Saturday, May 19, 2007

Challenges in Building Products

I just completed 7 years in the software industry in various roles - Implementation Engineer, Technology Evangelist, Software Engineer, Architect and Product Manager. In these roles, I have worked on software projects of varying complexity, maturity and in teams as small as 3 Engineers and in bigger ones (40 Engineers). It is interesting to look at some of the major challenges faced by these products at different stages in their life-cycle. To set the correct context, I must add that I have worked in the Enterprise Software space for most of my career.

Mature Products
: I started off as a Research Assistant in the Condor team and my first big project was to understand the requirements of the UW Hi-Energy Particle Physics (HEP) computer simulations and modify our Grid Computing software - Condor - to support the HEP simulations. Condor is one of the longest running university research projects and even in 2000 it was very mature with excellent documentation, user groups and large user base. During my stint with the Condor team, I learned that even mature products may require customizations and also about the importance of having a strong Implementation Engineer or Professional Services Team supporting the implementations and funneling back requirements to the core development team. If you are considering developing Enterprise Software, make sure you factor in travel and support costs of such an Architect or Team. Depending on the maturity of your product, these costs may be as much or even higher than the budget for your core development team.

A disappointing fact about Condor (at least to me), is the fact that it missed many opportunities to move into a mainstream market. Based on my interactions with Prof.Livny, I am nearly certain that he never intended to build a business out of his research project, but the entrepreneur in me rues the missed opportunities :( I must add that I tried mightily to evangelize this product when I worked for Prof.Livny and also when I worked at Optena.

Mainstream and Extremely Successful Products: Tom Siebel built his company and an entire market on his idea of Customer Relationship Management. Granted, he may not have been the first one who thought of this, but he brought an extremely feature-rich and useful product to this market and through aggressive sales and acquisitions built Siebel Systems into a well-known ISV in the space. Though I joined Siebel Systems only in 2002, I was friends with many of the original architects who started with Tom and my observations are based on my conversations with them and also my own experience working at Siebel Systems from 2002-04. Siebel's biggest challenges (in the post-2000 era) were -
  • Managing Growth and Product Quality- After a spectacular IPO, Siebel lost the wind in its sails following the 2000 bubble burst. The HR team definitely did not do Siebel any good by hiring (and firing) in what seemed to me to be a very random pattern of behavior. Maybe it was the fear of a wardrobe malfunction, but when Siebel made the switch from a Client-Server model to an Internet based architecture by introducing version 7 of its product, Ms.Quality Under-appreciated never attended the launch party (or any other parties - maybe I should say funerals - thereafter). I have heard horror stories of how Siebel 7 would not even install off of the CD that was shipped.
  • Implementation Costs and ROI - Bad Product Quality meant that customers had to hire pricey consultants to install and configure the software. It was not uncommon to hear about a customer spending $5 million in software acquisition and a further $15 million in deployment. This is just plain wrong - no matter, how good a product you have, if your implementation costs are not less than 20% of the sticker price of the product, you are going to lose the market - Period.
  • Lack of Faith - It is not uncommon for even the largest and most innovative companies (or individuals) to go through a bad patch, but the differentiator of winners and losers is often their perseverance and ability to work through a bad patch. It is in this aspect that Siebel Systems failed misearbly. When Siebel realized that hosted CRM vendors like Netsuite and Salesforce were drumming up a vigorous business, it decided to reintroduce its hosted offering (it may interest you to know that Siebel was one of the first players in the hosted CRM space with its offering, which it withdrew because of lack of business!). It also brought in a senior IBM executive - Mike Lawry - as the CEO to correct course and build new businesses. However, it never had complete faith in these (as well as many other) initiatives and ended up being swallowed by Oracle. In the same period, has seen a steady increase in its market-cap and has hired many of the ex-Siebel architects and engineers.
New Products/Company: This is one of the most painful chapters in my career. I quit Siebel and joined Optena as the first member of its engineer team. Optena's vision was to commercialize Condor based technology. Based on this experience, I now know about the following pitfalls while starting a company -
  • Beachhead Customer - The importance of the beachhead customer, who is willing to experiment with a very new and potentially immature technology is extremely important to a product's success. In the absence of such a customer, the product gets built in a vacuum and there is no way of prioritizing or even verifying whether the features are useful or not. Insisting beachhead customers to pay you may not be the best option if you realize the risk they are taking by using an immature product in production. If they are willing to pay, only charge them the fair-market price for the value and not what you expect your product to sell for 2-3 years down the line when it has established itself.
  • Innovation - Small companies must keep innovating, even if the beachhead customer wants only a few minor feature enhancements to the first version of the product. If you focus yourself on the enhancements required by the first few customers and stop your innovations, you'll end being a one-trick pony and will never be able to take the product to the next level in its lifecycle.
  • Hiring Successful Entrepreneurs - Optena lost its funding just when it was about to get a couple of major deals. The VCs realized that even with the new deals, Optena would never be a billion dollar company (this was shortly after Google's IPO). I daresay the VCs would have continued their funding if we had even one person on our rolls who had a successful trackrecord in building companies.
The interesting aspect here is that only one of the issues here was a real product problem with the others being strategic mistakes made by us.

Emerging Products: Our biggest challenge at Solidcore is ....

Sorry, I cannot talk about my role or the product at Solidcore as we are in an intensely competitive market that is growing rapidly. I promise to keep you posted on our progress.

No comments: