Google Wave: The Jesus platform of communication?

E-mail is like your normal post office mail. Send-wait-expect reply-get reply-start again. Only that the broking agent is electronic.

Instant messaging is like having a dots-and-dashes telegraphic conversation. Except that it happens over a digital network and in normal language.

Desktop sharing is watching your elder sister paint. Watch but don’t dare to touch.

Twitter is like sitting on the edge of an ocean. 140 character waves continue to hit the shores relentlessly. You sometimes walk and dip your feet in the water. Then you retreat and watch the tides again.

Facebook is like standing at the intersection of Broadway and 42nd Street (Times Square, that is). You watch different people, you hear their conversations. Sometimes a friend passes by and you have a short conversation with them. Often you mutter something that a passing friend picks up and responds to. Or she goes and tells that to someone else.

Zoho and Google docs are collaborative tools. Like 5 people standing at a giant white board, all editing the same item synchronously.

These are all examples of human behavior. Rather that of the human desire of communication. From the dawns of civilization 13,000 years back we have been interacting with each other for multitude of reasons. Unfortunately while the needs and the reasons for communication were always around, society found productivity solutions to them mostly based on priority. Once a need became a pain, people rushed to solve it. The next need either got solved independently or in some rare cases smart people discovered ways of interconnecting metaphors and extended a solution to solve another need. And this is why we have the list – which incidentally is way incomplete – of multiple communication solutions above (and what they borrowed from).

The inevitable had to happen. Someone either had to come and aggregate as much of the behaviour as possible (Facebook does a decent job but it is like a patchwork, not fundamentally changing any of the collated paradigms) or someone had to create the Supermarket of Communications. Google Wave (please watch the embedded video. 80 mins but worth every second) does the second – and does it spectacularly.

The first temptation of an engineer looking to amalgamate multiple paradigms of communication is to integrate them in their native state – a straightforward nice wrapper over the pieces (put GTalk into GMail. On selecting a conversation, open up a widget that shows collaborative docs between participants, and so on). Google Waves not only successfully stayed away from this approach but they re-engineered every paradigm they touched. E-mail is not like e-mail in Google Waves and so isn’t instant messaging and link sharing and photo uploading and wikis and document collaboration. The interconnection of these elements in human nature renders them just as parts of a larger whole. And Google Wave engineers them exactly like that.

Google went a step further to make all this open source. A vast population of developers will now develop extensions to Google Wave using their APIs. It is like how your i-phone apps make your phone much better than how you bought it from the AT&T Store. I sincerely hope that Google pushes ahead with Waves and while it does also thinks of how to monetize this initiative. The reason why Google’s search technology has improved leaps and bounds is because it is heavily monetized, unlike say Docs or Blogger. A build-first-market-later is typical of an engineering led company and quite the opposite of what the gurus preach, but it will be nothing short of a debacle if Google slows the momentum down on Google Waves.

Walpha plus Google: Best of both worlds

Wolfram|Alpha started dating Google sooner than I expected. This Firefox extension allows you to commingle Wolfram|Alpha’s output against a Google search retrieval. The Brin-Page smartness and the gumption of Stephen Wolfram – heady cocktail. I threw the same Bob Dylan rehtoric question at this newly dating couple and here is what it returned – each doing their bit to perfectionWA_G

PS: Walpha is sluggish than the big G. The latter fetched about half a million results in 0.34 seconds while Walpha took an additional whopping 15 seconds to display its wares.

PPS: Contrary to my expectations, Walpha hasn’t yet figured out “how many bags of wool does ba ba black sheep have”. Big G dug up some 6,000 answers.

Question: If Big G and Walpha were real life couples, how do you think they’d split the household work?

Search Engines versus Wolfram|Alpha

In the year 1978, film director, musician and author of stories for children, Satyajit Ray wrote a short story named “Compu”. Scientists had invented the ultimate computer that could answer any human question. The I/O metaphor was voice and the twist in the story came when “Compu” – the ultimate computer – started getting a mind of its own and behaved like a human being. Twenty one years later we can – with the benefit of hindsight – say that this is like putting Wikipedia into a voice enabled computer with semantic web understanding thrown in, but the story was quite ahead of its time for 1978.

The quest for understanding human queries and having a computer responding back has kept the scientific fraternity busy. However, as content in the web proliferated, the obvious solution to this problem was to let the web answer the question rather than creating an engine. This led up to search engines and focus shifted to making search results more and more relevant (with of course an eye on the ad revenue pie). The trouble with search is that it points to a source rather than answering the question directly. While that is perhaps a smarter strategy, it is less fun.

So in this cauldroun appears Wolfram|Alpha, the quantitative knowledge engine from Stephen Wolfram. All it does is “to make all systematic knowledge immediately computable by anyone”. It is much less smarter in handling non-computable stuff because it has a bravado (bordering on hubris) that makes it do its own stuff – not fetch results from the internet to read. So it is much more fun (it is also secretly sponging up all those questions it cannot answer today – like “how many bags of wool did ba ba black sheep have” and is figuring out answers to it as you read this). Walpha (affectionate nickname) puts forth an entirely different way of unearthing knowledge – so long as it has to do with computation. I would be keen to see if it throws up alternative monetization models, deviating away from the traditional advertisement based ones.

PS: People are discovering interesting aspects of the engine displaying strong character and a funny bone (“Compu”?). It has a taste for good music too as I found out

Will There Be Indian Software Companies Anytime Soon?

Why aren’t there more global software product companies out of India? Basab Pradhan examines

expertszonA few years ago at Infosys I used to regularly get the question – why isn’t Infosys doing more in software products? In response I wrote this piece for Rediff where I addressed the question of why Indian services companies don’t do well at products and why they shouldn’t even bother. Since then Infosys-incubated OnMobile has done quite well, but that in no way disproves my claim that the two business models – products and services – are so different that they don’t belong together (OnMobile was a separate company well before it found its groove).

Lately I’ve been getting a different question – why aren’t there more global software product companies out of India? (I use ‘product company’ as short hand for any information technology IP based company). Clearly, it isn’t for lack of software engineering talent.

The single most important reason is the lack of a domestic market for software. India’s market for business or personal software is very small market compared to developed markets. And you can’t build software products away from the market.

Designing software products requires a kind of iteration with end users that just isn’t possible unless you are ‘just across the street’ from your end user. Demos, prototypes, white-boarding with users who are willing to give of their time is what creates a good product design. Something that just can’t be achieved by using tools and methods, however sophisticated, in an offshore model.

This, by the way, is true for all product development, not just in software. You have to first create a customer base in your home market that can then become the springboard for international expansion. The auto industry is a global industry and to my mind there is no company that started making cars for a foreign market before establishing a beachhead at home.

Why is India’s software market so small? Partly due to the size of the economy and partly because average income is so low that productivity gains due to adoption of technology don’t  pay off unless that cost of the software is really low. This is true about all emerging markets.

The software products market is dominated by American software companies. As the largest integrated market by far, the US has a great advantage here. But that is not the only reason. The most competitive companies in an industry tend to cluster, as Paul Krugman showed us, and areas like the Silicon Valley and Boston are the industry clusters for software.

Can Indian tech companies succeed? They can – by first serving needs in the local market that are not served by global software companies. Such needs could be underserved because the global players find the market too small and the prices too low to make changes to their software or because the market need is unique. The price that would work for say a small business ERP in India may not be something that any US ERP vendor is interested in.

Pricing, or ‘the bottom of the pyramid’ path is a very plausible path to success especially with business software. The world over, businesses will be able to afford computers before people at home do. Small businesses in third world countries, very much like in India, don’t like to pay (too much) for software. But you do need business software to scale your business. A successful Indian business software company will know how to make a profit at price points where big software companies won’t.

The competition at the bottom of the pyramid is going to come from Open Source. Open Source is slowly creeping up the stack. There are even a couple of Open Source ERP names. But perhaps Open Source is an opportunity itself. Perhaps Indian companies should use Open Source to play the IP game. I have yet to see much happen on this front although this could have been closer to the knitting of the Services companies. Maybe we will soon.

About the author

BasabPhotoHomepageBasab Pradhan is formerly CEO of Gridstone Research. Basab blogs at

Don’t ignore the User’s environment

Behavior is a function of the environment. It is also a function of participants in the environment and a complex psychographic intertwining of the participants. Social networking is a great example. Corporates intranets have matured to varying degrees from being a storehouse of company policies to platforms of information dissemination. Some intranets, especially those from multinational, multi-disciplinary firms, are warming up to the fact that the workplace is after all a social environment with its own needs making effective connections. At this point, the trapdoor of Facebook-meets-corporate-intranet opens up, little realizing that the environments and participant psychology in the two situations are vastly different. And hence the need to cater to different environment driven networking demands.

I have witnessed this while building capital markets related products. Given the democratization of these markets, it is very likely that a Product Manager, Engineer, Tester – all – have particpated in it in some form or fashion. The choices they make during product development is often clouded with their own behavior. The fact that retail participation in capital market follows an entirely different workflow compared to institutional investing is overlooked.

Managing overlapping environments is an interesting challenge because it is not axiomatic that what I have mentioned above must always hold good. Take for example the building a professional software developer Q&A community (must watch – Joel Spolsky talking at Google about Stackoverflow). It is best that software developers just meditate over their own professional trouble-shooting pain points and build a system that effectively addresses them. Sadly, such examples are fewer and far between. And the challenge remains to carefully understand the users along with the environment they would be working in. Investigating in isolation either of the two will most likely result in this

The broken world of end-user training

It has been a while that I promised to write about Product Training. I was interacting with the Client Training group in my company when I remembered the promise and also realized how the current model is broken.

The trouble exists both at the supply end of the chain and at the demand end. Let’s see. Like not too many poeple graduate wanting to become teachers, getting talented people to join client training is difficult. In most organizations Client Training is considered a cost center, which means it is pernnially under pressure to keep costs low. Consequently one trainer has to develop multi-product (or multi-module) skills. Now this thing works fine in primary school (geography teacher also teaches history) but in the professional world is a disaster. Training the trainers is mostly inadequate with the product teams’ unrelenting focus on shipping the product and then getting busy for either the next round or the sexy stuff like wine-and-cheese launch parties. 

On the demand side – at least in India – user attrition is a big problem. Teams of users leave and a new set comes in. The new set has experience of using Product B so now the company that sells Product A needs to come in and retrain. The problem exacerbates if the industry has many products catering to roughly  the same user group. 

Some organizations outsource training arguing the not-our-core-skill theory. The decision is fine as a standalone but when you look at the intricate interconnections between Product Management, Product Marketing, Market Intel and Training a big chunk of the intercdependant relevance goes away with the outsourcing. Some companies do a smart thing by imposing switching costs for their products in the form of user certifications. Unfortunately the execution of this strategy also ends up suffering from the same supply side problems I mentioned earlier.

Consider keeping training departments of product centric firms close to Product Management. Over a period of time, all training that has to do with historical and existing functionality should be moved from physical contact to digital delivery. Once that happens, the training module can actually be placed within the product and perhaps at a cost (That F1-for-help thing is actually training too, in some form). Alternatively, there are many different ways to deliver digital content to the user fraternity (yes, an i-phone version is downright innovative and outright sexy). At the end of the day the idea is simple – improve the content of the training and move away from traditional physical contact method of delivery.

Got other ideas? Let’s hear them.