Thank you!

This advice for working on the right projects (and jettisoning the wrong ones) comes from Richard Daly, CEO of global genomic data processing company DNAnexus, which has seen its revenue grow 300% in the last 24 months — largely because of this strategy.

Train your company not to play with Dead Snakes.

What's a Dead Snake? Former Netscape CEO Jim Barksdale famously called problematic ideas 'snakes' and warned against Dead Snakes in a speech at Harvard. But in the context I'm talking about here (and to build on his ideas), it's a project or feature you’re building that's going nowhere. It might have looked very much alive. There might have been significant interest from a customer. It might have played a central role in your last board meeting presentation.

But it's dead and a distraction. A waste of time and money. A Dead Snake.

Dead Snakes can drain resources and distract you from the most important work your team should be doing. To gain an advantage, you have to learn how to recognize and dispose of Dead Snakes by:

The earlier your company can identify, avoid, and dispose of Dead Snakes, the faster you'll be able to build new products or features your customer actually wants, and grow your revenue accordingly.

DNAnexus, has generated our share of Dead Snakes. Over the last two years, we've deployed a new approach called rapid-cycle agile customer engagement to overcome this challenge — talking to more customers more often, and building new features for them faster than ever before.

Our market is high-velocity. Underlying genomic science is advancing every day, customer needs are rapidly evolving, and the software and computer architectures required to address these needs move fast. Keeping pace with — and even driving — evolution in the life sciences market has been critical to our competitive success, particularly when a competitor is a large enterprise. These competitors in our space have year(s)-long development cycles. Driving a short product development cycle with agile customer engagement has allowed us to iterate to meet rapidly evolving customer needs and maintain leadership.

Our strategy: Engage on a deep science-to-science basis with the customer, “acquire” the customer problem and run pilots in an agile environment to co-evolve. Acquiring the customer problem means sharing the task of finding a solution to the fundamental customer issue, and collaborating on a science-founded answer. This morphs the typical sales cycle into a hybrid development/sales process, permitting more rapid customer acquisition.

The upside: We've been able to attack on many fronts simultaneously, experiment with different features, and ramp up revenue growth. Deployment speed has become our primary advantage, which is key given how fast the genomic landscape evolves.

But we also discovered a major downside: Our Achilles’ heel became a proliferation of dead-end, N-of-One projects, which we dubbed Dead Snakes. An N-of-One project is a development effort valuable to just one customer, but not broadly applicable or reproducible. We learned how vital it was to “fail early” and drop these initiatives as soon as we could. Over time, we created a guide to identifying and avoiding Dead Snakes, as we've defined them.

Recognizing a Dead Snake

Dead Snakes take many forms — and, even though they're dead, they can still be dangerous. Here are the primary signs and triggers we've learned to watch out for when evaluating the performance and value of a project:

In our experience, about 30% of the opportunities that looked good initially turned out to be flops for one of these reasons. I'd say this estimation is typical at most B2B companies.

Building Your Team of Dead Snake Hunters

The best antidote to Dead Snakes is constant communication — regular, borderline-overbearing repetition. We have a couple of best practices in place to make this habit.

Weekly Letters

Each direct C-level report writes a letter that is due to me no later than 5:30 a.m. every Sunday morning. No letter should take more than 20 minutes to write, and it must include the key things the management team needs to know.

For example, our sales lead might share revenue data and target accounts. Engineering might talk about major projects and major blockers. There's no template or requirements, it's whatever information each manger believes is vital for coordination and prioritization. On Sunday, I get up, have two cups of coffee, combine these emails into a compendium, add my thoughts, and send it out to everyone by 7:30 a.m. It's an idea I snagged from VMware Founder and now Head of Google Cloud Diane Greene, and it's made an outsized impact on our ability to act fast as a team.

The opening I write for this weekly roundup sounds a lot like a short op-ed, detailing the most important opportunities, risks, and challenges we should be focused on. In fact, the first iteration of our Dead Snakes Guide started out as the top of this weekly 20-minute report. We wanted the frequency of Dead-Snake projects to be a “code-red” issue for the whole team because they were really affecting our overall abilities. The term was catchy, got expanded, and is a growing touchstone of the DNAnexus culture. Today, many of our Dead Snakes are first identified and dealt with in this weekly email.

At our staff meeting on Monday, we get to talk decisions and actions, and avoid going around the room for updates or status reports because we already have that information. The process took about a year to fully implement, but we now rely on it as a primary method for running the company. Ferreting out Dead Snakes is a core component of what we share.

With revenue scaling rapidly, we've had to level up our financial planning. We've dug deeper into customer acquisition costs, resource allocation, margins, etc. — and that's really shone a bright light on our snake population. If a project or product sponsor can't say, “Here are the resources I'm going to need, and here's exactly how they'll drive revenue,” that project is probably a Dead Snake.


If you're a founder or CEO, this meeting is the cornerstone of your authority and ability to achieve your goals. Every single time, you need to:

Thematic repetition is foundational. These elements must be interwoven throughout your employees' experience at the company, and you only get so many chances to articulate them clearly and explicitly.

At All-Hands meetings, make these components memorable with numbers. Share quantitative data and render it visually so it's easy to grasp. Try for complete transparency and clarity, because any project or feature that doesn't positively contribute to one of the four elements above is a Dead Snake.

Constantly pushing the culture to “fail early” and dispose of Dead Snakes is important and takes time, focus and repetition. But it’s crucial to maintain high velocity.

Every employee needs to be empowered to propose shedding non-productive or formerly productive initiatives.

In practice, speaking truth to power is very tough, but attainable over time with constant focus and management commitment.

After a while, you can condition your team to have a sixth sense about Dead Snakes. Projects will get initiated and enough people will know the standard triggers or have the muscle memory to kill ideas that won't drive revenue early. This is the flow you want to get into. That's how you gain speed as a team. That's how you become a blur your competitors can't pin down.

Avoiding and Disposing of Dead Snakes

To stop creating Dead Snakes in the first place, we made it a whole-company objective to have specific success metrics, timetables, commercial impact analysis and required resources outlined upfront. You want this plan to be laser-focused around a specific account that you want to win or a critical market objective — a single point of attack. Sometimes after this exercise, it's clear the snake is dead on arrival, and we can sidestep it completely.

The ideal is: when we drop a project, we drop it. Shutting down “fail early” initiatives remains a constant challenge, and requires continual reinforcement as a strategy. The advantage is the ability to fully re-deploy resources early. To do this, we’ve found that initiatives have to be totally shut down. Merely “trimming” is a failure mode in itself.

To really kill a snake, you need to sterilize the earth under it.

If we don't, we have found they have a tendency to turn into Blood-Sucking Zombie Snakes, taking up time and energy for way too long. Some PowerPoint presentation slides can become immortal long after they fail to describe successful initiatives.

That's why it's so important to celebrate the scorched earth left behind — rewarding the people who spotted and disposed of the Dead Snake early. “Great try,” we'll say. “Now let's get busy trying something else.” We won't just say this once. As the CEO, I'll be sure to mention it during the next All-Hands meeting and make it clear that there's no blame for failure.

At DNAnexus, it's my job to constantly hammer on the ultimate goal to maximize revenue, and defend the strategies that get us there — like backing out when new things don't work. Remember, at the end of the day, bandwidth will always be your most precious resource, and Dead Snakes are its greatest threat.

First Round ReviewThe best advice in tech
Subscribe for Exclusive Access