How to Unseat a Federal Incumbent Using Data Analytics

As Government Contractors go through the bid/no-bid decision process, organizational stakeholders analyze, measure, and weigh critical factors to determine if a winning strategy can be formulated and executed. One of the most significant factors to consider is the incumbent profile. Established incumbents performing on an existing contract are often believed to have the upper hand in the federal procurement game. An established incumbent brings institutional knowledge, existing federal stakeholder relationships, experienced on-site personnel, and leadership buy-in – all presumably resulting in a lower risk transition to the federal customer.

But what about re-compete opportunities where the incumbent is, quite frankly, failing? Unseating an incumbent with poor performance is an ideal scenario for any Government Contracting growth and business development professional, but typically this is only identified through the rumor mill. Contractor Performance Assessment Reporting System (CPARS) is source selection sensitive (meaning it is off-limits to Government Contractors) and getting Contracting Officers to spill on their experiences is a pipedream.

So how do you pick up on data signals that a federal customer is looking to replace their current contractor?

With CPARS being off-limits, Government Contractors must find alternate routes to the required incumbent data, so in comes FPDS-NG, USASpending,gov, and beta.sam.gov. Federal spending transparency laws makes federal contract transaction summary data available to the public through these systems after federal agency representatives manually enter information in data fields such as “incumbent contractor name”, “socioeconomic status”, ‘action dollar obligations”, “contract values”, “work location”, “type of work”, “contract award date”, “date signed”, and “period of performance”.

This data is dispassionate, factual, and does not express opinions about contractor performance. However, by analyzing circumstances of an incumbent contractor’s “awards” and “current state” a Government Contractor can infer whether the federal customer is likely satisfied or dissatisfied with performance.

Award Dates. It is no secret that there are seasonal trends in federal contract spending and that federal contract spending spikes during the last week in September (the end of the federal fiscal year). This tradition has long sparked debate on if these expiring budgets lead to wasteful spending, and in 2017 academic economists conducted a study that concluded that fiscal year end deadlines for signing contracts may incentivize a “rush to spend resources on low-quality projects at year’s end.”

In this study, the authors found an increased prevalence of lower quality ratings, schedule delays, and cost growth for IT systems contracts awarded in the last week of the federal fiscal year as compared to the other. These projects are genuine agency needs but may suffer from issues such as poor requirements definition and limited acquisition planning.

Further, while not mentioned in the study, new contracts awarded at fiscal year-end can face longer background investigation timeframes due to a deluge in new background investigations. Why does this matter? Contract awards occurring during the fiscal year-end “busy season” cascades into a high volume of background investigation requests. When background investigations take longer it impacts schedule and contract performance is delayed.

The chart below shows this phenomenon occurring within a dataset of 258 contractors on an IT service contract.

Firm-Level Data Points. Research published by the Naval Postgraduate School (NPS) identifies firm-level data points associated with an increased risk of contract termination or performance/integrity issues. In this research, firm-level risk characteristics were assessed against the entire portfolio of a particular vendor’s federal prime contract awards, as opposed to a single contract award. Another, NPS study identified other data indicators of contract performance risk. For example, Cost Plus Award Fee (CPAF) and Cost-Plus Fixed Fee (CPFF) contracts are significantly more likely to have CPARS failures. Also, high Contract Specialist workload correlates with an increased risk of poor contract outcomes according to CPARS.

A Government Contractor’s instability in the federal ecosystem, newly formed companies, and diversifying beyond prior relevant corporate experience into new functional areas, were found to heighten the risk of performance issues. Further, unpublished research found other risk factors such as:

• A Government Contractor’s percentage of dollars obligated in September (i.e., the percentage of dollar obligations received in September was calculated for all federal contractors. Higher percentages correlated with increased contractor risk. Some contractors receive all dollar obligations during September.)
• A Government Contractor’s overall de-obligation (i.e., work cancellation) percentage

By leveraging USASpending.gov, a Government Contractor can filter a recipient’s activity by fiscal year and view descriptive statistics in different categories such as a recipient’s top awarding agencies.

Being able to see data from all federal agencies give a government-wide perspective – but what hidden data points might you be missing?

Let’s use this question: “Did a Government Contractor’s federal business suddenly quadruple due to different agencies making large contract awards at a fiscal year-end?”

If you’re data-savvy, you can use scripts/code to calculate firm-level risk metrics within the USASpending.gov enabled by the bulk data downloads. The below risk metrics are calculated using a government-wide FPDS-NG/USASpending.gov dataset.

If you’re not-so-data-savvy- don’t worry! There is still plenty that can be inferred by examining the information in front of you. For example, you can sense when a Government Contractor has bit off more than they could chew. Consider the story of an inexperienced small business (SB) Government Contractor who won a large construction award:

• The awarded prime contractor had difficulty getting bonded because the construction project was significantly larger than jobs handled in the past few years.
• The awarded prime contractor’s account indicated that the original design was behind schedule, $7M over budget, and had serious constructability issues.
• The federal government’s facility was eventually constructed with significant financial and litigation issues.

How did we know about this contractor’s troubles? For one thing, a FPDS-NG contract modification record revealed a surety takeover, and an employee of the prime contractor (joined by the Department of Justice) filed a qui tam lawsuit alleging violations of the False Claims Act…and this just garnered by a quick internet search!

Construction is known to be a high-risk type of federal contracting, but similar principles apply with IT systems and other types of federal requirements. Publicly available data can reveal when federal agencies may have erred in judgment in selecting a Government Contractor, and these are the opportunities to target as the likelihood of unseating the incumbent increases the more frustrated a federal customer is with performance.

Scenario 1: Rapid Growth, Major Decline in Federal Business. A glance at a Government Contractor’s year-to-year changes in federal business can be found on USASpending.gov’s Recipient Profile. This Government Contractor’s federal business grew rapidly to $80M in a single year. However, this amount of business could not be sustained long-term. At the end of the day, rapid growth can increase the risk that a Government Contractor will be unable to effectively manage projects and satisfy federal agency customers.

Scenario 2: Unable to Handle Large Award, All Funds De-Obligated. Drilling down to a monthly view of a Government Contractor’s obligations can be revealing. With only an annual summary receiving a large award then abruptly having all funds de-obligated would be a net-zero and wouldn’t be noticeable on a bar chart. In contrast, a monthly detail view on USASpending.gov shows a $33M award that was canceled and de-obligated in full. This Government Contractor did not receive any further contract awards.

Scenario 3: Hasty Fiscal Year-End Awards.  As we mentioned previously, due to the federal fiscal year-end deadline, federal agencies obligate more contract dollars in September than any other month. That being said, a Government Contractor with an unusually high percentage of obligations occurring in September has an increased risk of failure.

Using the risk scoring methodology outlined above, there is currently over ~$30B* worth of at-risk contract opportunities up for grabs across the federal government. This is an estimated population of poorly performing incumbents with federal agency customers eager for new contractors.

In the game of unseating an incumbent, data is crucial, but we know contextualizing it to be actionable is king. Government Contractors will need to connect the data-dots to translate them into strategic plans to effectively target opportunities and federal customers. By utilizing the open-source data differently you can get a leg up on analyzing your PWin for contracts with an established incumbent – all while increasing competition and driving out underperforming competitors.

A note on the data: This estimate was derived from risk factors found to be statistically significant by a random forest machine learning model. Each federal contractor was then assigned an overall risk score. $37,303,616,689.80 is the total obligations received by contractors with high-risk scores. The risk scoring methodology is admittedly imperfect, and the accuracy of risk assessments improves when complemented by the expert judgment of contracting professionals. The quantity of dollars up for grabs is difficult to reliably quantify in precise terms, but there is high confidence that billions of dollars are ripe for competing contractors to win.

About the co-author: David Gill is a government contracting official and data scientist. He has signed Treasury/IRS contracts for information technology, tax administration, and special initiatives such as the Troubled Asset Relief Program. His data science skills were learned on the job while serving as a fraud data analytics program manager. David has led data science teams that analyzed tens of millions in transactions for fraudulent activity. He supervised technical data science work, briefed agency executives, and responded to oversight/audit inquires. David has written R programming code to analyze the trillions of dollars in government contract data available on USASpending.gov. For example, mining USASpending.gov contract data to produce one of the government’s most comprehensive lists of machine learning and artificial intelligence projects. Views expressed are not necessarily the official position of the U.S. Government.