Skip to main content

2 min read

Today was the first day of the 21 day COVID-19 lock down.

I spent my morning struggling to get groceries. Fortunately the nearby super market was open and most of the important items were present. 聽Items like milk and floor cleaner were over - but most vegetables, rice and yes, Maggi was available. If you see somebody else buying something, you also keep a couple of packets of it. It took 40 minutes to clear the payment queue - but well, that's a small inconvenience given the situation we are in.

Today, I read a few pages from Noam Chomsky's Understanding Power. Noam Chomsky is a acclaimed scholar with lot's of controversial opinion - so it's a fun read. One essay which caught my eye was his discussion on free markets.

He argues that

there is not a single case on record in history of any country that has developed successfully through adherence to "free market" principles : none

He gives example that most successful pockets of economies have grown with state intervention. United States had excessive state intervention from early days - right from defence funding which gave birth to the Silicon Valley tech companies to the earlier textile industry which survived because of high tariff on imported goods.

We see even this playing out today with China's protectionist govt. policy giving birth to multibillion companies like Alibaba, Tencent and Baidu. While back home in India, the poster child of startups, Flipkart was acquired by Walmart.

This made me think. Is the narrative of free markets just a hogwash?

Chomsky says

the "free market" ideology is very useful - it's a weapon against the general population here, because it's an argument against social spending, and it's a weapon against poor people abroad , because we can hold it up to them and say "You guys have to follow these rules," then just go ahead and robs them.

China never worried about being protectionist. Still its tech industry is doing pretty well and even challenging the US.

Is India's policy of keeping markets open, just falling into the narrative of "free markets are good" which other countries want us to believe in? Should India be more interventionist?

2 min read

Today, Prime Minister Modi announced lock down of the whole nation for 21 days. Apparently the threat of corona virus is much higher than it was envisaged to be.

More crucially, the US is having 聽hard time coping with it. India, with a much poorer healthcare infra is at a much higher risk.

What this means for us is:

  • More time spent locked down in front of computer
  • No outside running & exercises

Since, there will be lot more time at hand, and looking into screens gets boring & stressful after sometime. I have decided 2 things

  • Spend the extra time after work on reading books. I went through my personal library and dug out some older books which I always wanted to read. Hopefully, will post notes from them here in coming days.
  • Write a blog daily at night till the lock down is in effect. This would help me clear my head and hopefully share some notes here

Today I picked upon Glimpses of World History by Nehru. Its amazing how long term he plans. He is in a jail near Allahabad circa 1931 but is writing letters to his daughter Indira - to educate her.

I am still in the early history of ancient world. Two points which stuck with me:

  • India & China are the only civilizations which have continued in some form or other throughout the last 5000 years. Other ancient civilizations like Egypt only have some remains from the past, but there's no thread of culture/traditions running from the ancient Egyptian civilization to modern day Egypt.
  • Greece, though a very advanced civilization, always preferred to have City states. They never had a large kingdom like Persia or later day India. Is it because of any geographical factors or is it just that they are fiercely independent? Also, the Ancient Greeks are from a branch of Aryan race - Greek Aryans. Similar to the branch which came to India and was called Indo-Aryans.

2 min read

Off late I have been having lots of conversations with developers and engineering teams. Primarily because we are building an observability platform for microservice applications and developers are the primary customer persona for this product.

What I have observed is that developers are a curious lot - and always willing to explore new tools. Now, I have been a developer in my past life and have worked closely with devs most of my professional career. So, I completely identify with this. If a new tool is useful and I introduce this to my team/colleagues first - I immediately get the swag rights.

I am sure that this is the case in other professions also. Though there are some nuances.

A marketeer who has figured out a new channel with better RoI - firsts milks it himself and then only tells other marketeers about it or writes blog posts on it. This is also because if more marketeers start using that channel, then the attractiveness of that channel and hence RoI decreases.

Among developers though, there is no such competitive mechanics. If there is a more efficient tool, the larger number of devs use that tool - the more efficient everyone becomes. I also think that the rise of open-source is a reason why devs are more ready to share what they are building. Code has high marginal utility and zero marginal costs. Distribution is almost free - and the amount of value it can add to each new user is the same for the first few users as the later ones.

This mechanics of developers sharing the cool new thing they are working on - adding to their brand and increasing everyone's productivity is very unique to this profession. And hence working on new shiny things is cool.

If a salesperson discovers a new way to find emails of people, I think he will first milk it before letting others know. But if a dev finds a new tool that increases his productivity, he will immediately write a new blog post to share how he uses it to improve his output.

The Shiny new tool syndrome is actually a boon to the ecosystem. Of course, it needs everybody to be updated - as the tooling changes every 2-3 years. But if we are able to do more things, by learning from others and teaching others - why not?

4 min read

Data is the new oil. Deciding what to build in a product is powered by this oil. Especially in B2C products where its virtually impossible to talk to every customer and build product based on that. Instead data act as the indicators which reflect consumer feedback and is used in driving product decisions.

As product managers, we subject users to different "experiences" and based on the choices they make, it tells PMs whether it is helping customers or not. Product managers act like scientists who change small things in the product and measure consumer response. If the changes help in improving the metrics which the PM is aiming to move, then the experiment is considered as success, else it's a failure. In typical spirit of the scientific method, failures are not bad as long as the experiment was correctly designed and we learn something from it.

A few terms in this definition need special care and should be clearly defined.

  1. Metrics - These are numbers which the PM is trying to move. This could be conversion of users landing on the home page, or number of uninstalls within 7 days of installs, etc. All these numbers should be carefully defined to design the right experiment and move the metrics.
  2. Measurement - As the famous saying 聽by Peter Drucker goes
If you can't measure it, you can't improve it.

Though it is also important to measure correctly - so that the metric being measured actually reflects the goal.

For example, if you are trying to improve retention for an app, what is the best metric to measure. Is it uninstall rates or is it the retention cohorts? Should D7 uninstall rates be measured or D30?

Since, data is so crucial, it is also important that we are able to capture the right data points and analyse them in a variety of ways. 聽Different ways of looking at data can reveal patterns which have interesting insights to improve product metrics. This is especially important when PMs are deciding to build new features

This also needs that the data stack being used is flexible enough to support different ways of querying and it is easy to query them. If something is difficult to do, then humans will infallibly try to avoid that.

There has been huge increase in demand for data analysts - who can provide insights to the respective teams. But in my opinion, adding one more layer of people assigned just for the task of doing data analysis makes the process slow. If a product manager is trying to get data on how different features are used by customers, he should be able to just dive right into the data - and get the insights quickly 聽- in a user friendly way. Similarly, if marketing wants to learn how 聽users from different marketing channels are performing in terms of their retention, then they should be able to just make a few clicks to get this data. Data Analysts as bottlenecks Now, the issue is - this needs good knowledge of how the data is being stored in the DBs and what type of values they have. For non-technical roles like marketing, PMs, this may be non-trivial problem to solve. Just having a data analyst doesn't solve the problem. Also, you are not entirely sure if the results which an analyst gives you is the exact thing which you are looking for.

One the important metrics to measure data analytics ease is to measure how many people are using it. If more and more people in the organization are using data to power their decision, then the data platform is generating value for the company and hopefully powering good decisions.

Some current data analytics tools which are used by companies for visualizing their business specific data are:

  1. ReDash
  2. Metabase
  3. Google Analytics
  4. Grafana - For visualising the metrics

All these tools have their own pros and cons. I will try to delve deeper into this in later posts.

Meanwhile, would love to understand how do you visualise you business metrics currently? How happy are you with it - and what are the challenges you face?

Drop me a note on twitter @pranay01

2 min read

Landing pages are critically important for a SaaS product. How fast does it convey the value proposition of the product? Does it make potential users move down the conversion funnel? What's the conversion rate? These are the questions which many product people fret about.

One such brilliant landing page I stumbled upon recently is that of Hemingway Editor. In fact the landing page is itself the app and the demo. Landing Page of the Hemingway App The prefilled text tells you what the app does - and demonstrates it live. It shows features like analyzing complexity of the sentences, adverb usages, etc. And once you have gone through the text, you can just edit the passage to test your own writing.

  • The landing page, the demo and the product are the one and the same!*

The only other CTA is to download a desktop app, which is on the left top corner. It attracts attention because of the handwritten text font and the dotted arrow. It's as if your teacher is pointing something to you. You can't dare but take notice.

In the world of SaaS landing pages, this is the Zen state. Very hard to achieve. But this is simplicity personified.

4 min read

10 yrs is a long time .. long enough for a man to reflect on what happened, how it happened and how he felt about it.

2009 December - I was a recent graduate from IITM, in my first job. I had experienced the joy of earning your own money and was fully enjoying it. The monthly salary seemed good enough for me and I hardly spent about 20% of it. The work was interesting. I was working on fingerprint recognition - which I never thought would become so important with coming of Aadhar. Random company outing @ first job I never felt as rich as I did in my first job - though I have earned several times what I used to earn at that time. Someone wise must have said something around - Happiness being the difference of what you expect and what you get.

In 2011, got admissions from UCSD and Georgia Tech in their MSCS program - but decided to stay in India, doing MBA from IIM Ahmedabad. While working for 2 yrs as a software engineer, I had realised that its much more interesting to solve business problem - rather than doing things which you are told. Little did I know, that there are other ways to do this. Startups were not a thing then - Flipkart was just starting and we still used to buy books from book shops.

So here I was in WIMWI (Well known Institute of Management in Western India) as folks fondly recall IIMA. The first few days made me think that this was the worst decision of my life - the amount of work was just insane. But slowly, it got better. Everybody there was someone who had excelled in one field or other 聽- and the environment was very intense. The courses were not particularly tough, but you had to put time to understand stuff - and time was the one thing we didn't have. We used to spend 4-5 hrs sleeping and that too seemed a luxury. Last quiz of first yearIn our section - First year@WIMWI We used to joke that dogs at IIMA had much better lives than they students. Rushing to the classes in the morning, we would see dogs peacefully basking in the fresh sun - and we would be so envious of them. But things got better with time.

I got a summer internsip in RBS London at their trading desk. This was at the centre of the City of London - which is the no.2 financial centre in the world after Wall Street. During my first year at IIMA, somehow I had figured that finance would be a good place to try - as it was glamorous and was supposed to involved math. And I always thought I was good at it.

And London was a sweet heart. The moist, cold 聽weather was stark opposite to the dry, hot weather of Ahmedabad. From taking the double decker busses to reach office, yoghurt for breakfast, mexican wrap for lunch to tube back to lodge - all my actions were optimised to save every penny. Some random street in London circa 2012Isle of Skye - Scotland It was not that the internship stipend was meagre - in fact it was quite abundant. But, I couldn't make myself spend money after multiplying all prices by 80 (1 GBP was ~80 INR then) - so a 5 GBP coffee would quickly translate in my mind to 400 INR - and it was difficult for me to shell that amout for a mere coffee :)

(This is turning out to be a longer post than I initially thought. Will continue it in the next post)

4 min read

In our first year MBA class, we were taught the Efficient Market hypothesis. It basically implies that it is impossible to beat the market on a risk adjusted basis. If you want to beat the markets, you have to take more risks.

Now, this hypothesis relies on a class of people/agents who can be loosely identified as "entrepreneurs" - who identify any opportunity to make higher returns, make money out of it and correct the prices. Hence, markets again become efficient.

The thing to note here is that, this class of people which I have called above as "entrepreneurs" are those agents who operate on the margins and always looking to push the boundaries. e.g. If there is a new cheaper way to drill oil, entrepreneurs will exploit this technology by forming ventures which exploit this - and bring the prices down. The market thus gets corrected. If there is a new way to deliver goods cheaper by drones, there will be entrepreneurs who will start ventures to provide this service - and hence will make delivering goods cheaper. This will reduce the price of delivery and hence the logistics market will get corrected. You get the drill...

But, being in the space of creating ventures and constantly thinking about it for few years now, I think that not enough thought has been given to how these "entrepreneurs" are created. There are no single, well defined process by which entrepreneurs are created. Some accidentally discover an opportunity, some tinkered for long years to reach there. There is lot of luck involved. For example, Jeff Bezos came to know about the potential power of the Internet - by working for DE Shaw - which was a hedge fund trying to exploit the emergence of computers and the Internet. Ray Krock, who took over Mc Donald from its founder - and scaled it across the world. There is not set pattern for it.

Since the process by which entrepreneurs come to be is so random, how can something like the Efficient Market hypothesis - which relies on "entrepreneurs" exploiting every possible opportunity - be true. There are many areas where very few entrepreneurs enter - just because of the nature of that market - and these markets stand uncorrected for a long period of time. Markets which involves hardware are uncorrected for a long time - because very few entreprenuers go there. Areas which are loved by VCs get lots more attention - and hence are more efficient.

Of course, if by some random chance, an entrepreneur goes to an unattended area - and achieves success - then a lot of money flows into it - and the market gets corrected. But, till that entrepreneur stumbled upon that area - it was uncorrected for a long time. A case which comes to mind, is the area of RPA (Robotic process automation) which was unattended for a long time - till UiPath somehow stumbled into it - after barely surviving for 8-9 years. Of course, after UiPath - lots of money flew into this market and the market is overheated now.

So, the very mechanism by which markets get corrected is so flaky - so based on luck, that it always remains uncorrected for long periods of time - before it gets corrected.

The invisible hand of the market works through the toil of entrepreneurs. But the creation of entrepreneurs is unpredictable, and thus the invisible hand also plays to its own tune.

Efficient market hypothesis is just taught to students, to make them feel that the world works through well defined rules. While in reality, there is so much depth and complexity in the world, that simple rules like EMH can't possibly ever explain it.

One min read

Is quantity more important or quality? This question comes up in many facets of life.

Is hard work better or smart work? Is trying multiple things more important or carefully selecting the few you try?

The fact is that we much more control over one than the other.

We can easily measure how long we work, while its kinda squeasy to determine how smart. Of course, if your task is very simple with one metric - then how much you moved that metric is what matters. If you are a worker on assembly line, only thing which matters is how many cars you fitted.

But modern jobs are hardly so simple.

Still, the only thing we really can measure is how long we worked on something. How smart was it is only validated later - when you measure how successful that task was. At least working longer hours lets you try more number of things - and hopefully you stumble upon something smart.

So, there is really no dichotomy here. Work as hard as you can - at least in early days when you don't have much idea of what smart means.

2 min read

Being a founder is considered a milestone by many. You have finally taken the plunge to make your dreams a reality. Something you have been carefully tossing around in your mind, can now finally be tested against real people.

Is your product idea valuable? Will other people benefit if it becomes a reality? Does your startups existence create any value?

This also gives you immense amount of independence to do whatever you want. You can spend time on talking to users, researching the market or reaching out to people to get 'gyaan'. 聽In your head, everything does add value to take your product to the elusive PMF. The intentions are right, why won't they be? After all, you are the founder.

But indepedence is also a double-edged sword. A very broad canvas in early stage can lead to lack of focus. You can easily lose time - without seeing any visible movement in key metrics. And since time is always scarce in startups, you need to be miser about it.

Focus on the key metrics you want to move in a week, and ruthlessly prioritise to only take up those tasks which move these metrics. Be realistic, not optimistic.

This is something which I am trying to improve in myself.

Define metrics. Chase them. Reflect what went right/wrong. Iterate.