people, most fresh out of college and aching to make a dent in the world. We started out answering phones and responding to support emails from extremely passionate users. Within a year, we had outsourced the first-level calls and everyone on the team has chosen a new area of expertise to sink their teeth into. Some began running the Previews beta testing program, some went into data analysis to spot trends and feed the data back to the product side of the business, and some transferred to other teams like Usability or Engineering.

Me? I went into training and documentation. Shortly after starting there, I had camped out in front of my TV for a week and — with just a WebTV, a Laura Lemay book, and a Geocities account — taught myself HTML. So as we began to outsource front-line CS, I built a CS Agents intranet that would let anyone from any of the three call centers get to any of a few hundred support issues (the bulk of which I helped write) within three clicks. I took Andy MacFadden's incredible "Greater Scroll of Dialing Wisdom" document (still the best developer-written documentation I've ever read) and distilled it down to a Lesser Scroll for advanced connections agents, wrote a IRC chat FAQ and a scad of internal documentation, and flew to call centers in Florida, Kentucky and Salt Lake City to train the next wave of customer service agents, who would be the first point of contact with our wonderful users.

I even recreated a vitrual version of the WebTV Classic, Plus and DishPlayer services, using service screenshots tied together via imagemaps, so the CS agents who didn't have a WebTV box at their desk, or had to share one, could still follow along with users while explaining how to get connected. Now that the service itself is going dark, I have posted an archival version of this for posterity.

Those were the days. We would get calls from 70 year old users who were literally crying because they were so frustrated at not understanding how email worked, or how to get into a scrapbooking chat room. With patience and a healthy supply of metaphor, we'd explain it in a way they could grok, and send them on their way. And a few months later we'd get a call from the same user, complaining that the HTML in the website they were hand-coding on Angelfire wasn't working the way the expected it to. If we ever doubted for a moment the potential for a simple user experience to empower normal people to do truly mazing things, WebTV users would quickly set us straight.


WebTV's Original Mission

The original idea behind WebTV was simple. Provide an inexpensive, easy-to-use option for normal, non-technical people to get online. "The Internet for the rest of us" wasn't just a tagline, but a mission statement. If $2,000 for a PC seemed absurd, you could drop as little as $99 on a small device that plugged into your TV and gave you access to email, bulletin boards, IRC chat, and a rudimentary web browser, with as little pain as possible, for just $20/month. Add another $70 to replace the remote control with a wireless keyboard and you were off to the races.

While younger users (like me) might quickly outgrow the system and move on to PCs, the core WebTV users turned out to be the older or less technologically inclined half of the population, who didn't want to over-complicate their lives with too much tech, but still wanted to keep in touch with their out-of-town children, share baby photos through email, and connect through bulletin boards and chat with others who loved the same things they loved. Rarely have I ever met users who were so passionate about a product as WebTV users. To many of them, WebTV was the internet. And the fact that the WebTV service has stayed alive as long as it has is a testament to their devotion.


WebTV Plus, DishPlayer and TiVo

But getting the web to run on a TV was just the start. The bigger long-term bet was on integrating the web with TV, to create a seamless hybrid of the two. WebTV Plus took the integration one step further, allowing you to plug your cable signal into the box, and providing an extremely easy to use interface for searching TV listings, keeping track of your favorites, and even tying the shows you watched to related websites.

WebTV Plus was ahead of its time. Many of the innovations it pioneered (including the ability to click through to purchase or get more info on an item directly from a TV commercial) are things that today's technologists still strive to include in the next generation of set-top boxes.

And with the DishPlayer box (essentially a WebTV Plus built into a DishNetwork satellite box), they took it one step further and included one of the first-ever digital video recorders. DishPlayer and TiVo were in development at the same time, with offices just a few blocks from one another, and launched within a few months of one another. It still kills me that the DVR race played out the way it did.

At launch, DishPlayer had all the same functionality as TiVo, a much more intuitive user interface (most people forget the initial learning curve TiVo users faced), the ability to play games like Doom and You Don't Know Jack (hey, it was the 90s, these games were pretty cool back then), and the entire internet to boot, for essentially the same price. Sadly, Microsoft really didn't understand what they had on their hands, and put almost no money into marketing DishPlayer. Meanwhile, their pluckier startup competitor with a clunkier UI, far less features, and a cute animated mascot marketed the crap out of their product, and now TiVo is a verb and hardly anyone remembers DishPlayer.

Would TiVo be a verb today if WebTV's founders had said no to Microsoft's $425 million offer and run full steam ahead the way TiVo did? And would it have taken another 10 years for set top boxes to start infiltrating the living room like they're finally starting to do now? The world will never know. But I'll always have my suspicions.


How WebTV Fell, but Refused to Die

For the first year after the acquisition, Microsoft mostly left WebTV alone. They knew they had something amazing on their hands, but didn't really understand it, so they let us continue working our way. Besides the obvious flaw of not believing in DishPlayer enough to properly fund its marketing, Microsoft started to dig its teeth into daily operations.

Emily, the CS manager we all loved and adored, left the company and was replaced by a devious, soulless shell of a human being Whose Name Shall Not Be Mentioned. My own manager, Steve Kroll (who I would have gone to hell and back for) finally left the company in frustration, and was replaced by an old-industry manager who literally couldn't even type, much less grok the technology industry. And the two of them proceeded to replace all the passionate, knowledgable startup staff with business types who didn't know the product and didn't care about the users. I was lucky enough to find a new startup gig and left the company the day before our group moved into Microsoft's new Silicon Valley campus. By the time Soulless and Clueless had been fired (a few months after I left), they had already cost the company at least a dozen of its best CS employees and laid waste the the team's morale.

Some of the other organizations might not have ben hit as badly as CS. And indeed, there were a few of the old WebTV crew who spent the next decade of their career quite happily at Microsoft, both moving to the Xbox and Hotmail teams (or to the Seattle mothership) and doing their best to keep the WebTV service alive and kicking for as long as possible.

After the move to the Microsoft campus, the product remained on the shelves and in active development for several more years. They rebranded it MSNTV, upgraded the browser, migrated the email to run on Hotmail servers on the back end, and even came out with a box that supported broadband. Eventually, despite continued demand for the WebTV/MSNTV service, demand for new hardware decreased to the point where they stopped selling it. Despite the lack of new hardware, the service continued to live on for many years beyond.

The fact that today's shut-down is happening 17 years after the acquisition, and 16 years after the Dark Days should stand as a testament, not only to the herculean efforts of Andrew Levin and his team, but of the incredible staying power that was built into WebTV's business right from the start.

Today, we live in a world where most "products" on the web are subsidized by the advertising companies who use them to harvest ad targeting data from their users. No matter how beloved a product is, if it ceases to be sexy or cool, or provide enough ad-targeting intel, it gets the axe. (Google Reader users, you may now rant.) But WebTV was an old-school business. They made a product that served a need, and charged a fair price for the service. This gave them product a much sturdier foundation upon which to stand.

If I recall correctly, WebTV had a base of close to a million users at its peek. This number doubtlessly dropped steadily over the years. But with the majority of the service already up and running, the service could likely be run on a skeleton crew of maybe a dozen people plus server costs. I don't have access to any internal numbers, but by napkin math, at $20 per month per user, you would only need about 10K users to cover costs. Beyond that, you'd be looking at another $2.4 million in revenue for every additional 10K users. So even at only 10% of its peek user base, you'd still be looking at a business unit that generated around $22 million per year in profit.

It's hard to kill a product that refuses to stop generating income. I sincerely wish every entrepreneur today would take this lesson to heart. (Someday I'll write a post on the history of Half.com. It's a very different product from WebTV, but the lesson is identical.)


WebScissors and the Legacy of Jos

I made a lot of friend in my time at WebTV. Easily the friendliest and most prolifically talented of them was Jos Claerbout. He was one of those guys who just radiated friendliness. You couldn't meet him and not instantly adore him. And his brain simply wouldn't stop producing. He thought objecting oriented programming should be easier to learn, so he wrote the funniest, most useful object oriented programming tutorial I've ever seen: Don't Fear the OOP: Why coding Java is just like writing a trashy Western novel. He saw users having trouble making their own websites? He built his own tool, WebScissors, to make it easy for them to import images to their WebTV scrapbooks.

Sadly, he shared so much of his heart with the rest of the world that it gave out on him far too soon. Jos's death had a huge impact on everyone who knew him. His father, Jon, put together a beautiful site on The Life of Jos Claerbout that crystalizes many of our thoughts and memories of him.

Before he died, I had been talking to Jos about working with him to upgrade the user interface for WebScissors, and add the ability to view HTML source code. So I was deeply honored when Jon and Andrew Levin (who was maintaining the site on Jon's behalf) invited me to help maintain WebScissors for posterity, and add the features Jos and I wanted to bring to our users. To this day, WebScissors 2.0 is alive and well, and we have no intention of letting it go any time soon. It will outlive the WebTV service itself, as a testament to Jos's creativity and dedication to his users.

And it may even continue to be useful, as today's tablet users often face similar browser limitations to those that WebScissors necessary in the first place, all those years ago. Time will tell...


The Legacy of WebTV

WebTV was an amazing product that was far ahead of its time. It taught me the importance of usability, and how a simple interface and a few visual/textual metaphors could bridge the gap to make complex technology accessible to the masses. It taught me the importance of community, and how deeply users will love your product if you fight like hell to make it serve their passions. And it gave me the curiosity, inspiration and opportunity to start a career that still gets me excited to this day.

But the true legacy of WebTV is its people. WebTV alumni have gone on to play key roles in creating some of the most ubiquitous technology on the market today. I've lost count of the number of WebTV alumni who have worked behind the scenes at Apple to continue the mission to bring the internet to the rest of us. Andy Rubin, Andy MacFadden and their team are behind the Android smartphones that appeal so well to the masses. And quite a few of alumni have remained with Microsoft, contributing to the Xbox platform and other products within the company.

And let's not forget the users. Never, ever forget the users.


As WebTV fades to #191919,* I will always cherish the people it introduced me to, and the lessons it taught me.


* WebTV Trivia:
It can be really hard on the eyes to read contrasting text on a true-black background. Especially on a low-resolution TV screen.
So all of the "black" screens in the WebTV service actually used the shade #191919 instead of true black for the background color.

My Thoughts on the Netflix / Qwikster Controversy

I wrote this in 2011 when Netflix announced that they would be renaming their DVD service to "Qwikster" and separating it completely from the existing Netflix site. I wasn't actively blogging at the time, but I ran across this again and liked what I had written, so I've back-posted it here.


Core Focus – What problem are we trying to solve?

It has been said that if Union Pacific realized they were in the transportation business rather than just the railroad business, we would today be flying on Union Pacific Airlines. Rather than maintaining a strategic vision of the problem their company existed to solve (getting people and goods from point A to point B), they focused too closely on one particular solution to that problem (railroads) and were thus unprepared to respond when changing business climate made their chosen solution less and less relevant.

More recently, we saw Blockbuster repeat the same mistake. They incorrectly believed that their success was based on running an efficient DVD rental business, when in fact it was based on solving the problem of convenient access to high quality video entertainment. As improving technology made new methods of content distribution possible, they clung to their existing brick-and-mortar solution for far too long. And by the time they did try to explore other options, their years of inaction had given their competitors too much of a head start for them to stand a chance of catching up.

In order to understand the path forward for Netflix, we must first understand their core competence. As a consumer brand, the best way to do that is to ask not what products or services Netflix provides, but what problem they solve for their users. What need does Netflix fulfill that makes people love them as much as they do?

I would argue that Netflix's value to their customers is the same one that they wrestled away from Blockbuster: delivering convenient access to high quality video entertainment. If we keep this fundamental goal in mind, and focus our tactical decisions on achieving this goal, the rest of the questions become much simpler.


Brand Integrity – What's in a name?

Naming a company or product is easy. Building a brand is very difficult. Yes, you have to give your company/product/service a name. But more than that, you have to make that name mean something. Not a literal meaning, but an emotional meaning.

A brand name with a literal meaning can be useful in explaining the potential value to the first wave of customers. But as a brand evolves over time, the literal meaning is quickly replaced by what the brand means to its customers. And that meaning is defined almost entirely on how the brand makes people feel, based on their (and their community's) interaction with it.

When you consistently solve users' problems and make their lives easier, your brand will come to mean something special to them. "Netflix means entertaining the whole family without spending a fortune." "Netflix means dinner and a movie without having to go brave the weather." "Netflix means independence from an oppressive $50-80/month cable bill." "Netflix means never running out of new movies to keep my kids entertained and happy." These are the kinds of meanings Netflix users have blessed the brand with over the years.

When you have a brand this strong, introducing a new brand can be very risky. If the new product or service solves a different problem (like eBay vs PayPal, or the myriad of P&G brands), multiple brands that are expected to mean different things to different people certainly make sense. But if the new product or service simply solves a slightly different subset of the same problem (like Google's many search-related sub-brands), it usually makes more sense to keep the existing brand (and all the good will that comes with it) and differentiate the new functionality in context.

This risk only increases when the new brand is being applied to the product or service with which the users originally fell in love. Changing the name implies that the company no longer has faith in the re-named portion, and that it may no longer retain the qualities that defined its brand to begin with. It introduces friction, confusion and fear. It makes people worry. "Do they dislike this product enough that they want to take their name off the label?" "Are they just planning to sell it off to someone who's going to ruin it?" "Does the brand name even mean what I thought it did anymore?" Obviously, not questions one would ever want people to ask about a brand.

To determine whether a new brand or a sub-brand makes more sense, we must return to the original question of what problem you are solving for the users. If Netflix means convenient access to high quality video entertainment, instant vs mail order, movies vs TV vs games, and physical media formats are all just variations on that original problem and should therefore be represented as sub-brands.

Netflix Instant means instant access to (a somewhat limited selection of) great content in an instant.
Netflix [_____] means access to an endless library of content, delivered to your door in just a few days.
Netflix Games * means being able to play any game you want, for as long as you want.

Netflix still means everything it has ever meant. The sub-brands just help you tell a more tailored story to different types of users.

* Though Netflix does not currently deal in games, I include it here both in anticipation of that someday happening and to help illustrate these suggestions more abstractly than just to the two current services.


Convenience Is King - Streamlining User Experience by Platform and Preference

While the Qwikster re-branding has gotten a lot of press, the more passionate complaints across the blogosphere have been of a much more practical nature. Yes, the essence of what Netflix means has a great emotional impact. But the loss of convenience from having to go to multiple websites to do what they can now do in just one place has a much more pragmatic impact on people's lives.

To determine the best course of action, we must again return to the question of what problem Netflix is there to solve. If the goal is to delivering convenient access to high quality video entertainment, convenience is a vital part of the equation. This means getting them from interest to immersion as quickly and effortlessly as possible.

If you look at the path people take to get to the Play button, a few distinct experiences emerge. Users who don't yet know what they want to watch need a way to browse what's available, seeing recommendations and details about different titles that may increase their interest. Once they find something they like, they need a way to express interest in the title – this could be clicking a Play button, adding it to a queue, rating it, or marking it as something they already own. When they decide they're ready to start watching, they need a way to select a title – either from a browse page, queue or other entry point. Finally, when they have finished watching a title, they need a way to choose what's next.

How they navigate between these stages – and what content/functionality they are presented with – can be customized based on what we know about them. First, we know what platform they are on: are they accessing Netflix through a desktop browser, smartphone, tablet, set-top box, or gaming console? Second, we know which services they are currently subscribed to. Finally, we might know a bit about their personal preferences from past behavior.

Given this, it comes relatively straight forward to extrapolate how to guide the user through the process.

  • One entry point per platform
    For those accessing Netflix on a web browser, there should only be one website. For those accessing Netflix on their smartphone, tablet, set-top box or gaming console, there should only be one app per platform (and be as similar between apps for the same platform type as possible).

  • Browsing vs Selection
    Once in the system, users should be given the option of either browsing (and/or searching) the total collection of what is available to them or seeing a list of the titles that they have previously expressed interest in.

  • Streamlined Browsing (Requires BU Collaboration)
    When the user doesn't know yet what they want to watch or play, they also won't know which format it's available in. So browsing can easily be separated between TV/movies and games, but within each top-level category should allow browsing by intuitive filters such as genre, recommendations and past behavior, as most of Netflix's browse interfaces currently do. This is the portion that will require the most collaboration between business units, as the Instant and DVD groups must agree on that format in which the basic title info is presented, and then provide separate call-to-action hooks for each of the services that title is available for (Instant, DVD, Blu-Ray, teaser to look at the tie-in game, etc). But any extra coordination between business units will inherently pay off with a much more streamlined user experience.

  • Streamlined Selection (Allows BU Experimentation)
    Within the selection branch, it will likely make sense to have separate selection screens based on the specific services that a given user subscribes to. This can allow the various business units to have a fairly wide berth of innovation around how best to display their own content. The Instant team may choose to go with a tiled box art view with category links for narrowing large queues by mood (or who marked it, for family accounts, eventually). The DVD team may choose to reuse the mood/user filtering sidebar, but present the results in a priority-ordered text view. And the Games team may choose to use the same priority-order view as DVDs, but use a sidebar that allows filtering by platform, genre and game rating. Further customizations can be made based on platform (browser, tablet, console) and interface (mouse, touch, remote), to make it as easy as possible for a user on any given platform to quickly and easily find something great to watch/play.


Summary

People love Netflix. If they didn't, they wouldn't be reacting this passionately to recent changes.

Netflix means convenient access to high quality video entertainment. As long as a new service still addresses this core mission, it should retain the core brand (or a sub-brand thereof).

Convenience is king. While there are certainly logistical changes that can be made to both the user interface and back end processes, convenience to the user should always take precedence over the convenience of various business units.

Back in the Saddle

For the past few months, I've been toying with the idea of writing a full usability review for tribe.net. My motivation was partly to let them know what I thought they needed to fix, and how. And it was also partly so that I could use this usability review as a resume piece (not only for applying at Tribe, but other companies as well).

I was composing a general outline in my head, and keeping a running tally on the various improvements that needed to be made, and how to best apply them. But nothing was committed to writing yet.

Then Tribe put a link of their sidebar trying to get people to check out their own home page redesign prototype. They also had a survey and a "talk about the preview" tribe set up to capture users' impressions of the proposed change. I was somewhat relieved to see that I was not the only one who thought the new design sucked arse. It didn't fix any of the problems I had identified, and it broke a lot of the things that were good about the old design.

My negative reaction to seeing this prototype was enough to motivate me to site down and finally commit to writing the usability review that I had been composing in my head all this time. And, since it's often much easier to understand something when you see it instead of just reading about it, the usability review would need to be accompanied by a demo of how the site would look and work if my suggestions were implemented.

So I blew off the plans I had for the weekend, worked a few all-nighters, realized that the code behind their redesign had changed as well (also for the worse), and rebuilt their entire home page and the structural logic behind it from scratch. Finally, three days and far too few hours of sleep later, the home page is finished.

Of course, I still have to redesign all of the internal pages and finish writing the usability review itself. And I may also need to write a style guide for my demo version, so that if they do decide to implement it they don't have to reverse-engineer what I have done.

But while I'm working on that part of the project, I thought I'd post a screenshot of what I've got so far, so y'all can get a sneak peek at what I'm up to.

I'm terribly tempted to also submit this screenshot and a few of my reasons behind it to the preview tribe now, while everyone is still talking about it. But ultimately, I think it will be better to wait until the rest of the demo is functional as well, and it has the full weight of the usability review behind it. That way it comes across as a well thought out approach to the problem, instead of a "me too, me too" kind of thing.

Regardless, though, it feels great to finally be working on a real project again! I can feel long-neglected synapses sputtering and firing back into action again every time I get a tricky bit of code working. With any luck, finishing this review will give me the renewed self-confidence needed to get cracking on my Ethos project.

click for a full-sized version