A lot of different numbers enter into discussion when sizing federal technology markets spanning budget, contract awards, and reported spending. Of these categories, reported spending provides a view on how agencies are actually using funds. The publicly available portion of federal spending can be accessed through repositories like the Federal Procurement Data System (FPDS). As implementation of the Digital Accountability and Transparency Act (DATA Act) gets underway, the quantity and quality of the publicly released data is expected to improve as practices are standardized. Nonetheless, determining the size of current and future federal technology markets continues to challenge business development organizations.
Market Sizing Complications
Hurdles to market sizing can stem from market definitions, customer requirements, as well as nuances within available data. A couple of elements that are often referenced in segmenting federal budget and spending include Object Classification Codes, Product or Service Codes (PSCs), and the North American Industry Classification System (NAICS). While each offers a different sort of utility, they are also subject to limitations that impact sizing a market and assessing the competitive landscape. The simplest way to slice and dice spending information would be a breakdown along a single element of the spending data. However, analyzing data along a single aspect would be detrimental to market characterization due to the lack of flexibility and granularity.
Other restrictions revolve around the availability and contents of reported information. How agencies report spending data represents an inherent limitation in spending data. Reporting lags, most notably the Defense Department’s three month delay, can lead to significant fluctuations over time. Variation and errors within included details and terminology present an issue as complex as addressing partial information in contract descriptions. Challenges also arise pertaining to multiple products and services being acquired through a single contract action. Depending on alignment of elements with a market taxonomy, there may also be a risk of duplicative counting. The role of subcontracting can be difficult to address, as prime contracts dollars are more accessible. These drawbacks needn’t undermine market sizing efforts, but it is important to acknowledge them and bear them in mind when making decisions based on any findings.
Federal Industry Analysis Approach
GovWin’s Federal Industry Analysis (FIA) team takes all of these considerations and complexities in hand when we forecast the federal IT market (leveraging over a dozen different data sources) or present analysis of reported IT spending. Of particular use for making sense of agency procurement reports, the team’s approach has identified areas of reported spending that target technology products and services, as well as those that include a high degree of IT-related spending. For recent analysis of reported IT spending, the team completed a search of over 2.2 million FPDS transactions for the period from FY 2012 to FY 2014 across eight product and services categories. The scope was narrowed down to relevant transactions using an evolving selection of keywords. A number of additional steps are necessary to determine what this IT spending data reveals about vertical markets like information security, big data, or mobility. Our assessments further refine results through a review of NAICS, PSCs, and contract descriptions. Individually, these metrics may not yield an accurate view of spending. More often than not, there will be data – representing millions of dollars in some cases - for which the PSC code does not relate at all with actual contract description. It’s important to review results with a critical eye focused on data integrity to ensure that bad reporting does not skew the results. Additional measures must be taken to scrub the data to further improve the quality of the dataset for each market, up to and including manual review of questionable transactions.
Limitations and assumptions associated with the process are also carefully reviewed and documented. For example, manual entry of contract requirement information by government agencies contributes to the challenge of identifying and isolating relevant spending. Adjustments to collection criteria to address misspellings, typographical errors, and abbreviations may help to capture additional transactions relevant to your market that would otherwise have been omitted. Further review of results helps to reduce noise related to false positives – where context reveals a different meaning for keywords and terms. Even within PSCs designated for information system buying like software, terms that seem exclusive to technology-related purchases are applied to other mission-oriented and business needs. Once the dataset has been thoroughly scrubbed, analysis is completed to highlight key information and market characteristics. Even a “clean” dataset is likely to contain some adjacent spending or a small degree of transactions not wholly and purely related to a given market. This is partly a result of the data limitations, but it also reflects buying of IT solutions that span market segments, like bundling of products and services.
Getting a sense that this process is not for the faint of heart? While it can be a cumbersome and time consuming project, technology helps to streamline the approach. To be sure, employing analytical tools and techniques delivers significant efficiency improvements. Yet, some tools make it too easy to cut too many corners. An extensive market taxonomy does not guarantee an honest view of the market. In short, those tools are not a substitute for sound methodology and due diligence.
Curious about historic spending for IT markets like cybersecurity, big data, mobility, and cloud? Stay tuned! The FIA team will be publishing new reports that include this analysis as well as forecasts of contractor addressable spending.