Electronics Rocks – 2014

erocks_logoBeen thinking of writing about Electronics Rocks 2014 (eRocks), finally able to pen-down few things. For people who don’t know what eRocks, here is a brief – it is one of the most popular electronics conference organized by EFY media. Last year I attended as a participant found some interesting things in the conference. This year, after joining Emertxe we got opportunity to be a community partner of the event and offer a presentation on the Internet-Of-Things (IoT) design challenges. The event happened during October 10th and 11th at NIMHANS convention center, Bangalore that attracted 3000+ participants for the event. For my session about IoT which was the key focus for the conference attracted 200+ participants. Post presentation we have received very positive feedback from the community.

While there are many things about the conference, here are my top-3 learnings from the conference:

  • Not only Open software: The field of electronics has become more interesting in the recent years mainly because of open source software and easily available/affordable hardware. While devices like Raspberry-Pi, Arduino has already become very popular, I found some of the new devices like UDOO which are becoming very powerful around which many cool things can be built. Going forward I see many companies flocking into this space which is yet to be tapped to its full potential.
  • Product design: While the previous point gives opportunity to build around so many ideas, there seems to be very large gap when it comes to product design knowledge. While student level knowledge is enough to build a prototype (ex: Agriculture automation) making it as a complete product required a different set of knowledge. During my discussions with many enthusiasts I found there is a severe lack of knowledge about Productization using real hardware.
  • IoT is not new: While there is a lot of opportunity around the IoT space, in my opinion it is not something very new. Connecting devices to network (say LAN) is been existing for a long time, which has taken a upgraded as IoT thanks to multiple advancements happening in embedded & web application development. During my talk also I mainly stressed about this aspect, where fundamentals needs to be taken care to build products in the IoT space.

Here is the Slideshare link to my presentation, comments are welcome. Couldn’t spend much time across various tracks due to my time constraints hopefully next year I will be able to do better by listening into multiple tracks (ex: Jugaad innovation).

Product discovery – what exactly you call as “product”?

For my related post on involving customers during product development by making them co-creator, one of the readers comment that the same approach will not work in every product development. By involving customers during product development, it might look like a “service” to the customers by taking example of Google v/s Apple way of building products. In case of Google, they work on a product discovery approach, whereas Apple has taken the approach of “Don’t ask the customer what they want, many of the times they themselves don’t know” approach.  In my opinion, major differentiation between these two approaches come from what we “refer” as a “product”. Buying a “box” product  from the store and experiencing  it with hands-on is way different from accessing an online product. This has a significant change the way products are developed. Let me explain this from my own experience.

I used to work in a firmware development team, which had very close traction with hardware. We made certain firmware changes specific to hardware, any further changes to it will result in re-manufacturing the whole PCB. Also product milestones, changes/defect fixes were controlled (ex: All level-1 severity defects should be fixed before manufacturing release, which is the final release). Later point I moved to other team, where we were building cloud solutions from the scratch. Our first major launch was two days away where we had at-least a dozen level-1 bugs were open. When I asked how we can make the release, the product manager replied “We can go ahead and launch then regularly roll-in patch releases by observing customer usage. If customers are not facing any issues, it is fine. After all for this flexibility only we are moving to cloud”.  Well, this fundamental elementary thinking of how we look at defect fixing itself different when it comes to looking into products as something.

Thanks to advancement in cloud technology, tons of mobile and internet products are getting built every day by following the discovery process. They can release, iterate and then improve their products depending on how customers are responding to it.  Recently there was an article about Amazon’s product building which talks about similar approach. On contrast, when you are building anything that is closer to the hardware, taking this approach might create more problems. For example, given a trial a customer might say he wants infrared interface in it, which might result in months of time to tape out and re-design a new board. Whereas in case of web, if the customer wants a button to be changed, probably it can be done in few hours.

In conclusion, I would say product development largely depends on what exactly we call it as product. The development methodology should change and be in sync with it.

Ten types of Innovation – Concluding notes

With article on Bigbasket, the ten part innovation series comes to an end. When I understood the innovation types (created by Doblin) way back in 2011, my idea was to apply it from Indian context and make case studies fitting various types. It took two long years for me to complete this series with decent satisfaction.

Innovation has gone beyond building a particular product or service. By building something different doesn’t guarantee a business success, whereas ensuring customer derives value will. India, unlike some of the developed countries, is in the cusp of transformation where we have both traditional old school thinking and new school of thinking co-existing with each other. This made my inquiry to innovation all the more interesting. As and when I observed some innovative way to serve customers, I started mapping them back to Doblin’s model and came up with this whole series spanning across industries.

Please find URLs to individual posts as follows:

The Business of Open Source

In my previous post about Linux, I touched upon some of the key aspects that Linux lacks in order to become a successful consumer desktop operating system. However, as a programmer/engineer Linux is probably the best tool I can think of, when it comes to community based development and collective wisdom. Last few weeks I have been spending quite a lot of time studying the Open Source ecosystem and how various businesses are leveraging this ecosystem for their benefit. Let me lay down three major observations based on my study so far.

Open Source - Big deal of Business

Platform Linux – Unlike few years back (say 2007), the Open Source Linux has become much mature by looking into the Operating System as a platform. By adopting OS as a platform, products (mainly embedded) can do a lot of customizations depending on their need. This main advantage is aided by multiple middleware platforms (ex: ENEA) who offer options integrated like – build framework, Open Source package integration, patch configuration and management, tool-chain support for various architectures (mainly for ARM/PPC/x86) which is making very comfortable for any product to adapt Linux, which was a huge challenge few years ago. After getting the basic things working with the middleware platform, products can develop custom applications depending on the functionality. Based on my observation, I could see customizing and making Linux as a platform itself is a huge business, which most of the popular vendors (Windriver, Suse, Monta Vista) are doing. This is a remarkable change I see compared few years ago.

Quality at every step – The Open Source community of developers has attained a high level of maturity, which is built over two decades. Since the main source code for Kernel (kernel.org) and individual projects (souceforge.net) is maintained by volunteers, purely driven out of passion they ensure proper code review is done and approval process is followed before committing any change into the main branch. In case any issues, ‘self-detection-and-self-healing’ approach adopted by Open Source ensures it is rectified at the earliest. Also a bunch of benchmarking tools (ex: Linux Test Project) available to quality the changes made. Since everything is volunteer driven who stick to a set of common goals, Open Source is no longer a toy in the hands of geeks.

Corporate support – Now that big organizations have realized the power of Linux, they seem to support the community in a very strong manner. Every other Linux conference or event receives huge amount of financial and resource boost from these large organizations which is very heartening to see.

Again, these are my initial observations as I try to learn more into the world of Linux and Open Source. Some of them might look like very basic observations for an expert. Going forward, I will add more in-depth information based on my learning.

3Cs (Connectivity, Content, Community) of Aakash ecosystem

As a follow-up for my post on Aakash ($35 tablet initiative by Government of India), I thought of re-iterating the importance of building an ecosystem, which is critical for this whole new paradigm to become large scale success with real customers. No doubt, the student market in India has significant potential, which is waiting to be tapped. Aakash is trying to address this market, which need to overcome three major barriers, which I would like to call as 3C – Connectivity, Content and Community.

3Cs of Aakash Ecosystem

Connectivity – The first and foremost barrier is about establishing ubiquitous Internet connectivity across the country. The current Aakash specification reads WiFi and GPRS as primary interfaces for accessing the Internet, which itself is a major flaw. Wireless hotspots are not famous even in cities, where tech-savvy customers also need to struggle to get the Internet using public WiFi. Setting up wireless LAN infrastructure in rural areas has other additional challenges like – configuring/installing/maintaining the modem, non availability of power supply and running wired DSL lines by laying down copper infrastructure. GPRS solves the connectivity issue but speed is a major concern. In order to run educational applications like Multimedia/Streaming videos, downloading large size course contents, doing online assignments speed becomes critical. In my opinion this offers significant entry barrier for Aakash to become successful. In the past Simputer, was launched with similar objective, but commercially failed due to lack of connectivity.

Content – Assuming that the connectivity issue is resolved somehow (me being too optimistic!), the next obvious question is — what would the student access from the Internet? There should be a strong content repository in order to create value. Here I see two major classification of the content from the student perspective. First, the mainstream academic contents (which are currently in form of book as of now) need to be augmented with – practical experiments, use-cases, live demonstration, do-it-yourself kits etc, which will enhance the academic understanding. This would also (to a larger extent) solve the issue of rote based approach taken in majority of the schools.

Second, vocational learning content needs to be added for enhancing additional skills for the student. For example, when a student wants to build a project (say on road safety systems) he should be provided with content repository and associated tools, where he can access, understand scrape and create a project using the simple online tools. Another important dimension of the content is to add support for vernacular languages. Because majority of the education is still taught in regional languages, even teachers are not trained to teach in English.

Community – Third aspect is to have a community based approach for constantly monitor and enhance the content repository. This community should consist of programmers, education experts and teachers who can think and shape a vision for education. Eventually this should lead to create an ‘Appstore’ where education based applications can be downloaded and installed depending on the need basis. Since Android is used in Aakash, it can be leveraged very easily.

Each of the above mentioned barriers has multiple phases associated with it, demanding huge investment, which Government cannot support. There are few initiatives like Spoken Tutorial (from IIT-B), which eventually will not scale beyond point. Here is where the opportunity lies for private business entrepreneurs as called out in Kookskool blog (hosted by one of my friends). The lower price point of Aakash is definitely a pull factor, which is one piece in the whole puzzle called Educational transformation. Just by launching a ‘cheap’ device doesn’t bring any change, unless we address 3C challenges mentioned above. It shouldn’t become another failure in the lines of Simputer.

The Akash ecosystem

Akash, the much awaited ‘cheapest tablet’ got released yesterday, here in India. I could see different set of numbers – $35, 2250 INR, 2999 INR, about its price, quoted by various websites. I also understand the commercial version will be available from November onwards. Definitely the price looks very exciting, given Indian domestic market condition and comparing with other variants of tablets running Android.The concept of ‘thin’ device (mobile/tablet/PC), connecting them to web using GPRS/3G data connections and offering multiple services in terms of Apps, is in the news. Added to that interesting ideas (Apps would be hosted in SaaS model) are floating around this ecosystem for enabling monetizing aspects. Apart from the tablets, Android powered mobile devices are flocking the same space as price-point of such devices is sharply declining and the form factor of the device is increasing. 

On the PC side, the concept of ‘thin PC’ been there for while now. Novatium Navigator is one of the early entrants into this space, which offers Open Source Linux distribution with customized application ecosystem. Novatium has a different business model as they sell the device thro’ Internet Service Providers like BSNL. This proprietary device is sold at lower price point by bundling support/maintenance service for a nominal amount (on a monthly basis), thereby recovering the cost over a period of time. While there is lot of talk (or hype?) about this new ecosystem, I am still not able to see a single product becoming a big success in Indian market. For example, the first generation mobile VAS providers (Java enabled phone + Stand-alone App + GPRS connection) generated similar buzz around 2005 time-frame there are hardly any product which has become ubiquitous, used by millions of Indians (there are few successes in mCommerce space though).

This raises few more questions from my side:

  1. Assuming there is a potential market that needs to be tapped, is the ‘real’ need of ‘real’ customers is well understood?
  2. Majority of these devices/services are available only in major cities, which hardly contributes into the total market size. Why still the gap exists?
  3. The customer behavior and the way they look at computing should be changed by innovating around the way software is built. Vernacular language support is an immediate case that I can think of, where the focus is pretty less

Tail-piece: After publishing this post, I came across an interesting link from PluggdIn, where one of the recent survey claims Indian mobile customers spend more time in using data than for traditional voice + SMS. The article also states that the issue of distribution & lack of willingness (to pay for content) remain intact. May be this is where the core issue lies? An average Indian consumer has got used to getting everything for FREE, when it comes to software.

The Product Manager and the Business Owner

I see a lot of instances where these 2 roles are somehow being confused, leading to insufficient attention being paid to these roles. The purpose of this post is to call out the distinction between these roles for sake of clarity. Let us put in my thoughts on how these roles are different.

Business owner – Responsible for assessing the market opportunity, identifying target customer segment, setting revenue targets, putting together the sales / support strategy and overall scaling the business. The Business Owner role largely operates in the WHY realm. WHY is this market worth going after? WHY are we, as a business, best equipped to do this? WHY does this constitute a strong long term play for the business?
In a small company or a single product company, the CEO may well be the Business Owner.

Product Manager – Responsible for building the right product to enable the business to meet these targets, Ensuring that the product built has an USP or a differentiator among alternatives present in the market. Demonstrates, through live instances, that the product actually solves the customer need that it sets out to solve. The Product Manager role largely operates in the WHAT realm. WHAT is the right product / solution to address this market need? WHAT is the best user experience to be built? WHAT does your product deliver that others in the market dont? WHAT is the right go to market strategy?

The extension of this would be to look at Engineering Head as the HOW guy, but let me just stick to the PM and Business Owner here.

These roles may have their own independent reporting structure, but need to work very closely together to build a successful business.

Many Product Manager profiles and job descriptions that you see are a “hybrid” of both and for smaller businesses the same person doubles up as the PM and the Business Owner. While there is nothing wrong with you doing that, understanding the distinction between these roles is important. It is critical to keep a check on the extent of your focus on each of these and be aware of the balance. For a business in scale both these are huge focus areas and need dedicated mind share.

To start with you can take up both the WHY and the WHAT, but have a good sense on WHEN you want to hire someone who can complement you on this.


– NWritings

Performance Management – Keep it simple

In one of the earlier posts here, Performance management was called out to be one of the least understood domain. I, in fact, think its one of the MOST OVERRATED process. No, I am not saying its not important… I just feel that many companies have over complicated a fairly straight forward function. Here is how the whole “process” works in many companies:

  1. The employee is asked to fill out a long form, which includes her achievements, areas of improvement, support received, support not received, strengths, weaknesses, aspiration etc etc. As expected, most employees refrain from saying things that will adversely impact their hike / rating / promotions…. This takes a few days at the very least.
  2. The manager then goes over this with the employee to give her a good hearing and charts out a plan to address areas that need to be addressed – the plans put out are more often than not, not implemented subsequently. Watch out for the same set of inputs in the subsequent appraisal to understand whether they were indeed implemented or not.
  3. The manager then reviews this with HIS / HER boss to normalize things across teams and functions and comes up with the final hike / rating and promotions.
  4. The decisions are then communicated to the employee as part of a pretty long and typically stressful meeting. The focus is on retro fitting the hike that the manager decided to a rating.
  5. After a few days, the employee comes back with concerns about the way it was handled…. “My rating was lower than expected”, “would have liked more hike” etc etc.

Now the whole process is not only pretty long drawn (I have been in companies where this process took 1 1/2 months or more… twice a year), it also takes away a fair amount of mind share and focus from the core activity of building a product.

An alternate approach that I have found much more effective has the following characteristics:

  1. You, as a manager, is responsible for working very closely with the person and feedback is given then and there – for both good and bad performance. “Your handling of this customer complaint was impressive – here is why”, “Your commitment in getting this module out was awesome”, “You screwed up on this sales call – and here is how I think you should have handled it”. The results of the good or bad performance is fresh in the mind when the feedback is provided. You dont say during the performance appraisal “You messed up the documentation 2 months back”. This also gives a chance to the person to rectify things as quickly as possible. If you do not work closely with a person, get her immediate supervisor to do the performance appraisal. You dont.
  2. There is no surprise (positive or negative) during the performance appraisal – the person pretty much knows what to expect based on the continuous inputs received. If the person is surprised at the rating that you give her, it just means that you as her manager have not done your job and had her waiting till the performance appraisal to tell her how she has been doing.
  3. In short, feedback is then and there and continuous.
  4. In response, also build a culture that allows people to raise their concerns as soon as it impacts their work. To their “boss’s boss (if the concern involves the boss herself). The person cannot say, “I did not get clear specifications for this functionality 3 months back” during the appraisal.

This changes the whole performance appraisal from a “big band one time stressful” to an “on going continuously improving” process.

The common justification provided for the “big band one time” is that, it lets the employee to step back, reflect and set medium / long term goals. While it is important to “step back, reflect and set medium / long term goals”, I dont believe they need to be during the performance appraisal. It needs to be in response to an event (your competition launching an update ahead of you, the person realizing that he is working on a technology that doesn’t interest him, the person wanting to move into a more customer facing role, the person feeling he cannot work with a particular individual and wants a change).

These (and a lot more, in a typical product organization) require course corrections and changes to the way we work and set goals, but these hardly need to be part of the performance appraisals.

Keep it simple!!!!


– NWritings