GitHub Copilot’s revenue or paid users will show no annual decline by July 2027 — resolves 2027-07-14

Coding assistants are moving from suggesting code in an editor toward agents that can carry out whole tasks. The question is whether those agents will shrink the market for suggestion-based tools.

SecondSource expects both approaches to coexist. This card tests that view using official revenue or paid-user disclosures.

Through 14 July 2027, SecondSource expects GitHub Copilot, or a copilot-style product among the top two by market share, to show no year-over-year decline in officially disclosed annualized revenue or paid users. The estimated probability is 60%.

The test uses whether the reported annual change is negative, rather than a present-day baseline number that SecondSource has not verified. It does not require both revenue and users to be separately disclosed.

ARR means annual recurring revenue: recurring income expressed on a yearly basis. “Copilot-style” here means a coding assistant that suggests code. If revenue and paid users are both disclosed but one is flat or up and the other declines, the original rule does not say which to use; that needs clarification at assessment, and this card is not settled. The ranking measure for a top-two substitute is also unspecified. If Copilot figures stop being reported separately and are folded into an inseparable category, the original rule voids the forecast.

The original rule also does not specify how the option to use a substitute product interacts with voiding the forecast when Copilot figures stop being disclosed separately.

The original material behind this call has not yet been linked here in a form readers can open and check.

Read the most recent official disclosure of ARR or paid users, from Microsoft earnings or a GitHub announcement, and compare it year-over-year against the prior-year figure.

What would overturn it

A negative year-over-year disclosure flips the growth call for copilot-style products.