The Myths and Realities of Outsourcing Engineering
Steve Szirom
I have been a keen observer of the trends in the IC industry which pertain to the current trends in outsourced engineering and design tasks. By outsourced engineering, I am referring to such products and services as IP cores, FGPA cores, and IC design. The management trends in these areas are being driven by California companies. More than half of all IP companies that have been founded in the recent few years have been founded in California.
The management of an IP company and that of a design house are somewhat similar. The classical design house CEO is coming from an engineering background…someone who has a vision or just a smart engineer who wants to run his own show. His management expertise is probably strong engineering project management and more limited in the areas of business development, marketing, and other business related tasks. He works in a regional or a national customer environment where being close to the customer has its inherent benefits. The CEO of a classical third-party IP provider may also come from an engineering background but in this company model he has to have a much better mix of business and engineering skills. He will be dealing with a product (an IP core) that will be resold to many customers (hopefully). The customers are all over the world with strong emphasis in the USA, Japan and the Pacific Rim regions.
In the case of these small (10 to 50 person) IP and design service companies, the role of the CEO becomes of paramount importance. In these types of companies the key productivity asset is the design expertise of each engineer. Every design-house company can buy the EDA tools, hardware and software, but the distinguishing variables are the CEO capability, the company process and culture, and its engineers' brainpower inventory.
In a large company, the engineer is far removed from the CEO in the organization chain. The dynamic of a large company's engineering management is driven by influence of many department engineering managers. The sum total of all these different managers all adds to the bottom line at the end of the fiscal year -- it is not dependent on one CEO but on many department manager "CEOs"
In the case of these small entrepreneurial IP and design service companies, the role of the CEO is the most telling success factor. Let's take a look at some of his influences in the small enterprise:
- Experience in development, operations, marketing, and finance, such as business unit director of mid-large chip maker
- Background (fast growth in resource constrained environment)
-
Managerial team building skills and ability to delegate
-
Intelligence and high energy
-
High integrity, quality, strong work ethic environment
-
Openness in internal and external matters
-
Moderate ego and humility
-
Coach and head cheerleader
In this column, I will focus my comments on the managerial team building skills and his ability to delegate, i.e., to create a "process". By the process, I basically mean that system which finds customers matched to the companies strengths, that assigns the best mix of engineering talent to the design project, and that makes the efforts of the team collectively engineer a correct solution on time even with inefficiencies in the process. The company "process" is as set of tangible and intangible set of procedures, rules, business methods, and company culture which is never perfect.
If we look at today's Silicon Valley's engineering community, the recent trends have shown that these engineers (the key revenue generating intellectual assets) are known to be more loyal to their industry than to their individual employers. Part of this is due to the short-term financial views taken by electronics firms (layoffs dictated by quarter-to-quarter financials). But another factor is also inherent in the entrepreneurial and creative forces that exist only in Silicon Valley -- the world's foremost skill center in silicon IC design.
One of the main reasons why engineering outsource companies have difficulties is due to their inefficiency in staffing their design projects. Clearly, the challenge for the CEO is to answer some very fundamental questions regarding the motivations of his most important asset -- his engineering team. Why does an engineer want to work at my company and why will he go through superhuman engineering hoops to deliver the product correctly and on time even in a system with an imperfect process.
Here are some broad guidelines that will help to improve the process at IP and design-houses:
-
Create a strong culture and foster teamwork
-
Clearly define roles of the "architect" engineers and "back-end implementor" engineers; keep the architects involved with customers
-
Recruit several key superstars and implement a cycle of progressive training (hire college graduates, train during apprenticeship program, discover talents and assign more progressive roles)
-
Harness the experience of the experienced engineers by letting them conceptualize new products and improvements
-
Constantly monitor and re-invest in the design center infrastructure as newer foundry technologies render older version of tools obsolete
-
Have a good system for valuing the contributions of the engineers to the bottom line (pay appropriately)
-
Treat engineers as "professionals" where post graduate degree, apprenticeship, on-going training, and specialization is required
-
The outsourced engineering concept is here to. The IC industry's enabling technologies, such as fabrication processes and EDA tools are becoming more and more commoditized. Entry barriers will be lowered.
Companies will be spheres of excellence in various niches-- the value-added concept. The companies' efficient processes, knowledge of markets and customers, and skills in managing employees (heavily engineering oriented) will be determinants for success.
|