R&D Tax Credit Knowledge Base

TaxCredit.ai launches GAAP Capitalization Reporting ⚡️

Written by TaxCredit.ai | Apr 30, 2024 5:26:21 PM

You may be wondering... How is Software Capitalization related to R&D Tax Credits?

Both Software Capitalization (under Generally Accepted Accounting Principles) and the R&D Tax Credit (under IRS rules) require an analysis of engineering activities at various phases in the software development cycle.

In the context of R&D tax credits, TaxCredit.ai classifies data from GitHub, GitLab, and Azure DevOps as Qualified Research Activities or excluded activities according to IRS rules, then calculates each engineer's R&D time for the period (among other things).


For software capitalization, TaxCredit.ai classifies engineering data according to GAAP Capitalization rules instead. Our AI determines
the portion of each engineer's time spent in the Application Development phase (capitalized) vs. the Post-implementation phase (expensed as incurred). 

Many accountants may not realize they can track software engineers' time without hassling them. With TaxCredit.ai, engineers aren't required to do any additional work- accountants can leverage data from the platforms their engineers already use.

Why are accountants investing in Software Capitalization Reporting?

Primarily, it's to satisfy external financial stakeholders like auditors and investors. Often, growing tech companies do not have a formal process in place to record the amount of time their software engineers spend in different phases of the software development cycle, as is required under GAAP.

Getting engineers to record their time specifically for accounting (or tax credit) purposes is difficult and impractical, if not impossible. We know that from our own experience as a software company!

And that’s not a knock on engineers - it’s not in their job description. Nor is it a productive use of their time with the lightning-fast way agile teams move in today’s world.

As a result, the company may receive comments about methodology deficiencies and potential misstatements during their annual audit, like the one below:

TaxCredit.ai bridges the gap for accountants, transforming raw data from the platforms their engineers already use on a day-to-day basis into actionable insights and calculations for Software Capitalization reporting.

Investing in a data-driven Software Capitalization Reporting platform like TaxCredit.ai will strengthen substantiation of your company’s accounting methods, remediate deficiencies identified by auditors, and showcase your commitment to financial excellence for external stakeholders.

Keep reading to explore how TaxCredit.ai helps tech companies with software capitalization!

Which costs or activities are required to be capitalized under GAAP?

Under GAAP Capitalization rules, three phases of the software development cycle must be considered in order to determine the amount to be capitalized and amortized:

  1. Planning
  2. Application Development
  3. Post-Implementation

The costs associated with the Application Development phase are required to be capitalized, while costs associated with the Planning and Post-Implementation phases should be expensed as incurred.


The three phases of the software development cycle and how the expenses must be treated under GAAP. TaxCredit.ai calculates time spent in the two engineering phases.

TaxCredit.ai calculates Application Development and Post-Implementation time for tech accountants

TaxCredit.ai ingests your engineering team's Git data and uses an AI classifier to categorize each task among the two engineering phases that often comprise the overwhelming majority of the software development cycle for agile teams:

Application Development

  • Features
  • Functionality
  • Improvements
  • Pre-release bug fixes
  • Updates

Post-Implementation

  • Post-release bug fixes
  • Maintenance
  • Troubleshooting
  • Similar activities required to keep a software application working as intended, but not adding to the application’s functionality.

What about the Planning phase?

The Planning phase often represents a small(er) portion of the software development cycle compared to the engineering phases (Application Development and Post-Implementation).

While code versioning platforms like GitHub, GitLab, and Azure DevOps capture all of a software team’s actual coding activities, they don’t capture the time people spend in the Planning phase, doing things like creating tasks, assigning tasks to engineering staff, or spending time in product meetings.

We approach Planning phase exclusions one of two ways:

  1. Using data from a project management platform like Jira, ClickUp, Asana, or similar, we're able to estimate of the portion of time a person spent creating tasks, assigning tasks, or in product meetings.
  2. We'll ask your VP Engineering to make an estimate of Planning phase time for each person based on role/job title.

Do note, engineering supervisors often spend more time on Planning phase activities than engineering staff, so we'll recommend excluding a larger portion of Planning time for supervisors (engineering managers, VPs, and C-suite) than for engineering staff. It’s also common to see Planning phase time increase incrementally with the person’s position on the org chart- the further they are removed from the ground-level engineering, the higher the exclusion.

How long will TaxCredit.ai take to implement?

In short, it’s little to none. Your engineering team doesn't need to be onboarded to TaxCredit.ai.

Software engineering teams use platforms like GitHub, GitLab, and Azure DevOps every single day throughout the year, so the data required for a historical analysis like Q1-2024 or fiscal year 2023 already exist - it just needs to be imported into TaxCredit.ai by one of your engineering admins.

After import, TaxCredit.ai works its magic, converting raw data into actionable insights and calculations that accountants understand, like how much of each person’s time/compensation should be capitalized and your company's total capitalized cost for the period.

What’s the data import process like?

It's quick! One engineering admin can often import data for the entire engineering team by linking to GitHub or exporting data into human-readable text files. The manual export/import method allows engineering admins to see exactly which information they are providing, and we have instructions that make the process easy.

Importantly, TaxCredit.ai does not read your source code or intellectual property. It uses metadata about all of the changes made to the codebase during the period.

After import, TaxCredit.ai takes care of the heavy lifting, instantly classifying every task completed by each engineer as either Capitalized or expensed as incurred.

Let's connect soon!

Please get in touch with us if you'd like to see how it works over a quick Zoom demo. Additionally, you can download an anonymized TaxCredit.ai GAAP Capitalization Analysis below to evaluate how your existing methodology stacks up!