U.S. flag

An official website of the United States government

Dot gov

Official websites use .gov
A .gov website belongs to an official government organization in the United States.


Secure .gov websites use HTTPS
A lock () or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Through DoD’s Valley of Death: A Data-Intensive Startup's Journey


  1. Home
  2. Through DoD’s Valley of Death: A Data-Intensive Startup's Journey
Through DoD’s Valley of Death


The often discussed “Valley of Death” is a phenomenon faced by many startups when trying to do business with the Department of Defense (DoD). The Valley of Death is a journey (typically one to two years long for survivors) where a vendor transitions a prototype or commercially available product to a DoD contract. There are many reasons why this period is hard, but a few are expressly applicable when considering Use Case development for a Data Intensive Startup. 

A Data Intensive Startup is any vendor seeking to generate value for DoD through the collection, aggregation, or transformation of data. This term applies to traditional information technology (IT) firms, but also includes a growing percentage of non-IT firms that leverage data and information connectivity in their value chain. A large number of traditionally hardware-based firms equip their machines with data-collection and -reporting devices—for example, a sea buoy manufacturer that equips its buoys with data-collection and -dissemination devices. But even commercial software and data firms often are unprepared to deal with their first DoD line of business. 

DoD has prioritized an integrated data strategy that intends to guide operational concepts and technical solutions toward capabilities that enable commanders to make decisions at the speed of relevance in today’s information age. As a part of this strategy, DoD faces the challenge (and opportunity) of integrating a number of high-quality, inexpensive commercial (i.e., GPS) data feeds with expensive, classified (government-only) data sources (i.e., military satellites). 

A growing number of ventures are exploring technical solutions that present sensible answers to one or more of DoD’s stated data problems, but many do not know or fail to consider the strategic milestones along the Valley of Death. Let us first explore a few business-related milestones that are often ruled by a vendor’s available resources. Then we will look at several operational aspects that will increase relevance of a given Use Case developed by a Data Intensive Startup.

Resource Constraints Versus Business Strategy

To ATO or not to ATO

Data startups that aggregate commercial or publicly available unclassified feeds provide a valuable service to DoD, but a critical business strategy decision involves how this service will be delivered contractually to DoD once the technology reaches maturity. An initial decision is made whether the company intends to maintain a subscriber-based service via a web client or pursue software installation on DoD end points. 

The benefit of maintaining a web client subscriber-based strategy is that it enables the startup to avoid the lengthy Authority to Operate (ATO) certification with a relatively undefined timeline. This subscriber-based strategy provides optimum agility for the startup and reserves vendor resources to pursue non-DoD clients in parallel. Also, it allows the vendor to proceed with software updates without needing customer approval. However, DoD typically funds subscriber-based services in annual cycles with discretionary funding, which makes the subscription model inherently vulnerable when crossing the budgetary Rubicon. Also, this strategy confines the service to end points with a connection to the open Internet and limits website development within DoD’s network management access requirements. 

If the vendor pursues an ATO on DoD networks, the startup may be able to leverage several long-term advantages. ATO generally conveys legitimacy of a product due to the information assurance measures performed in any ATO review. Also, ATO typically enables enrollment into approved catalogs for installation on government endpoints, which raises the potential for success in marketing to other government agencies. Typically, a separate ATO process is needed for each tier of network classification. Aside from the extended timeline, software updates typically require some government oversight for the update, and the deployment of the update must occur in cooperation with the government networks’ resource management teams.

Figure 1. Interagency Crisis Information
Figure 1. Interagency Crisis Information

Source: Joint Publication (JP) 2-01, Joint and National Intelligence Support to Military Operation
Cross-Domain Strategy

Startups that aggregate commercially available information must maintain a connection to those original data feeds in order to continuously move those individual data feeds up the value chain. DoD maintains many unclassified machines with open Internet access where this can be possible. However, unclassified networks may or may not be the primary network for communication and planning for a given operational community. 

As depicted in contemporary military movies, an operations center usually has a large number of computer terminals. One reason for a large number of terminals is the ability to maintain connectivity with multiple levels of classified networks. Figure 1 provides a notional example of how U.S. agencies exchange information across multiple networks. Figure 2 provides a notional example of how intelligence is shared between the United States and partner nations in a multinational intelligence architecture. 

A “cross-domain solution” (CDS) involves software and/or hardware-enabling data to move from one level of network classification to another. Aggregators of unclassified information need a CDS to ensure that their data arrives and is usable on higher-level networks. Below are two options that could be considered for a CDS that may or may not coincide with a vendor’s profitability strategy.

ATOs for Every Network

The most expensive and time-intensive strategy would be to pursue an ATO at multiple levels within the network environment. In this case, an unclassified instance of the vendor software would aggregate unclassified data sources and send the consolidated data package across the CDS. The data would then be opened and manipulated on the classified network by another instance of the vendor’s software. Given the ATO journey discussed previously and the need for contracts on each network, significant business risk is introduced with this strategy. Also, the vendor must ensure that its software version management strategy supports an environment where a newer version of software installed on Network A remains interoperable with the older version on Network B.

Figure 2. Multinational Intelligence
Figure 2. Multinational Intelligence

Source: Joint Publication 2-01, Joint and National Intelligence Support to Military Operations
Interoperable Data

The strategy most desired and consistent with DoD’s Data Strategy would arguably be for the vendor to provide data to the CDS in a common file format that could be ingested by existing applications on higher-level networks. For example, a vendor that aggregates geospatial data could export its resultant data in a Keyhole Markup Language (KML) file format. Many popular commercial and military geospatial applications can ingest KML data files, which means that the vendor’s aggregated data could be more easily consumed by the suite of available software at each tier of classification. Structuring data in common file formats would also help the resultant data products to publish in DoD virtual knowledge bases like the one illustrated in Figure 3. The downside of this strategy for the vendor is that the service remains in the unclassified domain and may not earn the name recognition that it deserves. Absent name recognition, the user community may lose familiarity with the service in the presence of competing service providers on classified networks.

Figure 3. Virtual Knowledge
Figure 3. Virtual Knowledge

Source: Joint Publication 2-01, Joint and National Intelligence Support to Military Operations

The Best Available Versus a “Program of Record

The largest and most enduring budgetary line items in the DoD are categorized as Programs of Record (PoRs). Congressional budgets allocate major funding in support of PoRs, and PoRs usually represent efforts that span multiple years. 

Even though Congress goes through the authorization and appropriation process each fiscal year, the lawmakers do so knowing that the PoRs must be sustained across their lifespans. Budget line items that exist outside of this construct have numerous ways of receiving funding, and DoD has been increasing the number of investment vehicles available to deploy funding to emerging capabilities (e.g., Other Transaction Authorities (OTA)). Many startups begin work with DoD via OTAs or some similar financial vehicles, but history shows that long-term financial stability for DoD contractors typically comes when a vendor transitions to a contract related to a PoR or is acquired by another contractor who is. 

The most direct and disruptive way to inject a new data capability into established PoRs is to associate the Data Intensive Startup’s offering with published program requirements. Established PoRs not only have published requirements but should have measures of effectiveness (MOE) and measures of performance (MOP) associated with those requirements. Given these knowns, the Data Intensive Startup can concentrate business development efforts on PoRs that have a demand signal and persistent funding to purchase the offering. Also, by knowing the requirements’ MOE and MOP, the Data Intensive Startups can optimize resources to provide a product significantly better than the status quo.

Operational Relevance During Use Case Development

Data Intensive Startups must articulate the value of their data-oriented offering. A large percentage of startups are solution (or technology) focused. In simplified terms, startups often focus on the technical quality of their offering rather than the value generated when that offering is placed within its military context. For example, a Data Intensive Startup that can aggregate unclassified data streams to generate insights on a particular geographic area should be able to map the importance of those insights in multiple domains.

  • Are the insights relative to multiple domains such as military, political, economic, social, informational and/or infrastructure?
  • Do these insights relate to an ally, host nation, or adversary’s Center of Gravity?
  • Would engagement at any identified point(s) result in a decisive outcome?

Figures 4 and 5 offer two examples of how Data Intensive Startups can communicate value in recognizable ways for DoD stakeholders. 

Data Intensive Startups should be able to articulate what their product or service provides by using relevant DoD terminology. While DoD often adopts jargon developed by industry, startups should perform some due diligence to ensure that they speak to their potential customers in relevant terms.

Figure 4. Centers of Gravity
Figure 4. Centers of Gravity

Source: Joint Publication 2-0, Joint Intelligence
Figure 5. Systems-Oriented Template
Figure 5. Systems-Oriented Template

Source: Joint Publication 2-0, Joint Intelligence
Show the Value

Data Intensive Startups should be able to articulate when their product or service is valuable to the Warfighter. Joint Publication 2-0 provides a robust discussion about how DoD performs Intelligence planning. Data Intensive Startups should be able to explain their Use Case in at least two ways with respect to timing. Data Intensive Startups should understand how their value addition changes based on the particular phase of conflict and whether the normal value delivery pipeline is available. This is important because platforms’ operational behaviors and technical configuration change during different phases of conflict. 

Related to the time domain, the Data Intensive Startup team must be able to assess their offering against typically “timeliness” metrics employed by DoD. Many analytical processes are able to extract deep insights from unstructured data, but it may take months of review by experienced researchers with specialized technical training. In order to find operational utility, a data service must be able to provide relevant insights in a timely fashion. The definition of “timeliness” can vary by context, but assuming that timeliness meets the minimum viable collection threshold, further investment in “timeliness” metrics typically demands a proportional response to “costliness” metrics. By contextualizing the timeliness trade space within the entire value chain (also known as “kill chain”), Data Intensive Startups can understand if incrimentally greater investment could yield a positive impact on DoD’s ability to “close the kill chain.”

In addition to knowing when a Use Case will be valuable, it is important to understand who on the team will be able to transform that value into effective outcomes. This not only adds detail to a Use Case, but it provides insights on the types of knowledge, skills, and abilities of the primary user.

  • Does use of the product require specialized technical training?
  • Does use of the product align to the existing job responsibilities of a given team member, or does the use of the product represent growth to already demanding duties?
  • Which agency or military Service has legal authority to act on the data insights (e.g., U.S. Coast Guard (USCG) has maritime law enforcement (MLE) authorities, and the U.S. Navy does not)?
  • Does the actor with appropriate legal authority require support from partner agencies to effectively intervene? (e.g., USCG MLE detachments often ride on Navy ships to optimize the USCG legal authorities and the U.S. Navy’s greater resources)

Finally, it is important to understand where the tool will be primarily used in physical space and in the digital architecture. Figures 1 and 2 were previously used to discuss the importance of a Cross Domain Strategy with respect to transiting a multi-level network structure. Additionally, these diagrams can help us analyze where the solution is used in physical space as well as in the digital architecture.

  • Does the product require connection to the open Internet?
  • Does the product require any unique or high-performance hardware?
  • Does the product require any third-party support software?
  • Can the product be used on non-U.S. Government computers (e.g., multinational partners, nongovernmental organizations, commercial firms)?


DoD needs a large number of Data Intensive Startups to satisfy the demands of the integrated data strategy. For ventures to survive the Valley of Death and provide disruptive innovation to DoD, Data Intensive Startups must ensure that their business strategies and product Use Cases equip them with what they need to thrive in the realities of DoD’s operational and technical ecosystem. When so equipped, solution providers will better articulate the value they add to potential customers and ultimately provide the Warfighter with data solutions needed for 21st-century warfare. 


Defense Acquisition Magazine January to February 2022 cover

Read the full issue of

Defense Acquisition magazine



LANDRETH is a U.S. Navy civilian engineer, acquisition workforce professional, and U.S. Navy Reserve submarine officer. In 2018, he was named one of the Top 10 Federal Government Professional Engineers by the National Society of Professional Engineers. He holds degrees from the U.S. Naval Academy (B.S.) and the University of South Carolina (M.Eng.), and he has completed Joint Professional Military Education at the U.S. Naval War College.

The author may be contacted at [email protected].

The views expressed in this article are those of the author alone and not the Department of Defense. Reproduction or reposting of articles from Defense Acquisition magazine should credit the author and the magazine.

tweetsubscribePrint Button