Friday, June 26, 2009

SMB Stack

Pradeep and I were together at Larsen & Toubro from 1995 to 2001. Pradeep was in-charge of cement sales for Gujarat (a major industrial state in Western India) while I was rolling out SAP at L&T Gujarat. We completed the SAP rollout in 01 and I moved onto join Satyam. Pradeep continued in L&T till 2003 until L&T hived off the cement business to to "Birla Group" (now Ultratech Cement). At that point Pradeep started his on small business trading in cement. He, gradually, added steel , ready mix concrete, cement and clinker export and expanded his business to whole of western India. Today Pradeep runs a Rs 400 crore ($ 20 Million) business all built from scratch in a matter of six years. Of course , the construction boom in India, helped Pradeep a lot in addition to his extensive experice in Infrastructure products.

I got a call from Pradeep three weeks back while he was visiting his daughter at the East Coast. After exchanging greetings and our stories, we settled down for a long telecon. He wanted to put in IT systems to run his business but was little skeptic of cost and ROI given the current economic situation and pull back on real estate / development projects.

As a "die hard" SAP fan , my first and quick solution was to implement "SAP Business One". It didn't take me much to unerstand Pradeep was well aware of new SAP solutions for the SMB segment. However , he felt business and market situation was not ripe enough to get into implementation of a complex integrated systems. He wanted to start with something simple , easy and quick and may be build in baby steps as the situation improves.

I asked him to give me some time for research and hook up next week.

Over last two weeks , I had number of calls with him and together we explored numersous solutions. I gave him a demo of Google Apps , showed him heap CRM. Personally I explored Sugar CRM and how Sugar could be chaeply hosted in Amazon cloud. While we are still in discovery phase , two things came to me which I feel worth a mention here:

  1. Theere are number of free open source (or very cheap subscription based) business solutions out there for small businesses. These solutions are there, practically, for every industry.
  2. These solutions are good from short term standpoint but long term TOC will be very high coz there is no 'off the shelf' interagtion available.
I guess we all understand both these points, but realizing them first hand gave me a totally different perspective. It also suggested that open source movement needs to be channelized now in the sense there is hardly any value in building a new CRM solution. But integrating existing CRM solutions eg Sugar with other open solutions can be immense business value . For example CRM solutions can be integrated with Google Check Out . HR solutions can be integrated with Linked in .. and so on . Another key area is to put together the methodology and buisiness processes to make these systems easier from deployment standpoint.

I thought the first step is to create a directory of open source solutions and put the information in one place. With this at the back of my mind, I stared working on smbstack where I want to stack good open source business solutions to create a base refrence location.. Hopefully it will grow into something more interesting and of value to to small busienss owners.

Please write to me if you want to contribute or have suggestions ...

Wednesday, May 20, 2009

Why Agile? | Planigle

Great and Simple Intro of Agile at Planigle.... worth reading
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Traditionally, software development has been done in a 'waterfall' process. First you collect the requirements. Then you design the software. Then you implement it. Then you test it. Release cycles have typically been on the order of every 6-18 months.

Approaching Things Incrementally


Agile development approaches things in an incremental fashion. You break things into time-boxed chunks called iterations or sprints (frequently 2 weeks). Within the iteration, you do the full life cycle, but on a much smaller scale (only focusing on the most important functionality). You typically come into the iteration with a high level understanding of the increment that you are trying to take on (often called a story). At the beginning of the iteration, you define the more detailed requirements (acceptance criteria). You develop and test the story within the iteration.

Staying Releasable

The goal is to keep things such that every iteration could be potentially released. Each iteration is demonstrated to the customer (or a proxy such as a product manager). This results in changes to the backlog of stories for future iterations. You can release at whatever frequency makes sense for your customers. Hosted services might actually release every two weeks. Smaller releases are encouraged as they provide more opportunities for feedback from the larger customer base."

Being Predictable and Transparent

Agile has many advantages over waterfall. In waterfall, you have the illusion of control. You typically have a project plan which covers the duration and show all of the key tasks that must be done. In reality, the project is typically much too complex to accurately schedule. For example, once you finish development, it is very hard to predict how long it will take you to find and fix all the defects.

In agile, you tackle things in much more manageable chunks. It is easier to predict how long items will take that are days long (rather than months). As for defects, these are kept in check as you go. It is much easier to diagnose / fix a defect if you know it was introduced within the past two weeks (rather than 6 months). With agile, you know you can make your dates. The question is what will the functionality be? In waterfall, both dates and functionality are in question.

Focusing on What Will Make a Difference

Perhaps even more important than the predictability gained is the productivity gains. Agile is all about prioritizing and tackling the most important stories. And then re-prioritizing every iteration based on what you've learned / changes to your marketplace. As a result, you might only get the first couple of stories done related to a particular set of functionality before you realize that it is enough (i.e., it is more important to do other things than to continue on that line of functionality).

In waterfall, you set the functionality up front. So at the end of the 12 or 18 months, you have the system that would have been a good fit for the market 12 or 18 months ago, not the market of today. Without the feedback loops, you tend to go deeper into things than you need to (resulting in unused features). Finally, because you're releases are so infrequent, users tend to pile on with their requirements since they know that if it doesn't make it, they have to wait a long time for their next opportunity. Agile's productivity gains are as much about what you don't do as they are about what you do.

Tackling the Challenges

Agile isn't without its challenges. It takes time to adjust to the new approach (and to get the infrastructure in place and working well to support it). Another challenge is that you have to re-educate your customers (internal and external) to think in terms of priorities rather than fixed feature sets. But agile is worth the cost. Once you do it and see how well it can work, you'll never go back.

Wednesday, April 29, 2009

Agile Methods 101 Bar Camp 2007

I had been looking for some core stuff on recent work done in method development area with a view to understand the evolution. This presentation from Diane Strode caught my eye for sheer simplicity and research basis. There are ton of other skillfully crafted presentations on "Agile" method(s) on Slideshare.

Sunday, April 26, 2009

QUOTE this!

This one is cool and inspiring ..both at the same time in one stroke ..

Saturday, April 18, 2009

SAP Sales and Workflow on Blackberry

Long wait but I saw this available on ecohub today. Not sure if its generally available or not . The Demo of Sales appeared rich in functionality (as usual) but lacking in interface (another as usual).

I had to fill up a long form to access the "Sales demo" which was nothing more than few screen shots. No info yet on the devices supported.

Worst part - You need to contact SAP Sales Rep for Pricing info !!!!!

Friday, April 17, 2009

Red Hot blog of the day


Shutri got the Red Hot Blog of the day , today .

Thursday, April 16, 2009

App Dev & Program Management - At Cross roads

PMI has a pretty much "be all one all" approach for any sort of Project Management endeavor. Be it a "Construction Project" or a "Marketing Project" or for that matter an IT project looking either for a new application development or implementation of a packaged solution.

However , things have been changing fast (as expected) when it comes to managing a project in technology space. I think we have interesting time ahead to see how the old project management practices handle the onslaught of Agile method.

I recd an interesting mail from Forrester this morning .. here are the excerpts ...

1. Successfully adopt Agile software delivery methods. Those old processes are choking you, and while many organizations want to become more agile, that shift requires a change in thinking, which isn't always easy.

2. Take a new approach to project management and business analysis. We cling to tradition when it comes to these software delivery practices, so weeding this part of the garden can be difficult. But we know that the old approaches aren't working as well as they must.

3. Optimize the effectiveness of your quality assurance practices.

4. Consider product-centric, not project-centric, approaches. Many traditional IT firms take a project-centric approach to delivering software. They manage the projects, but not the product: the software itself.

5. Re-evaluate and simplify your application platforms. Your application portfolio is most likely as full of weeds as my garden. But the platform landscape is changing, and your choices for efficient, effective application platforms are increasing.

Experiments

Quanta

Loading...