Rohan Jayasekera's thoughts on the evolving use of computers -- and the resulting effects

Occasional thoughts by Rohan Jayasekera of Toronto, Canada.

My Photo
Location: Toronto, Ontario, Canada

I've been online since 1971 and I like to smoothe the way for everyone else. Among other things I co-founded Sympatico, the world's first easy-to-use Internet service (and Canada's largest).

View Rohan Jayasekera's profile on LinkedIn Rohan Jayasekera's Facebook profile twitter / RohanSJ
Subscribe in a reader

Or enter your email address:

Powered by FeedBlitz

Friday, November 24, 2006

2.0nd chance

I didn’t blog earlier about the the Usability Camp held in Toronto on Nov. 14, 2006, World Usability Day. The organizers, Mira Jelic, Jyotika Malhotra, Richard McCann and Ilona Posner, were astute enough to invite as one of the speakers Michelle Ivankovic, Senior Designer at Umbra. Umbra is Toronto’s own “worldwide leader in casual, contemporary, affordable design for the home” and I’ve been a fan of their products for at least 15 years, starting with an indoor garbage bin that I still use (since joined by a second one for recyclables). Michelle spoke about how they improve the usability of everyday items such as their holder for paper towels. I found this highly relevant to the event, as a reminder of how using a computer is an everyday thing just like using a paper towel holder, so that usability is just as important.

I find it shameful that the computing industry has done such a poor job of usability. Maybe Web 2.0 is our 2.0nd chance.

Digg, Over-Extrapolation, and Monoculture 2.0

I just read So What Can Netscape Do Now? Here Are Four Options. in Deep Jive Interests. Tony Hung understands that Jason Calacanis threw away an existing business, the Netscape portal, by changing the existing product to a new one (a Digg-alike) that did not appeal to the customer base. He didn’t even get the Digg crowd (a much smaller group than the existing Netscape users) to switch. (Dr. Tony thinks first-mover advantage is very difficult to overcome, but it all depends. Remember Friendster.)

What would cause Jason Calacanis, a very smart person, to make such a huge mistake?

Bad assumptions/axioms.

My guess is that he believed that (a) what the cool kids were doing would spread to the larger population, and (b) Digg users were the cool kids. If so, he was far from the only person in the Web 2.0 world to have these beliefs; many have drunk of the Kool-Aid.

(b) first. Digg users are not the cool kids. For one thing, they’re overwhelmingly male, and in addition I’d bet that overall they’re pretty dorky. They don’t seem to be interested in what other cool kids are, like clothes. Meanwhile, one of the front-page stories when I checked just now was “Having a clue as the main requirement for open source”.

As for (a), cool kids are herd animals more than the rest of the population. Popular nightclub? Not for long: a newer one will become cool and everyone will go there instead. (Which is why Ken Schafer and Ross Rader predicted that MySpace would lose its #1 spot.) Fortunately for the general economy, not everyone makes purchasing decisions based on what is most popular at this very moment. And not everyone wants their news items selected that way either.

To be fair, there is also reddit, which has a slightly older demographic, and furthermore tries to deliver items of interest to the particular reader by checking popularity among readers with similar histories. But, so far anyway, it’s not all that different from Digg.

Is the Digg approach viable within communities that are less skewed? I think Jason Calacanis thought it would be, hence the reinvented Netscape. That he failed does not prove that it can’t be done, so let’s watch and see.

More for the Monoculture 2.0 file.

Wednesday, November 22, 2006

Product Ownership

Leila Boujnane points to a TechCrunch item that reproduces an internal Yahoo! memo by Brad Garlinghouse, Senior Vice President - Communications, Communities, and Front Doors.

In the memo, Mr. Garlinghouse asks the question
Equally problematic, at what point in the organization does someone really OWN the success of their product or service or feature? Product, marketing, engineering, corporate strategy, financial operations… there are so many people in charge (or believe that they are in charge) that it’s not clear if anyone is in charge. This forces decisions to be pushed up - rather than down. It forces decisions by committee or consensus and discourages the innovators from breaking the mold… thinking outside the box.
This situation exists in lots of organizations, with the same consequences. In a company that makes its living selling one or more products (by which I mean to cover all of “product or service or feature” — see my earlier post What is a "product"?), multiple people have a legitimate claim to make decisions about a product. If a product isn’t marketed properly, it won’t sell. If it isn’t built properly, it may not sell — or if it does, maybe not for long. If the financials aren’t right, it may sell — but lose money.

Mr. Garlinghouse’s solution, which in my moderately long career is the only one I’ve seen to be successful, is to have exactly one person in charge of each product, with no overlap. The marketing, engineering, financial operations, etc. people for that product all report to that one person.

I believe in this structure so strongly that once I even eliminated my own job in order to make it come about. I had been in charge of the software development for four different products and helped bring about a reorganization that put everyone working on a particular product into their own group, software developers included. Fortunately the company kept me around until I managed to spot a need for my services elsewhere within the organization.

What does it mean for one person to be “in charge”? I want to draw the distinction between “delegating” and “assigning tasks”. If you delegate something to me, that means that you trust me to do it to your satisfaction. You go about your day assuming that I will do the job. If I screw it up, you will be very disappointed: you trusted me, and now you have to deal with whatever damage I caused. In contrast, if you merely assign tasks to me, you’re not trusting me: you don’t let go, and expect to oversee what I do. If anything goes wrong you may blame me, but you must also blame yourself, because you were the one in charge. So: if you delegate to me, you put me in charge; if you merely assign tasks to me, you stay in charge.

At Yahoo, Mr. Garlinghouse would have a bunch of people called General Managers, and put each one in charge of an area comprising a group of related products (because Yahoo has tons of products). That is, he would delegate his authority down to them. They would make the decisions for their areas. If Mr. Garlinghouse didn’t like a particular decision, he could overrule it, but he wouldn’t do that too often or he’d undermine the authority that he’d supposedly delegated.

But what about smaller companies?

I believe that the same role needs to be filled, though usually with a title other than General Manager. At a single-product company, the person in charge of the product has to be either the CEO or the COO. Nobody less will do, because the one product is the only thing the company has to sell. Another person can be named Product Manager, but nobody should be expecting that person to be the one in charge of the product. It’s got to be either the CEO or the COO (and not both).

If the company has more than one product, then it becomes reasonable for each product to be the responsibility of a different person. That doesn’t necessarily make it desirable: if a company has only two products, for instance, it may very well still make sense for either the CEO or the COO to keep the responsibility. But if the company is sizable and has many products, it becomes impossible for one person to handle all products, and delegating product responsibility is essential. This is when each product needs one and only one person in charge of it. The obvious title for that person is Product Manager, but it’s the role that is crucial and not the title.

Monday, November 20, 2006


This evening’s DemoCampToronto11 was notable in its failure: I’ve been to most of the previous ones, and the vast majority of the presentations have been decent. Not so tonight: of the five presentations, there was only one (Sunir Shah on “Design Bibliography”) that I’d consider a good one. The others:

-AutoSSL was certainly a good topic: an original Internet product being created by the presenters themselves. But the presentation was a combination of a demo (fine) and an attempt at a marketing-type presentation (not fine). There were objections from the audience to the diagram shown, which although not done in PowerPoint seemed to violate the spirit of the “no PowerPoint” rule of DemoCamp. I think the idea of the rule is that you should show the thing itself, not show something about it, and the audience made its unhappiness known. (Interestingly enough, the demo itself failed to work but this didn’t seem to bother the audience particularly. “Just give us a demo” — even if it fails!)

-Firestoker was a similar lost opportunity, again an original product improperly demoed. Every demo needs this to be explained: what is it? I had to figure out what Firestoker was by looking at a bunch of screenshots.

-Andrew Reynolds’ demo of the Selenium test tool was well done, but he was just demoing something he’s a user of (not builder of), and already available to all.

-My Studio Assistant: Arnold Wytenburg showed something that I have no doubt would be useful to many artists, but it’s just a website builder for a particular category of business, like so many others. Nothing particularly unusual or interesting about it.

Solutions? The DemoCamp organizers are already on the case, Jay Goldman having announced at the end of the event that they’d be looking at “refactoring” DemoCamp, and inviting comments. Here is my first thought:

There is already something in place at DemoCamp to address the what-is-it question, in that every presenter starts off by answering four questions: Who are you? What will you be showing us? What do you hope to give the community? What would you like back from the community? But it’s not working. One modified approach might be to replace these questions by a recommended (or required) structure for presenters to follow, something like this (where “X” is the thing being demoed):

1. Say who you are.

2. Set the context. Say who would use this. And if X is meant to solve a certain problem, say what that problem is. (The “pain statement”. See this video about elevator pitches by Sean Wise, also starring my former Sympatico colleague Peter Evans in a non-speaking role.)

3. Give the demo together with an explanation of what is being demoed. If a problem is being solved, say how X solves that problem. If not, say what X is, and how it’s useful or interesting or otherwise worthy of the audience’s attention.

4. Give any commentary on implications / uncertainties / future plans / whatever. It is crucial that this come after the demo.

(Repeat step 2-4 as many times as desired: sometimes there are multiple features or multiple uses.)

5. Make any offers to, and requests of, the community.

6. Take questions.

Tuesday, November 14, 2006

Today is World Usability Day

Today, November 14, is World Usability Day. From its charter, the following message, which I endorse:

Technology today is too hard to use. A cell phone should be as easy to use as a doorknob. In order to humanize a world that uses technology as an infrastructure for education, healthcare, government, communication, entertainment, work, and other areas, we must agree to develop technologies in a way that serves people first.

Technology should enhance our lives, not add
to our stress or cause danger through poor design or poor quality. It is our duty to ensure that this technology is effective, efficient, satisfying and reliable, and that it is usable by all people. This is particularly important for people with disabilities, because technology can enhance their lives, letting them fully participate in work, social and civic experiences.

Human error is a misnomer. Technology
should be developed knowing that human beings have certain limitations. Human error will occur if technology is not both easy to use and easy to understand. We need to reduce human error that results from bad design.

We believe a united, coordinated effort is
needed to develop reliable, easy-to-use technology to serve people in all aspects of their lives, including education, health, government, privacy, communications, work and entertainment. We must put people at the center of design, beginning with their needs and wants, and resulting in technology that benefits all of us.

Therefore, we, the undersigned, agree to
work together to design technology that helps human beings truly realize their potential, so that we can create a better world for ourselves and future generations.

We agree to observe World Usability Day
each year, to provide a single worldwide day of events around the world that brings together communities of professional, industrial, educational, citizen and governmental groups for our common objective: to ensure that technology helps people live to their full potential and helps create a better world for all citizens everywhere.

There’s more, but it detracts from the message. Also, I am shocked that I had to fix a number of typographical errors, since such errors reduce usability and yet this text came from the “Usability Professionals Association”! Perhaps eternal vigilance is the price, not only of liberty, but of usability as well. I encourage you to speak up whenever you see poor usability and have some way of telling those who could improve it.

Sunday, November 12, 2006

Mashups schmashups

I had hoped that by now people would have stopped talking about mashups as one of the great Web 2.0 things, but they’re still doing it. John Markoff wrote today in the New York Times that
The classic example of the Web 2.0 era is the “mash-up” — for example, connecting a rental-housing Web site with Google Maps to create a new, more useful service that automatically shows the location of each rental listing.
The current Wikipedia definition of a mashup is “a website or web application that seamlessly combines content from more than one source into an integrated experience.” But most mashups, including the canonical example above, are nothing but the use of a mapping engine to graphically display locations. Just as I told that Markoff’s sentence above was a quote, so that it would be displayed in a suitable way, location data is often usefully displayed in map format, and that’s all there is to it. That there is also content included from another source (such as the locations of streets) is incidental: if the mapping engine used a snapshot of street information, instead of pulling it from a database, in most cases it would be just as useful.

When people supposely use Google Maps to map those rental listings, they’re not actually using the Google Maps that many of us know and love. They're using the API that Google chose to make available to programmers at large. There is no actual “mashup” of two websites. Instead there is the use of a mapping utility that is “called” (as we programmers would say). This is really no different from, say, a web page counter utility that shows at the bottom of your web page how many people have looked at it, e.g. . You don’t think of a site that has a page counter as “Web 2.0”, do you?

Friday, November 10, 2006

The Web 2.0 monoculture

I just have to speak up about this, because not only do I believe it’s bad for Web 2.0 but it’s proving to be bad for me personally. The Web 2.0 world is a monoculture: a system that has very low diversity.

As I began this post at the end of October (yes, I’ve been meaning to finish it for a little while) I checked for who else had written about this. It turned out that this was actually a pretty current topic and I was behind on my blog-reading, e.g. on October 16 Anil Dash posted Life or Death for Web 2.0, which I highly recommend to gain some understanding of the problem.

In his post, Dash refers to an article in that day’s Washington Post by Shankar Vedantam, Why Everyone You Know Thinks the Same as You, a scary article for those of us who are used to thinking of the Internet as a positive force for connecting people. Scary but I highly recommend reading it too (it isn’t long).

What do these names have in common: Anil Dash, Shankar Vedantam, Rohan Jayasekera? They’re all South Asian (Dash and I were born and raised in North America, while Vedantam moved from India to the USA in his early 20s). Perhaps there is something worth looking into here: are those of us who are North American, yet strongly influenced by a different cultural background, either more sensitive to or more inclined to be concerned about a perceived lack of thought diversity?

Note that I said “thought diversity”. Others have already expressed their concern about, in particular, a lack of diversity in ethnicity and gender. Sure, that’s an issue, but I think that will work itself out over time. Other types of insularity exist too, though, and get less recognition.

I co-founded Sympatico (Canada's largest ISP), which as far as I know was the world's first mass-market Internet service. One of our portal content people remarked to me one day that she was hoping for more diversity among us. I agreed, but we were thinking different things. She wanted people with physical disabilities, aboriginal people, etc. I wanted right-wingers.

Our group was pretty lefty, and this was not representative of the population we were serving. I recall being horrified when, long ago, I tried out Yahoo’s chatrooms and discovered that they were largely occupied by “trailer trash”. But if that’s the real population, let’s do our best to serve it.

Later I happened to hire someone who didn’t come from that lefty thoughtplace and in fact wore a mullet. Of course I hired him purely because he was really good, but getting more thought diversity was a very nice bonus as far as I was concerned.

Let’s talk about Apple. A lot of people in the 2.0 world love Apple and its products, and regard Microsoft with suspicion if not outright hatred. These same people profess to love open source and dislike monopolies — yet they ignore Apple’s closed-architecture approach with the Macintosh and the iPod, which may well have a lot to do with Apple’s success. As I see it, Apple is far more monopolistic than Microsoft is these days. Is it such a good thing that the Web 2.0 community is so Mac-heavy when this is so unrepresentative of the Internet-using population?

Then there’s age. I’m over 40 and in the Toronto geek community that means I’m a foreigner. (When I was a kid I would never have predicted that my skin colour would cease to be a problem and my age would become one.) I think there are two reasons: others think they can’t relate to me because I’m weird (they don’t know anyone else like me), and because this is a tech community they assume that I’m out of date (ironic given that they’re now rediscovering things from the 1970s, as I wrote earlier).

Music? Apparently the best musicians all have names starting with “DJ”. So much for originality in names, let alone content (if you remix something that someone else did, in the 2.0 world you somehow get more kudos than the original artist).

Back to Anil Dash, Shankar Vedantam and me. It seems that we have a shared inclination to spread things to the general population, which I don’t think is the case among most Web 2.0 people, who want to spread things to each other. I take great pride in having helped spread the Internet to the masses in Canada (and having set an example for similarly-minded people in other countries). Dash encourages the masses to use spreadsheets for all kinds of things, as he does. Vedantam says that “I moved into science journalism because it seemed like a very neat way to capitalize on knowledge that I already had. In modern culture there is a huge knowledge base, but we haven’t spread that knowledge very well. I feel that spreading it may be as important as generating it or acquiring it.”

So if you believe that Web 2.0 has monocultural problems, do your best to diversify its population — and not just in ethnicity and gender. Start by not thumbing your nose at anyone who uses Microsoft Windows, votes Republican/Conservative/equivalent, is over 40, or likes to listen to Shakira.

Thursday, November 09, 2006

Web 2.0 and IT departments

A couple of days ago I attended Toronto’s second Enterprise 2.0 Camp. My thanks to Tom Purves for organizing this, and to Firestoker and PROFIT Magazine for sponsoring.

I’m not going to blog about the event in general, but a comment Tom made during his presentation is something I’d like to talk about. Tom remarked that IT departments like Web-based applications because they’re easier to support: no need to install or update software on all those PCs. I’ve heard this comment before, and I can well believe it.

But are the IT departments doing themselves out of jobs? Once might respond that applications still need to be installed and maintained on corporate servers. But more and more the enterprise is becoming dependent on Web-based services outside the firewall. Knowledge workers depend on Google and, for instance, and as other services become available that are helpful for their jobs they’ll use them too. The IT department might install something locally, but probably not for a while, and even then why not just use the application that’s available to you even when you’re not in the office? (Sure, there are VPNs, but using them is an inconvenience.) Also, if you’re working on a project with vendors or other partners, it’s easiest to use applications that aren’t behind anyone’s firewall.

In the late 1970s I worked for what was then called a timesharing company, I.P. Sharp Associates. We sold access to applications that ran on our servers (as they would now be called), and because the public Internet didn’t yet exist we had our own far-reaching network (supposedly the world’s largest private packet network). A lot of our business came from corporate users who wanted to get things done and couldn’t wait for their always-backlogged IT departments to provide what was needed. So in effect we helped them do an end run around their IT departments. As business-oriented applications increasingly become available on the Web, this will happen again, but on a tremendously larger scale.

Score another point for Nicholas Carr’s Does IT Matter?

UPDATE: since I posted the above, Business 2.0 Magazine has written a similar story. You can read it here.