CMMI Institute

Newsroom

The latest information for media, analysts, and others interested in the CMMI® Institute and process improvement.

Capability Counts Series: Software’s Eating the World: The Best Software Company Wins

By Hillel Glazer, C-Level Mentor: Emerging Technology Centers and Performance Jedi, Entinex, Inc.

This title is based on an op-ed, “Why Software is Eating the World,” published several years ago in the Wall Street Journal. The author is a well-established venture capitalist. The “eat now” in venture capital is something that got me thinking of how my own experience has parallels to what's happening in nearly every company.

Organizations across all industries have to become software companies. Why? Because everybody wants an app or e-commerce portal. Many don't think of themselves as software companies. But whether they like it or not, they need to become software or product companies, which puts a lot of pressure on building product.

I've spent the last five years in start-up companies and being a student of how investors make decisions, especially early-stage investments. They rely on their investments to quickly gain traction in the market. However, instead of coming up with a perfect idea from these companies, the capitalists expect them to crank out idea after idea and experiment with what works and what doesn't. They're looking for rapid feedback, learning and adjustment in their investments.

It turns out that this same way of looking at traction fuels the data that cutting-edge product developers need. Investors and developers look at things from opposite ends of the same equation, often not realizing they were connected in the middle with an equal sign.

Investor economics and product development economics are effectively the same, and the CMMI community will greatly benefit from looking through the lens of the economics of what investor communities are looking at from software. CMMI professionals will also benefit because they will be able to more clearly articulate their value proposition, economically and from a business perspective.

Support Calculated Risk and Failure
To help reduce time-to-market, measure all things to do with waiting and then attack the causes of that waiting. For example, see how many things are currently in process and what they are waiting for, how long they have been waiting, and why are they waiting. Any time they need to go upstream in a workflow, they are going the wrong way. Going upstream means the product is not moving toward the market. If there’s any competition whatsoever—and often that competition is nothing more than consumer attention spans—going upstream means delay. Every delay is time lost, and time lost is lost traction. Ultimately, lost traction is lost revenue.

Long before Apple solidified its fan-base, they would capture market share by merely shipping products and updates sooner than anyone else. Nokia, by comparison, was so much slower at responding to the market that they were nearly wiped out as a company after sitting atop the mobile phone market for years. And, not until they retooled their engineering practices to allow them to crank out model after model was Samsung able to mount a defense against Apple’s product delivery onslaught.

Also address risk tolerance. Instead of taking calculated risks, we are spending a lot of time analyzing what we think we want to do instead of analyzing actual data. We need to analyze performance data of our own process capabilities and how well these ideas are meeting customer or user needs. Gathering data relies on putting out a product. You can measure what it took to put it out, and then you are able to measure the response from the market.

We often don't know until it's too late whether our idea is going to work in the market because we are not getting real feedback. Since we have become averse to putting out products that may or may not work, we are not generating the data that we really need! You can see this is a vicious cycle.

Allow an environment of calculated risk and for that environment to be tolerant of failure, as learning comes from failure. It's not just random failure or throwing things at a wall to see what sticks. It is educated design and disciplined experiments. Provide people the freedom to fail and the safety net of knowing that failure is ok—because it’s the only way we learn.

The good news is that it is relatively easy to correct. Personalities may need to move around so the learning and the failing can happen. There is certainly only so much failure an organization can tolerate before things completely fall apart.

Educate, But Don’t Overwhelm
Spend more time educating everyone and getting stakeholders to understand what we are doing, how we are doing it and where we are heading—but know they may not be ready to take on the entire journey. Organizations must understand why they are doing things that seem as if they are not making progress.

Everyone wants to make progress faster than others, and this desire can often be faster than the physics of the situation allows. Sometimes to keep people interested and motivated, we need to stop. I work hard to avoid the trap of letting them peek behind the curtain—doing so might not work out well because they are not ready. They've been so focused on what they do, they are not concentrating on performance and working out the product development flow. When we begin to implement performance measures into the workflow, it can be very overwhelming.

Try to not feed the entire process all at once to an audience that is not ready to absorb it. Besides overwhelming them, it can backfire to where they decide they don't want to do it. Be cognizant of what the company is doing, the organization of the team, where they are and helping them not overfeed on what you are doing. You want them all to have it, but if you allow them to get too far too quickly, then they don't have the foundation on which to make decisions down the road.

Moving Forward
A lot of these ideas come from the commercial market, but they still apply in non-commercial or regulated spaces. These performance jumps are achievable in heavily regulated, severely constrained and highly technical companies.

They help identify what to do despite a lot of limitations that are not imposed by anything they have any control over. For example, there are certain things you can't change, such as regulations. You are not going to get out from having a bleeding edge where you are literally not sure what you are building because you don't know how it's going to work yet. But you have to start someplace!

These challenges are frequently found in industries with significant external oversight. There are many eyes looking at things and there's not a lot of leeway to change the rules. These constraints are often used as a crutch and some say there is nothing you can do here. I don't believe that's true. It makes it more of a challenge to come up with a dynamic solution to a complex situation.

Tie CMMI to Business Decisions
CMMI can help improve organizational performance and we must tie this to things that business leaders make decisions on. Companies can lose sight of how the performance of the practices tie to business questions and needs.

CMMI can help expose process performance that tie to operational and business performance. This could then be extended to tie to risk. An argument can be made that lacking engineering data and feedback mechanisms that could have informed decision-makers, by focusing only on processes certain parties to the matter plowed through many warning signs designed to prevent exactly what happened.

We must think of things in terms of business systems instead of processes and make it easier for executives to understand the value of what we do on the process, operations and performance side of things. This increases our value and credibility. Speaking for the business and the profit contribution to this in economic terms is the best path to succeed with CMMI.

Software-s-eating-the-world-v2.jpg