Big Data is about giving a wow effect about your service to your users

I am not trying to give the usual definition of what is Big Data as there are umpteen definitions already. 

Big data is a new wave [that has been used by many companies for so many years already:-) ] to make existing proceses and applications smarter :-)

It is like the smartphone revolution that we have already seen. Most applications have a smartphone interface because it makes the application smarter and improves user productivity.

Look at what Facebook kind of services itself did with your personal data. They suggest friends and bring in people you would not have thought of as friends. This is increasing the expectations of the business on application teams. It is driving companies to use their existing data and make their applications smarter, in the end help them make better business decisions.

Here is another example. Say you have moved in to a new city and are not familiar with the train timings. However your friend or a friend's friend, lives in the same city. She uses the "Train Station finder App" which she really likes. For you to know  about the App, in the traditional way, would normally have been via newsletters or advertisements of the Train Station itself. But now, with the data that facebook kind of service has, they can now suggest that you use it since you just moved in to the same city. Now when an app marketplace suggests  or advertises that to you, you are happy as you think that is relevant news for you.

And now with mobility there is a lot more data. Location. Social. Mobile. Tablet. Ways for an application to be human friendly. Lots more unstructured data that needs custom rules/queries applied. Incidentally that means, terabytes of data that need newer technologies to store and process for meaningful consumption.


In other words, Big Data is about making your applications give a wow effect to your user.

Empathy with experience vs apathy being best of breed

Our HR team asked us to read the book Get rid of the performance review, by Samuel A. Culbert.  That book has a lot of bad mouthing of existing HR practices :-) BTW I will post my review about the book in another post.

Now getting back to my observation while reading that book which was also triggered by a discussion with a family friend who is in the HR team of a large IT Services company.


Hiring an experienced HR manager who does not understand your company vs promoting someone who has been an employee all along as HR manager has a huge difference. The first one knows industry best practices and knows how things are in other companies. He ends up being more mean and will keep the company's prospects foremost.

However the one that is promoted from down the ranks, understands the nuances of the company culture. The later one may lack knowledge of the best practices followed in the industry. However his ability to work with the "HRed" is surely better and can help create a conducive environment that helps build a great company.

So if you want to be a people friendly company, then you know what you should do.

But more importantly weigh your options well when choosing between experience over best of breed.

$100 Million niches normally exist in a $1 Billion Software Market

Product feature prioritization is a constant challenge that product managers face. Now when you focus on a huge market opportunity, focus first on a niche segment.

This is especially necessary if you want to keep your marketing and support costs low.

Take Project Management as an example.

Project management tools are needed by development teams.

Project Management is needed when consultants do small projects for hundreds of clients over the year. This is another niche.

Small teams (less than 5) working as a close knit group would want to track common tasks they work on. This could be further segmented by considering that the potential users of your tool could be working from remote locations.

To make things more clear, if there are people working remotely, then chat integration, desktop sharing could be important than say a Gantt chart which is more useful for very complex projects with tens of people. If you have an independent consultant doing many small  projects a year, pricing per project will make it expensive for these small customers.

Identifying the segments is very critical.

Addressing it one after the other is the best strategy which is like addressing the $100 million niche before taking on the $1 billion market. Probably that is a hint on how you could prioritize your feature requests.

I may be generalizing an observation a bit, but I hope you get the drift.

Ride a technology wave; Be first to market with your product

If your marketing strategy is to be the first in the market and being seen as an innovator, time to market is very critical.

Now that is obvious. However ask yourself, does your product release strategy reflect that? Or are you wavering in your decision about a market potential and later on losing the first mover advantage? Some times its ok to put a lot of your eggs in to one basket when the stakes look juicy.

Lets take an example. Being an innovator means, riding a new technology wave like Ruby on Rails or Node.js or Cloud or iOS and building a product that complements that ecosystem.

To make this marketing strategy work, here are two important aspects.

First release of the product can be ugly code! But go to market first or very very fast. You could rewrite your release 2.0 to suite your ego.

Some companies have a strategy of releasing lots of point tools instead of massive integrated suites. This is good as a customer who needs only a small piece of the suite capability can buy that specific piece. Its also an easier sale. Lower cost sale. Solving a specific problem well. And most importantly, free cross  selling for your "mothership" - integrated product suite.

Now let me get back to the second aspect. If you are confused on whether what you do has to be released in to the market as an "add on" to your existing product, or be released as a separate product, do both. But first, release as a separate product. Especially if the capability is credible enough. If you did not understand the second point, read the previous paragraph again.

One mistake we can make is, put our best people on routine stuff and miss a market opportunity completely.

Accessibility of your application. What interfaces?

If you have a collaboration application that improves productivity and is going to be used day in and day out, here are some alternate interfaces that your users would like for your application. You would have known most of it for sure, but I thought of listing it here for the record as I have observed the usefulness of some of them in www.socialcast.com (an enterprise collaboration tool).

• Desktop App that shows information inline.
• Outlook Connector to view the info in dedicate folders
• iPhone App
• iPad App
• Android App
• Bookmarklets using Add Ons for Firefox, Chrome, Safari
• Browser native viewers by using Opera Extensions, Safari extensions, Firefox Add Ons, I.E. Add Ons, Chrome extensions
• Share (post) via email
• Show messages as
• Blackberry App
• Google gadget or GMail Gadget
• Facebook App to show message in facebook
• RSS Feed
• Vista Gadget

3 Billion Devices Run Java - Oracle marketing Java

Oracle monetizes and runs a lot of their technology well. Sun was a technology company that helped many others but itself make money :-) 

Here is how Oracle is now positioning Java, something that Sun Microsystems failed to do. Here is a screenshot of the Java update installer that sells to the masses what Java can do !

A nice lesson on technology marketing. Even if you make rockets, spell it out in a language that is easy for your consumers to understand.

What should Mail for businesses do?

Today email is like the blood that runs through the veins of a business. Businesses use it for communicating with customers and collaborating with employees. It is used for various reasons like sending an invoice, sending a generated price quote, scheduling a meeting, replying to technical queries, sending out bulk mailers about special offers, and lots of other tasks that help generate revenue.

If the number of emails sent for, say, replying to technical queries increases, it becomes hard to track it in a usual mail box. So the need for Support Management Software arises. Likewise the need for scheduling multi-user meetings and keeping oneself more organized, resulted in full fledged calendering software.

At a smaller scale email is actually good enough. Keeping your email unread is a basic way to remind yourself of a pending task.

Now when the number of interactions increase, that signals there is more customers and employees which signals more revenue for the business, which means there is a need for some special software. This software could be specialized software like CRM, Support Software or even better a mail suite that helps you get the job done right from within your mail client. For example, a reminder for contacting a client after two weeks, could be an email turning to unread state at the specified date and time and coming right on top based on a reminder you set for the last email with the client!

I would call these mailing solutions as "smart mail suites" as they give some sort of modelling (some additional capabilities) for a normal mail box.Now for every million dollar business, there will be tens of $100K businesses. These $100K businesses, I will call micro-businesses. For these micro businesses, smart mail suites are good enough. They can't afford to pay for integration of specialized software. They also find it an overkill to jump between interfaces to use these specialized software too.

Everyone needs an integrated view of their business communication. That usually needs a  million dollar software integration project to make happen in a big business. Why create the problem in the first place for a micro business, if you can provide those essentials in a "smart mail suite" ?

We could go one step further and tweak this use case for mail for micro businesses, mail for SMBs and more.


I will close with one additional point. Your hosted mail solution should obviously be miles ahead of hotmail.com of 1990s?

IT Team Job at stake?

If your IT Team has been caught napping when the virtualization, cloud and now SaaS wave is sweeping, I would say its your fault. Sorry for being harsh.

The main value I see about the cloud is accessibility of data. Virtualization and public clouds are an enabler for SaaS. You now have more ways to access it as your data has left the "sacred corporate premises". 

Previously IT Team would harp about security and prevent any form of access from outside. However that stiffled innovation and productivity of the employees.

If you had ensured your business critical functions (applications) were made more accessible, maybe with lesser rules, using other technologies,  the up take of SaaS could have slowed down. This could have given the IT Team more headaches. But headaches create work :-)

Steve Jobs: Learnings

Lots of praise has been showered about, how Steve Jobs was great in building insanely cool products like the iPad, iPhone and iPod.

However there may have been a greater strategy there. He may have had a vision of accessibility of data for consumers. It was about data all the way.

At many popular points of human interaction like listening to music, the way we work, entertainment, he dropped gems on the way. Gems like iPhone, iPad, Apple TV, iTunes.

He has then set up Apple to continue making data accessible via iCloud. To complete his actual vision?

Steve Jobs : Rest in peace.

What is DevOps - a short note

DevOps is heard quite often now. I will try to give my perspective.

Application management teams are normally the bridge between development and operations. To help the technology stack developed by the development team to be utilized to its optimum extent in production, there must be collaboration. This is facilitated by development team using APIs and instrumenting the applications with application specific metrics that relate closely with the actual functioning of the hosted application.

For example, just saying CPU Utilization is high does not mean much in an application context. However telling there is a surge in number of purchases via the shopping cart right now and that is causing abnormally high utilization, means a lot. This "application specific input" means a lot for operations. Getting this kind of collaboration going is DevOps for me :-)

An IT Management product could help the development team integrate these business metrics in to the same Operations Dashboards that your NOC will use. The development team would ideally use programming languages like Java, .Net, Python etc. or protocols like JMX, WMI and SNMP to communicate with the management tool.