Figure out what your destination is

I love Mahabharatha and all its characters. You can derive many parallels to your day-to-day work. Each character is either a design pattern or an anti-pattern. Lord Krishna makes a remarkable statement - "The river's final destination is the ocean, no matter how many big mountains are on its way". Each person has to figure out how to reach there. Pardon me if I am being too philosophical...

I think this particular sermon from Lord Krishna is the right message for IT Engineers who are working on huge projects (>100 members) and who feel restricted by the "policies/guidelines", that the projects enforce on them. 

If you know your destination, the "project bureaucracy" is just one more big mountain for you to cross. 

Are you getting bored with your mundane Java coding job? Figure out what your destination is. If your destination is to become one of those "programmers who earn more than a CEO", then the mundane coding that you are doing is probably adding up to the "10,000 hour rule"

Are you getting bored with the data modeling work? Figure out what your destination is. All you have to do is to find out learning opportunities to perform more work on data side, which might include integration or mining or visualization or many more..

Just figure out what your destination is and things will get lined up. If it still doesn't line up, then you are definitely not in the right place.

Keep your desks clean

Every time, an executive representing the customer visits an offshore IT campus, you would see e-mails like these spamming the developer's inbox from the senior management of that company 

"Keep your desk spaces clean and neat" 
"Do not leave Tea cups especially in open areas" 
"Be prepared to address any questions at your desk from the client" 
"Keep the lights on"

"Please come in business attire and be dressed properly"

Does the customer really care, whether your desk has the unfinished coffee lying around? No. 

He cares for what you deliver after drinking that coffee. And not what will happen to that unfinished coffee.

And the timing of such emails is great. "Only when the customer visits" I think the management stop worrying about desk cleanliness and start worrying about code cleanliness. 

Should desk cleanliness be a factor in becoming a preferred vendor for a customer?

The developers have to be prepared to answer any question from the client. Why should he  be prepared? Is this some kind of examination or what? The developer should be already  knowing what he is doing. So, if the customer has any clarification on the code that the developer has written, he should already know it. And I doubt the customer is going to fly 10,000 miles to find out why a piece of code is working in a particular manner. Should the developer be prepared for a customer visit? I think the management should be prepared for the customer visit, because they will have to answer tougher questions regarding deliver quality. And let the developers be how they are.

Service your customer at the right time

I went to an Indian restaurant near my home for lunch. My wife is not in town and I am not the greatest fan of my own cooking. Nowadays, most of these guys offer the "one-size-fits-all" buffet for lunch, which is good for the owners, not so good for the customers. Still, I went forward, took the plate and went straight to the "Non-vegetarian" section.

I opened the food warmer on the appetizers, which was named - "Tandoori Chicken". And I didn't find a single piece. I was disappointed, but I was so hungry that I decided to move on to the rest of the buffet table. I served myself some hot naans and curries. And went back to my table. On my way back, I told the waiter that the Tandoori Chicken container is empty. He told me he was sorry about it and offered to bring the chicken "fresh" from the kitchen, RIGHT AWAY.

I loved Tandoori Chicken and was ready to wait a few minutes before starting my meal. 5 minutes went by. There was no tandoori chicken. I wasn't able to fight my hunger. So, I started eating my main course (which typically is naan bread and curry).

After 15 minutes, the waiter brought me my Tandoori Chicken, hot from the kitchen. By that time, I was having my desserts. Now, my penchant for Tandoori Chicken, forced me to eat them, even though it had missed its chance. The worst part was that it tasted terrible. To serve me quick, they brought it without the "Tandooriness" in it.

Now, you serve late. You have already missed the bus. But you still try to catch up, which is good. But you serve bad. I am sure I won't go back to that restaurant ever.

When I was driving back, I started empathizing for the customers whom I would have worked with. There have been projects when we have delayed deliveries. We missed the bus. Even when we delivered good afterwards, it was late. Our code just missed the customer's taste buds. I remember we cursing them for not being empathetic for our hard work and great quality code. I just realized - "It doesn't matter"

It is very critical to know how important is something for your customer, not just quality-wise, but "time-wise". Even if I received a great pizza from Pizza hut, but if it is 30 mins late, I wouldn't be too happy with them.

How EZPass taught me to reason beyond the obvious?

It was a simple lesson learnt. Whenever I cross the tolls in the EZPass lane and happen to see the front-of-me motorists'  EZPass status alerting - "Go, Low Bal(ance)", I always thought that the guy had forgotten to re-fill his EZPass. It is right most of the times, but not all the times. Now the "minority" reason never occurred to me and I always thought that could be the only reason.

A few days back, my EZPass also started showing the same alert whenever I crossed the lane - "Go, Low Bal". Now, I had setup automatic refills and how could that happen. Then I realised that I had replaced my lost credit card a month back and the EZPass had the automatic payment on that lost credit card.

The core reason for my EZPass status alerting was not that I forgot to recharge it, but I expected the recharge to work automatically, but didn't work because of a failed  payment. That could have been the reason for many of the motorists that I see who gets alerted, though not a majority, as you don't expect the credit card payments to fail.

The obvious reason (motorists forgetting to refill) was so obvious that it completely masked me from other potential causes (failed payment gateways) and only experience taught me that there could be multiple reasons for a failure.

So, whenever there is a screaming, obvious reason in front of you for a failure, don't get dissolved in it. There might be also non-obvious reasons for a failure, which you might completely miss.

Attrition promotions

The Indian IT industry suffers from a real bad phenomenon. - Attrition promotions.

Let's say you are a super technical brain and is the most sought by your client. The client loves you and thereby your boss also loves you.

Now, you decide it's time to move on. Because the job doesn't interest you anymore. So, you open up the monopolistic email client "MS Exchange" and start happily typing your resignation letter. And then you click "Send".

A few hours later, if you are a super performer, you will be stunned with the number of calls you receive asking you to reverse the decision, as you are the most critical "bolt" in the engine.

Finally if nothing works, they promise "Promotion".

And then if you are person who loves power, you would gleefully accept it and work for another 3 years, before you send the next resignation letter.

Now, this is what I call as an Attrition promotion. To stop losing you, they promote you.

The industry is flooded with such wrong promotions at the wrong time. You should get promoted, when you are found capable and not when you want to leave the organization.

Instead of a promotion, the organizations can challenge their employee with some meaty work and equivalent pay. Promotions signal a wrong direction. Promotions mean that the organization is ready to wide the peak of the ladder.

Don't promote to make the guy stay in your company. If you find him capable, do it upfront or else don't do it.

A hair dresser's "Just for fun" past-time

Yesterday, when I was having my haircut, after a long 3 month break, my hair dresser asked me...

Hair dresser: "Sir, where are you from?"
Me: "India"
Hair dresser: "What do you do?"
Me: "Guess?"
Hair dresser: "So, if you are from India, you should be a software guy. Right?"
Me: "Perfect. Thanks to clustering algorithms in data mining"

I knew what he does, but I went on to ask him, what else does he do? He told me he takes MIT Online computer science classes once he goes back home and has been learning programming to write some apps for iPhone.

So, I asked him why is he doing all this?

"Just for fun".

Wow. That hit me like a bullet. Learning computer algorithms after a day's job of standing and dressing people's torn hair and that too for fun...I personally was ashamed of myself.

"When was the last time I learnt a new programming language and that too for fun....And I call myself a software guy..."

Realized that being in IT for a decade, has stained me with the whole notion of "cost-benefit" analysis. For everything, I look for a cost and benefit.

"What benefit do I get by learning Python?"
"What benefit do I get by learning clustering algorithms?"

It was a rude awakening. Probably fun is the benefit....

Leadership is support....

Do you feel like calling your boss when you are in a problem hoping he is there for you?


In my opinion, leadership is about the support you can provide to the community that you touch upon - your team, your customer, your vendor, etc, when they need you.


I want to take an example of one of my colleagues, who is basically an account manager. His role doesn't demand him to fix a java or a shell script, but in case of one, I feel like reaching out to him. His role doesn't demand him to take care of low quality conference bridge issues. But in case of one, I reach out to him. And he does all this without a fuss. For me, he is a leader.


I know whom to reach out in case of an issue. Are you one of that kind? A supportive leader or a tyrant boss.

Working with your competitors...

The IT Services business is very very tricky. And to win business, you are supposed to do anything...


You need to stay competitive. And you need to stay with competition...


The latter part is a very unusual situation when you are supposed to work with competition to deliver value to the customer. Now, how do you do that? You would have won the deal against the same competition and post that you need to work along with them. 


To give an example, take a Manufacturing services company getting serviced by multiple IT service vendors. Nothing unusual about it. Now, typically in a datawarehousing project, the data is pulled from multiple sources and then integrated into a datawarehouse and then reported from this environment into a presentation layer. 


We have 3 distinct ecosystems- sources, data-warehouse & presentation. Now, if each ecosystem is driven by a separate IT services vendor, and all 3 of them compete against each other for every project, how can you expect harmony to exist between the 3, which will naturally flow into the application that is getting built. 


Lot of times, projects get delayed not by the team performance, but purely by such strange de-risking practices followed by organizations to make the competitors work together. They never can. There will always be the line where information sharing will be seen as an "evil act".

Learning to document..hmmm...no...communicate

We document something or the other everyday. 

We document for "formality" purposes. 

We document to "save our asses". 

We document for "milestone deliverable" purposes (payments would be dependent on it).

But how many times have we documented to communicate?

Every time, I compose a design or an architecture document, it starts on a good note. I enter the "table of contents", with a lot of thought and structure and then as the documentation progresses, the interest wanes. The objective starts with "communicating the design" and then finishes with "documenting the design". A more deeper thought on it revealed that when I start documenting, I start with documentation, not communication...High level design documentation, low level documentation, support manual documentation, etc...These are all boring words to anybody. Who cares to read the disclosures? Who cares to read the "Conditions apply?". Because these are all documents. Their primary purpose is to save themselves and not to communicate something to their customers. That's why they are a failure..

Now documentation is a boring word...I researched the two keywords - documentation & communication on Google's Ngram Viewer and this was the graph...

(Communication vs documentation)

It is a nice comparison between the 2 words. One basic inference is that "communication" has been in use for a long time. Documentation seems to be a word of the post-40s. And communication has been used in a lot of books compared to documentation, illustrating the fact that its communication you would want to do rather than documentation in your projects.

Once you start thinking in terms of communication, your whole perspective towards composing any kind of artifact changes. You would want to communicate your design process, challenges you faced, your solution, benefits and drawbacks. Its like selling your design process. This will free you from MS Word documents. It can be a simple presentation or a movie or an audio file too or just a diagram.

Documentation..hmmm...no...Communication is an art..Its one of those pieces when done right, the satisfaction is fulfilling. Ask the book writers and scholars.

Formal meetings in informal setups

I entered this field of pre-sales/sales without any kind of formal education, so you can expect me to be a bit rusty. One month ago, I was in a casual dinner setting with a potential prospect. We were having dinner, chatting about projects, states and their cultures, etc...It seemed like everything went well that evening. I was giving my usual DW and BI story, maturity, roadmap, etc...I was invited into this meeting to primarily impart some potential pitfalls about a tool that the customer was planning to buy.


It has been more than a month and we didn't hear back from them, in spite of repeated followups. I was wondering why. Today, I received a call from my sales manager, who was also in that dinner meeting that evening. He told me. 


"Senthil, I would like to give you a small piece of useful tip. Never use your mobile during such a meeting, even though it might be a casual dinner meeting. Either excuse them and use it or don't even use it". 


After some analysis on that particular incident, I recollected that I used the mobile to text my wife that I will be coming home late. And some more analysis revealed that I had taken the dinner meeting with the prospect a little too light. After I was done with my job of educating the prospect on something that they were looking for, I was relaxed and forgot that I was still with a prospect and not a customer. I should have been more careful.


I realized that it was a bit rude from my end for having used my mobile. I wouldn't have certainly used my mobile, had I been in a conference meeting room or more of a formal setup. The restaurant made me feel that it was a very casual meeting and I felt it was okay to give up some of the "formal" rules.

Formal meetings in informal setups can be tricky. Never fall for the casual ambiance. It is still a customer meeting. I screwed up that day and I am thankful to the gentleman who called to correct me.

Internal competition & scarce resources

On our lunch table, we were having quite an interesting discussion. The question which was under scrutiny was

Why am I locked to one project? When will I be released? Hope I don't retire in this project.

Lot of organizations practice what is called as "Internal Competition". Internal fights are primarily instigated by the top management layer to get the best of the products from different project teams. In some cases, the same product might be developed by 2 different teams and the best version will be given the chance to continue it further.

This culture of internal competition can also be noticed in a lot of IT Service companies. For any IT Service company, their product is their people. They sell people and their experience on the projects And you will find that most of these companies would have services/offerings compartmentalized by "industry" groups. So, they sell people with good technology and industry skills. 

Now, when competition is induced between these groups, they fight for the scarce resource's (people's) time. They wouldn't want to release these people to other opportunities in other divisions. Its a classic example of "Controlled Labor", where the allocation of scarce resources is not controlled by the demand, but by "internal competition". And whenever this imbalance in economics happen, we know the result is not desirable.

Let's say I want to rent out 5 houses. Should I rent it out on the basis of the market(demand) or should I rent it out on the basis of some kind of rent control scheme ( like renting out to only bachelors who are vegetarians)? We all know the ill effects of rent control schemes. That's what is setting out to happen in IT Services industry. Technology skills are getting locked. And they are locked hoping that their skills would be required for future usage in the same industry group, whereas another industry group might potentially lose a deal, unable to staff the skill set. And in effect, the organization stands out as a loser.

Internal competition is good, but not in a way that it locks the scarce resources that the entire market is dependent on.

Filling timesheets

I am sure a lot of us would have had a tryst with the "nonsensical" work of filling up time sheets. This was initially introduced to keep a track of how long does an employee spend his time in a project, so that it could help them for 2 purposes

1. Bill the customer for the exact amount of time the "employee works".
2. Calculate exactly how much time he works so that it will help in future estimation of projects.

But how effective is this process? Measuring productivity has always been a great challenge. Because productivity is not just about time. It is about quality, efficiency, growth, etc.... And what is the solution that we have today? Time-sheets.

Why would I want to enter that I worked only 5 hours? Because I know that if I enter 5 hours I will be questioned, even though I had finished my work. Its like the prisoners dilemma problem in game theory. You don't trust the system, because the system doesn't trust you. And the system also thinks the same way.

Now, what about entering more than 8 hours? That is also a tricky problem, because you will be questioned on your effectiveness to finish the job and on top of this there would be 100s of clauses in the SoW which would detail that you wouldn't bill the customer more than 8 hours a day. So the magical number of 8 hours a day was established and employees just enter that number.

So this madness continues on a week-week basis to enter 8 hours/day. And in this process, we are not solving both the issues mentioned in my first paragraph. We will keep billing the customer 8 hrs/day.

In my opinion, organizations should first open out this channel to accept realistic time lines from the employees and bill that same back to the customer. Yes, it will cut corners, but that's alright in the longer term of things. There will be certain unpredictability, but its alright.

I think CMM guys should make this as a mandate in their evaluations !!!

We belong to 2 parents

In this post, I want to address one specific case of attrition. Attrition of employees from their "biological parents" to their "foster parents". In IT services, we are bound to 2 sets of parents at any given point of time. And I guess that relationship is what causes these kind of fall-outs.


Biological parents

These guys are the ones whom you work for. They are the ones who recruit you and position you into projects. They are the ones who send your paycheck every month. You belong to them.

Foster parents
These guys are the ones for whom you perform the work. They are the ones who nurture your skills. They feed you with challenges & requirements. They are your customers. In fact, they feed you and your biological parents.


Now, when you spend a lot of time with your foster parents, your loyalty and faith starts tilting towards that particular organization. You will be honest to them. You will give them unbiased opinions. You like them. What happens to the biological parents? They still own you.


I have seen a lot of my friends leaving their biological parents for a more stronger relationship with the foster parents. Is it unethical? No. I don't think so. It's purely the way this industry works. When there is osmosis, we will have reverse osmosis too. Its talent here. And it will move towards higher concentration. If the "biological parents" have to somehow stop this osmosis, they should be gearing for better people policies, which will attract their employees to stay back. Ideally, they should be more concerned about their people leaving to join their competitors more than their customers. Because when your own employee leaves to join your competitor, you have leakage of your IP, whereas if he or she joins your competitor, you have promotion of your IP.


I feel its perfectly okay when your employee joins your customer. He will spread the good word of your organization.

Why should a techie do pre-sales?

As a technologist, I used to always stay away from the world of sales. I hated it. And I hated sales guys even more. Because they would always sell some kind of service which we would either not have or not have currently.

A couple of years back, my boss wanted me to realize how stupid I was to think like that. He asked me to do pre-sales. In my opinion, a pre-sales consultant's job is the most discretely defined of all the positions that lead into the executive management space. A pre-sales consultant's primary responsibility is to convert a lead into a customer. Never was I faced with such a challenge in my entire technical career. All I have to make sure is not to lose a customer by delivering some crappy piece of software. Now that's preservation and not conversion.

This posed me immense challenges because my mindset had to change completely. From a cubicle-guy to a board-room guy. From a SQL-Editor guy to a Powerpoint-guy. From a internal-guy to an external-guy. The last point is very important. Pre-sales makes you more of an external guy. An external guy is one who understands the market. He is the one who understands the customer. He finds problems rather than solves problems.

So now, why should a techie learn pre-sales? Simple. To find problems. The more and more you grow up in your expertise ladder, you realize that you need to find problem patterns. As a senior consultant you are expected to know the problem before hand itself. And have a pattern available for the same. And pre-sales helps you to get into this mindset. Pre-sales will help you to lookout for problems before problem finds you. It will make you think in terms of broader problems and even broader solutions.

You talk to a pre-sales guy. He will give you the problem, design considerations and solution. You talk to a techie. He will ask you the problem and then he will give the design considerations and solution.

I need supervision.

Can you supervise yourself? One of my company's co-founders had published a book called 'The Professional' and he talks about professionalism is about self-supervision. A professional is about a person who knows when he has completed his job; he shouldn't be relying on somebody else to approve the completion of his job. If he does, then he is an amateur or apprentice.

Somehow, I am not convinced about this principle. And the reasons are as follows

1. The world of IT services is ever-changing. In that case, how do I know what is complete? I need some more eyeballs to tell me whether what I have done is correct or not, whether I have completed it or not. I need supervision not only to tell me if I have completed my work, but also who can monitor me on a day-day basis. To track my progress. To tell me if I am on track. To tell me if with this rate, whether I will miss the bus or not.

2. Our requirements are not as static as an undertaker's job. Its pretty fluid. We have to adapt and change. In such a changing process, you need expertise to validate your work. You need supervision to watch over your direction.

We all know the importance of reviews and followups. A world without reviews and followups is a highly idealistic world. A world of programs. Speaking realistically, I wouldn't let my own team self approve their own work, no matter how brilliant the team is. I will certainly get into some amount of detail and followup. I will certainly want to track their progress and not wait for the last day to get a status. Its not because I doubt their capabilities, but because I know that only constant feedback and tracking improves the quality of work.

Getting stagnated...

I have spent my past 8.5 years of my career doing Data-warehousing and Business Intelligence. What next is something that I keep asking to myself, almost everyday? One of my managers stopped by and asked me.

"Are you getting stagnated?"

I was taken aback. I wanted to know why did he feel so. He told me - "Because you are working on the same thing for 8.5 years".

In Indian IT Services, we have this strange theory engulfing every software engineer -

"Every 2 years you need to get promoted and become a Project Manager very soon".

And once you become a Project Manager, you should become a Delivery Manager and then an Account Manager and then...It depends on tall the corporate ladder and how wide it is....

This phenomenon is affecting the whole technology arena. The IT Services industry in India wants an engineer to transform from an "Innovate-new-things-guy" to a "Get-things-done-guy". An "Innovate-new-things-guy" has a lot of technology background and a "Get-things-done-guy" has a lot of management background. I would say the latter is the Supply part of the equation and the former is the demand part of the equation. And the equation needs both the parties. Somehow our IT Services managers are causing an imbalance to this equation by drying up the "Technology pool".

There is never stagnation in the technology pool, because technology itself is not stagnant. Do doctors get stagnated? They sit in the same chair for years and years and do the same kind of diagnosis, though every case would have its own complication.

There is always a career for specialization.

My biggest challenge right now. Follow-ups.

Last 2 weeks in my project were pretty challenging. Reason - 'Follow-ups". I hated the job of following up with someone to find out if they had done their job.

I hated it because I felt I was knocking into their privacy.

I hated it because I felt it was not under my control anymore.

And I hated it because it makes you feel like a beggar.

The more and more I hated it, the more and more I realized that it was inevitable. I had to do it. It is unavoidable. Now, I had this big task of doing something that I hated. Whenever I faced situations like this, there is one thought which keeps me kicking - "If 'I' could learn and write SQL, then I can do anything". SQL is something that I hated a lot. I used to score "Single digit numbers" in my training tests. But today SQL is my career. And I do it passionately. So I decided to apply the same philosophy to follow-ups too. Turn something that I hate, to something that I have to do it passionately...

Vidaathu Karuppu (A Tamil phrase for "an omen that never leaves your back")

Treat it as an omen that will always sit on your back. You just have to be-friend it. Once you start perceiving it as an "always existing" to-do, you will start designing permanent solutions for it. My very first hour in my day is a "Follow-ups" calendar item. I send reminders to all my "follow-up contacts" or give them a call to find out.

Sending a PENDING Status Mail

I send an email with the Subject line as PENDING. It is little bit rude, but I find it working.

Fix an interval and keep reminding

Fix up an interval like 1-2 days or 3-6 hours or any limit depending on the criticality of the process and keep reminding.

Highlight the impact and also the benefit of finishing on time

Make sure you highlight the impact of the delay to them. Also highlight the benefit of finishing it on time.

I use one of these techniques to get my followups closed. But I still struggle, because it needs a lot of patience.

I made a big mistake. Farewell is as important as Introduction.

I made a big mistake. When I signed off from my previous project, I didn't send a "Farewell email" at the right time. I sent a week later.

I worked out of my client's place. I sent an email to all my colleagues onsite (the term used to refer to client's location in the offshoring business) the day I left from the place. But somehow missed to send to my counterparts at offshore. I just sucked.

I didn't realize till one of my colleagues was surprised that I had left the project and I hadn't even conveyed my move. I felt miserable.

Sending a farewell email at the right time communicates a lot of emotions. It says that you care to communicate your decision and you respect them. It says that you would want to stay in touch and finally it says how much you will miss them when they are not around you.

I realized that farewell emails are equally important as an introduction email and it has to be drafted in a nice manner at the right time and not for the sake of sending one.

Cost of Default - A "Derisking" technique

I parked my car in my usual Park-Ride facility and was walking to catch my light rail to my office. I had to enter the parking lot # to buy my tickets. I remembered it, but I was reluctant to take a chance. My train had arrived, but I was prepared to miss it and go back and check my lot#. I walked back to the lot and verified it. My memory was right.

When I reflected back on that incident, I thought it was too risky to rely on my memory and the cost of default was not proportionate to the risk that I take (The cost of towing and the frustration that accompanies with it was just too much to take). The imbalance made me to go and verify it.

Why don't we do it for our own projects? Before we release a piece of code, do we think if it is worth releasing a buggy piece of software? Do we weigh the cost of default to the urgency? Probably, if we start to, it might help us to avoid a lot of unnecessary chaos that we are causing to the IT services industry,

Smaller teams vs Larger teams - What should you work in?

I have been always faced with one dilemma throughout my career of 8.5 years. "Working in smaller teams" vs "Working in larger teams".

Ok !! Let me define first...

Smaller teams - not more than 4.
Larger teams - more than 6.

I have experienced both these worlds, but I keep toggling always thinking the grass is greener on the other side. Or probably its paler on the current side...

I would prefer to work for smaller teams because

1. A smaller democratic setup.
2. Easier to steer the ship. Don't have to worry about too many moving parts.
3. Work gets noticed/criticized/rewarded immediately.

I would prefer to work for larger teams because

1. Highly dynamic and challenging environment.
2. Complex human chain.
3. You will learn scalability (I was born in India).
4. Most of the interviewers ask you how big was your team size and not how small was your team size.
5. You will learn to be more realistic.

Hence, at the end of a project, just like TamilNadu politics, I just vote for the other side. I think both the kind of projects are necessary for the evolution of IT services. And if you bias your opinions towards one of them, you will soon be gobbled by the nature of the game...