On what time horizon do you work? With 20 years of the web now behind us, our field has a fair amount of history, which has left us with a few theoretical barnacles to shake loose. But it also raises questions as to the time horizon over which predictions ought to be made and roadmaps designed. And it raises questions on what our industry obsesses about in the here and now as opposed to what it could think about in the long run.

History will teach us… something?

When Tim Berners-Lee first demonstrated the technologies behind what came to become the world wide web, he had little idea of the impact his invention would have. But what has happened since may allow us to look back and better understand how to look forward. While it took a couple of years for the web to move from academic and research circles to a more mainstream and commercialized space, some of the early battles are still reflected in what we see today.

For example, there is a constant struggle for control of the viewing experience on the internet. In the earlier days of the web, that struggle was represented by the browser wars with a large fight exploding between the dominant browser creator at the time (Netscape) and the dominant Operating System vendor at the time (Microsoft). The same fight can be seen today in the struggle over whether web applications or dedicated platform specific applications ought to rule the roost in the future.

The death of Netscape left the web barren for a few years as few contender to the supremacy of Internet Explorer emerged until the Firefox project finally found its footing. Over those years, the web largely stagnated, partly as the result of a deflation of the financial bubble that had arisen over the previous year but also partly as a result of the lack of a credible contender to the dominant browser brand.

Looking forward by looking back

When looking at the future, we can be a lot more hopeful today, because the odds of a winner-take-all have been lowered in some segments of the market. For example, the rise of iOS has been counterbalanced by the rise of Android and an increase in the number of web-based apps. With such a diversity in the mobile space, it also will make it difficult for new entrants to succeed. Marketplace tend to reward only a couple of groups and we are already at a point where we have 3 players. This is bad news for Microsoft and Nokia who are trying to play this from behind and have a real uphill fight to gain any serious traction, no matter how great their offering is.

In the tablet space, the story is a little more worrisome. Apple’s current domination of that market could lead to a new stagnant space unless a contender is identified. Amazon’s positioning of the Kindle Fire as a potential alternative may present opportunities to restore a balance, forcing all players to bring their best game. At this point, it appears that the marketplace has widely rejected other offerings in the tablet space so Amazon’s entry makes this an interesting case. The question at this point is not whether Amazon will topple Apple (it won’t) but whether it will introduce a product that appeals to a segment of the consumer market and forces Apple to keep improving its offering (and here, I would present the iPod roadmap as an example of how innovation dies in a marketplace, even if that marketplace is dominated by Apple: from a product standpoint, the iPods sold today have not really changed since the introduction of the iPod Touch four years ago).

On the standard front, we are seeing some fierce battle lining up in how browser supports new features of HTML5 and related technologies. As I pointed a couple of weeks ago, this is a good thing. The current diversity of browsers in the marketplace forces every player to attempt efforts at being on the cutting edge. This is the kind of thing that will speed up adoption of new technologies across the landscape as a whole and will help the open web become a strong contender for the future of computing. Technologies like HTML5, CSS3, WebAudio, and WebGL are helping the web become more operating system-like and, as such, increase its chances at becoming a strong player on any platform.

Roadmap 2025

But looking at today and looking forward, we are seeing a marketplace that is increasingly fragmented. So what’s a startup with limited resources to do? Should one put all their eggs in a single basket and bet on a single platform? Should a company look at introducing fewer features across all platforms, spreading their product roadmap over a longer timeline? Or should developers go to the lowest common denominator and target through the web browser?

The answer is it’s complicated but, at its core, one needs to think about what is needed to move the company forward and get it to reach early measurement markers while still focusing on the future.

At Keepskor, we call that Roadmap 2025.

I know it sounds extremely ambitious for a company to think to 2025 when it doesn’t yet have a product in the marketplace. But the issue for most startup is not ambition but the lack of such thing. By focusing on Roadmap 2025, we get a sense of how the marketplace is evolving and how we want our product to evolve accordingly. Frequent readers of TNL.net know that this column is not really concerned about the next few months and tends to focus more on the underlying cultural and business trends that can have decade-long impacts. This is why things like the impact of DVRs on the television business were discussed here over a decade ago, or more recently, I looked at what challenges existed for books and television. The 2025 consumer market will be impacted by those changes and the way in which applications designed on the internet interact with that will also be impacted. As will the dominant platforms, one of which, we hope, will be Keepskor.

But how does a Roadmap 2025 look in the present? Well, put quite simply, the present is an early alpha. The minimum viable product the marketplace will accept for something much larger. But the MVP for a platform that will exist in 2025 is substantially bigger than the MVP for a simple app. And building it requires higher levels of abstraction than a normal product. For example, no one really knows what the winning front-ends will be in 2025. What we know is that they will be different than the ones we are used to today. So we design our platform to ensure that input and outputs are separated from everything else. In the short run, this creates an added layer of translation that increases latency (measured in milliseconds) to every transactions but a slowdown we consider acceptable for now.

Iterate backwards

The basic agile method of development looks to scrums that are moving the product forward in an iterated fashion, tackling one component or another over short cycles. The level of clarity as to what will happen in the next scrums is generally dependent on how far forward you’re looking as each scrum can move a product to the adjacent possible. In a long roadmap model, one iterates backward. For example, if you are looking at what your product looks like in 2025, do you know what it looks like in 2024? You may have a rough idea as to how the product should evolve along the way. Maybe you get a sense of what it looks like in 2020 or 2015?

I call this backwards iteration. Move from the ideal state and start planting base camps backwards. It’s a strategy I learned from long-distance skating (I’ve done runs of up to 65 miles on a single day). No one wins a long distance race by focusing on the end point. The win is only by focusing on the next 100 yards and iterating those forward. But before you do so, you need to have a sense of what you need to do by each of those 100 yard markers. And that means considering how tired you will be and how much slower you will be at the end of the race and working backwards to assess the best balance of expanding and depleting energy vs. yards won.

In a business, this allows you to build a model that is pretty accurate for the first couple of years, and only mildly less so afterwards. It also allows you to go back to the roadmap when major events happen and identify whether they will have any impact on your business (and if they do, whether this impact needs to be mitigated now, and can be mitigated by moving an item forward in the roadmap).

From a communication standpoint, it also makes it easier for people getting involved in your business to get a sense of where you’re looking to go. This may mean that some investors are not interested because they don’t buy into your vision (and that’s OK because those investors would be an issue further down the road if you’re still going down some of the same paths) and also means that some people will not be interested in joining the company because they don’t agree with the steps moving forward. I consider that early filtering as it ensures that your company gets proper buy-in from all stakeholders, thus lowering the chances that employees will leave once they are fully committed.

Study the past, understand the future

But a roadmap 2025 plan is not something that one pulls completely out of thin air. What it is, more than anything, is a set of assumptions based on close study of part events. For example, the majority of the Keepskor roadmap can be explained by events that have already happened. This grounding in history and historical trends helps us ground the business into models that have been proven to work in the past and which, when combined, turn out to be substantially bigger than the initial ideas.

When looking more closely at history, I get a sense that most areas of history have some cyclical component to them, with a pendulum swinging in one direction or the other but always moving us forward ultimately. The study of previous cycles is probably the greatest tool one has in understanding what the future can look like.

However, there is one thing to note: The distant future is easier to prognosticate than the closer one because things actually do change at a much slower pace than most would expect. In the 1990s many dotcommers, present company included, made what looked like wild statements about the future and how the internet would change everything. Circa 1999 and certainly in 2001, those seemed like crazy pronouncements but, in hindsight, the world we described as coming to bear at the turn of the century is pretty much the world we live in today: the revolution has just taken longer to take hold.

A personal note

For years, people have asked me why I was writing TNL.net as I didn’t seem to derive any real income from it and the things I wrote about were, on the whole, only tangentially related to what I do. Over the last year, I’ve come to discover that the writing of TNL.net is a somewhat selfish act. First, it allows me to crystallize a lot of my own thinking: the process of writing and editing an entry forces me to focus on a topic area very closely. It also forces me to review source material and historical data to ground and inform my thinking. And then, there’s the part where you, the reader, enter. Your emails or discussion help me get a better sense of some of the flaw in my argument.

Update: I generally start writing early in the week and fill the entry up over time until it gets published on Sunday evening. This week, Fred Wilson posted “Long Roadmaps”, a great entry that covers some of the same ground. I guess there must be something in the air here in NYC that gets us to think about those things.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title The long view. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Tivo and Digital Media Servers. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

Right now, this page is served to you either in HTML or RSS. Those are two different languages. One, HTML, is understood by your web browser. The way text is bolded, for example, is that I put in a little tag that told the browser to bold this word. The browser reads the tag and presents it appropriately. The other, RSS, is understood by what we call RSS readers. Those are programs that you use to subscribe to a channel. A channel is something that you would receive every day. That way, you don’t have to go and check the site to see if it’s been updated. Your program goes and gets the information. RSS is an XML-based language. What it means is that there is a lot of information in that channel that is there just for the benefit of that channel, to allow to present only the newest news to you.

A couple of years ago, Tim Berners-Lee, the man who created the web, looked at his creation and realized that there was a jumble of pages and that, in order to make sense of it, we needed to give things a little more structure. So RDF was born. What it is, basically, is a way to organize the whole web so that computers could talk to each other without humans in between. This has potential uses and here’s an example I thought up: The smart calendar.

Joe and I are working on a joint project. Joe is in London and I am in New York. I want to arrange a face to face meeting with Joe. Right now, I either call, email, or contact Joe in some way and we figure out a time when we can meet face to face, then agree on a city in which to meet, then make the necessary travel arrangements. What if I went to my calendar, typed in meet face to face with Joe, and my calendar and Joe’s started discussing when the best time and place would be? My calendar would check my availability and Joe’s. It then would check if any of us has any travel plans in each other’s city. Based on those, The two calendars discover that I have a trip to Paris set for next Tuesday to Thursday. As a result, the calendar would recommend that I go to London to meet with Joe on Friday. If both Joe and I agree to this, my computer would then go to the travel reservation system, check prices and flight times, book a flight from Paris to London on Thursday Night, cancel my Friday morning flight from Paris to New York, book a flight from London to New York on Friday Night, cancel my hotel stay in Paris on Thursday night, book a Thursday night hotel stay in a company approved establishment near Joe’s office, and notify Bob (who’s also in London and with whom I had scheduled a conference call) that we can meet face to face when I’m in London instead of doing it on the phone.

Usually, this would have taken several discussions, a whole slew of new flight and hotel reservation changes, and a lot of wasted time. Using a semantic engine, all this would be automated. A lot of computers would have talked together (first mine and Joe’s agreed on time and place, then my computer talked to the travel company with which I had my flight and changed that reservation, then my computer talked to a number of airline companies to see who had the best price on a flight from London to New York, then my computer talked to the hotel reservation system in Paris and canceled one night, then my computer talked to the company computer to see what hotels it approved near Joe’s office. It then talked to computers in the several hotels in London to find a room with my preferences and within my price range. Having done so, my computer talked to Bob’s computer to tell him to change the appointment from a phone conference to a face to face meeting. Bob’s computer talked to a computer in his building to book a conference room. Having done all this, the hotel and airline computers then talked to my company’s accounting systems to agree on billing) because they all talked similar languages (or could point to a translator who would explain how they could. THAT is a practical example that would save money (finding the lowest price on airlines and hotels, reducing the number of trips) and increase productivity (saving time spent on certain tasks) thus allowing me to spend more time on money-making tasks.

All this is still a long way away but if the dream of a semantic web is realized, it will become reality.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Geeks to English. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

First of all, by acquiring Linksys, Cisco gets a strong foothold in the small office/home office market as well as the hobbyist/consumer market. Why? Largely because this is where Linksys’ strength is. What Cisco gets out of this is a new source of revenues in a market it has had troubles getting into. The announcement that they will not change the name of the company and will let it run as an independent unit seems to point to that end.

Second, it provides Linksys with strong support in enterprise sales. Linksys has been getting into the enterprise largely through the back-door, with employees installing cheap wireless routers in offices. Now, with Cisco’s backing they can get into the enterprise as part of a more complete solution.

The next question is what this does to Cisco’s strategy as a telecom vendor. Linksys already sells voice over IP products. There may be some people at Cisco edging their bets in terms of the telecom bet. It could be that they figured that VoIP could also come in through a cable connection, which in itself would undermine sales of equipment to large phone companies and move more telephone traffic into the hands of cable companies. Alternately, Cisco could start selling solutions that would allow for a complete end to end solution for VoIP: Large offices would use Cisco’s existing solutions and smaller branch offices could take advantage of the Linksys offerings.

However, the real prize in this acquisition is in the wireless space. Linksys has already established itself as one of the strongest players in that market and Cisco will probably take advantage of those gains, eventually demising its own offerings in that market. The advantage for them is consolidation of product lines and an early toe in the 802.11g arena.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Cisco acquires Linksys. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

On the left, you’ve got Google, the 2000 pounds gorilla of search which is now looking to expand its advertising program beyond its site and is fighting to not have its name associated with searching on the Internet. However, playing to its advantage is that it now owns a patent on its ranking technology.

On the right, you’ve got Overture, which once had a business that most people figured would fail (after all, who would pay for a placement in a search engine?) but somehow managed to prove the naysayers wrong and is now going out and buying itself a new seat at the search table.

At stake is the future of search but it may be much more. It looks like the market is reshaping itself to become not just about search but also about targeting. Give the right search result, attach the right ad, and all of a sudden you’ve got a redefinition of online advertising networks. The funny thing is that we’ve been there before. This was what Doubleclick promised but eventually abandoned due to privacy concerns. I suspect that the search engines are headed in a similar direction and that history might repeat itself here (and I suspect that we may see more people starting to worry about privacy issues similar to those that plagued Doubleclick a few years ago.

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title History repeats itself. You can follow him on twitter here or receive his weekly newsletter by subscribing here.

What I came to realize is that what works for a computer does not necessarily work in a wireless environment. The main issue is input and output. A wireless Palm is great to get information but somewhat difficult to use to send out email (typing in graffiti being the biggest challenge so far) and WAP works well to get little bits and pieces of information but is limited to a set number of characters (depending on which version of WAP you’re using, you will get an allocation of between 1500 and 2000 characters). As a result, sending out something like this newsletter over WAP does not seem to make sense.

However, a new breed of services is now popping up and it could be the next big thing: connecting to the Internet by just dialing into a phone number. In order to test this out, I checked out several services: Tellme Networks, which launched last month to a barrage of publicity, and Quack, which has been quieter but is also launched. Other contender for the space (and all of those are launching “soon”) are Internet Speech, BeVocal, and 888TelSurf.

Tellme Networks

Mike McCue is no stranger to bleeding edge technology. I met Mike back in 1995, when he was running a company called Paper Software. Paper did VRML browser, at a time when VRML was not only cool but also seen as the potential future of the net. Netscape acquired his company and Mike ended up as VP of technology there. While there, he looked at the future of the Internet.

However, as many other Netscapees, Mike ended up leaving Netscape and founded a new company: Tellme Networks.

Their goal: to make the web as easily accessible as using a phone. Pulling talent from both sides of the browser war (Netscape and Microsoft), Tellme was a very secretive operations until a couple of months ago, when it announced its offering: a phone service that gives you access to news, weather, sports, stocks, movies and restaurant info.

An interesting concept but how well did it hold up? Would I be stuck in one of those horrible menus (press 1 to access news, press 2 to access stocks…) or would it be better, I wondered as I registered for the service. To my surprise, it worked very well.

There are a number of interesting features in this service. First of all is the text to speech engine, which is one of the best ones I’ve heard so far. It takes bits of information and relays them in a clear and understandable way, which I found quite amazing. The voice recognition algorithm also held up very well, clearly understanding what I was saying (for those of you who’ve heard me speak, you know I have a fairly strong French accent, which makes this all the more amazing).

But what about the selection? Well it is, in one word, adequate. Not great but not bad either. I was able to get stock quotes (an interesting feature allows you to use the touchpad to spell out a stock symbol if the service does not recognize the name of the company you gave it), grab the latest news from CNN (the only general news provider tellme offers), pick up the weather in New York, check out a movie schedule, get some restaurant info and be connected to the restaurant for reservations. Also interesting was the phonebooth, a feature that allows you to make 2 minutes phone calls anywhere in the US for free (brought to you by AT&T). All and all, I was impressed enough to save the number into my cell phone memory.

Quack

Quack offers basically the same things as tellme (with the exception of the phone booth) but requires you to set up personalization largely through the web. As a result, the service may seem a little kludgier. On the plus side, however, the voice recognition algorithm gets some points. For example, I requested stock price on Internet.com on both services. Tellme was unable to find Internet.com and asked me to enter in the stock symbol using the keypad. Quack immediately recognized the stock, gave me its symbol and, while playing ads, went on to retrieve the information. What was disappointing, however, was that their text to speech engine is not as advanced as tellme’s. The voice on the other end of the line was unmistakingly a computer when it came to particular bits of information (weather reports, stock quotes) compared to the syntax used by tellme.

Another plus for Quack is their caller-ID system: The service also recognized that my call was coming from New York and immediately gave me the weather report for that city (I had to request New York on tellme’s service).

All and all, I found very little difference between the two services, with the possible exception of Quack carrying traffic reports (which tellme did not offer when I tried it). However, since I am one of those New Yorkers without a car, traffic reports held little value for me. If I were outside of New York, I am sure that this service could be a godsend.

Looking at the future: the other contenders

Internetspeech and bevocal seem to be aimed at the same market. They both plan to offer similar services. InternetSpeech seems to go a step farther by trying to capture an audience that will be able to say a URL and get what’s on that page read to them. This could be interesting but can get kludgey as the example on their site shows. Going to Yahoo means that their service will read the links but also shows that it doesn’t know how to stop on a particular link.

However, a somewhat more promising concept is that of 888Telsurf, which plans to not only offer information but also access to an online calendar and address book as well as over the phone reading of emails. Advertising revenues will finance their service, like the others. Unfortunately, none of those services are available now nor do they have announced launch dates.

Where is it all going?

Obviously, those services do not plan to stand still. Each of them is trying to establish a beachfront as the new “portal” to the Internet. I personally believe that a lot of those will end up either being acquired or striking significant relationships with the current group of already existing portals. Ultimately, all of them will have to follow the lead of 888telsurf by offering email, address book and calendars as part of their standard package.

While they are all squarely aimed at the consumer market, I think that at least one of them will break away and start offering the technology as a software offering, allowing corporations to set up private intranets with dial in interfaces. For example, one could be able to call into the main office to get part numbers on something they are selling, or check the latest sales statistics for a particular region.

Another potential use for the technology developed here is in the Ecommerce arena. Why not use that technology to offer a call-in service for ordering. Wanna get that new book from Amazon.com? Why not call 1–800-AMAZON (or whatever their number will be) and order it via phone. This could go over the scare hurdle that some people still have about online commerce.Or at least, it will give it a new dimension.

The other thing that could help those offerings become more popular would be the integration of comparative shopping features. I can see myself in a store, looking at prices on a particular item I want to buy, then pulling out my phone to check if someone will offer the same item for less.

Either way they play it, I think that this group of companies will present the first serious challenge to WAP, unless more content providers start offering WAP-enabled content (at current time, the selection is fairly weak).

Tristan Louis is the founder and CEO of Keepskor and writes the influential tnl.net weblog, where this was initially posted under the title Challenging WAP?. You can follow him on twitter here or receive his weekly newsletter by subscribing here.