MathWorks

Great working environment. Everyone has an office.

People respect and invest in the success of the company and each other.

MathWorks is financially very strong and stable. For a decade, it has been consistently growing.

A lot of interesting ideas and advances in the technology. Every new hire can learn a lot, whether college grads or veteran developers.

Considering that the main products were created in the 80s and early 90s, it is really amazing that there are still many innovations happening in the products.

Deeply penetrated markets. Well-respected in the industry and a lot of loyal customers.

Good work/life balance.

Good communication with customers. Customers love the products, and they are helping to shape the future of MathWorks.

If you want to find a good example of Jim Collins' books, this is the place. It shines in quite a few areas that Collins highlighted in his books "Built to Last" and "Good to Great". For example, the 'cult-like culture' is an intentional result of years of cultivation.

Overall, the main markets are R&D and education institutes for MATLAB and control design automation for Simulink. These are small and well-established markets. MathWorks products have a very strong standing in these markets, but the market itself is not growing rapidly. This means relatively weak salary and slow financial growth for individuals. With the stakeholder bonus, the total income is average in the design automation market.

Another typical pitfall for tech companies from the 80s is that there is no career growth outside the management track. If you are good in a technical area, you are expected to become a senior team lead (STL) and then a manager. It is impossible to stay in the engineering track. As a result, quite a number of really good engineers have to leave the company to pursue technical advances.

The weak mid-level managers is another result of the same problem. The mid-level managers were good engineers 5-or-10 years ago. They are in a manager role now, but some are weak at leadership skills and try to play the engineering role by making critical technical decisions and try to squeeze time to do coding works. This is bad for both parties; it discourages the engineers, creates conflicts, and even makes it hard for the managers to grow in their management role.

The engineering levels and management levels were invented when the company was still small and no one had that many years of experience. They left very little space to grow when the company became big and people became older. As a result, it has become harder and harder to get promoted because there are not so many levels to grow and there are so many people who need the promotion.

Over 10 years ago, it was decided that MathWorks will follow 'the Toyota way' and then became a process-oriented company. "There is a process for it" has been the trend ever since. For a company currently with 3K employees, it feels as if the company were 10x its size. There are a lot of lengthy processes and useless meetings, mostly because the processes were invented by the managers. A lot of people/parties need to be involved for very trivial issues. Maybe someday the executive team can realize that the Toyota way was born out of Japanese culture for production systems, and US software companies really need to customize it for their own needs.

Over the last few years, the company has gradually fallen into the pattern of Collins' other book, "How the Mighty Fall." For example, most teams have lost their focus. (If every team has more than 10 focus areas, is it still appropriate to call it a 'focus'?) Luckily, the top executive team seems to be aware of the issues and are improving it.

Create good career paths for top engineers to advance their careers; do not force them into management roles.

Here is another idea: when many people have 10-20 years of experience, invent new levels so that people can grow into.

Think again about fully embracing processes. Sometimes it may not be a good idea. Try it on a small scale and listen to feedback.