FUTURICE / GITHUB EXECUTIVE BRIEF

Futurice Copilot
impact summary

A compact view of platform usage, Copilot maturity, engineering impact, and the opportunities most likely to convert adoption into measurable value.

High-leverage adoption opportunity: 82 users remain in Cohorts 0-1. Engaged users average 10.1 merged PRs per month versus 0.7 at baseline, while aggregate merge time for Copilot-active users was slower.

CURRENT SNAPSHOT

Executive overview

Licensing and impact data as of 13 Sep 2026

Copilot Business

90

licensed seats

Copilot Enterprise 0
Copilot AI units

527,554

current consumption snapshot

750 Coding Agent AI units
Conversion opportunity

82

Cohort 0-1 users

80.4% of measured cohort
GHE metered consumed ACTIVE

270

GHEC Classic / Metered

GitHub environment Futurice Enterprise slug / GHEC classic
Authoritative account Futurice - Enterprise Cloud Salesforce 0010V00002NEuH3QAL

OBSERVED COHORT ASSOCIATIONS

Copilot impact

Latest impact service window: 17 Aug-13 Sep / 28 days

POSITIVE SIGNAL

Code shipped

14.1x
Engaged
10.1
Baseline
0.7

The impact service observed 703 merged pull requests and estimates substantially more code shipped for engaged users than at baseline. This is an association, not a causal productivity claim.

NEGATIVE SIGNAL

Overall time to merge

1.5x
Copilot-active
11.6h
Baseline
7.5h

Users with observed Copilot activity were slower to merge in aggregate. Investigate workflow composition before presenting a speed benefit.

Read these signals together. Higher output did not translate into faster aggregate delivery. These are observed associations, not proof that Copilot caused either result.

Engaged / Copilot-active

Users with observed Copilot activity during this 28-day reporting window; this label applies to the merge-time comparison.

Baseline

The comparison group used by the impact service, representing users without observed Copilot engagement.

PRs merged per month

The impact service estimate used to compare Cohorts 2-3 with Cohorts 0-1.

0

No AI

No observed Copilot engagement; development follows existing workflows.

1

IDE only

Copilot is used inside the IDE for suggestions, chat, and individual assistance.

2

Agentic

Agents complete bounded development tasks while a developer directs and reviews.

3

Multi-agent

Multiple specialized agents coordinate across a broader engineering workflow.

COHORT PERFORMANCE SIGNALS

How usage maturity relates to delivery activity

102 users represented / latest 28-day impact window

Delivery activity by Copilot adoption cohort
Cohort Users Avg lines added / day Avg PRs merged Median PR merge time
Cohort 0No AI 28 2.7 0.6 7.5h
Cohort 1IDE only 54 89.6 1.1 16.9h
Cohort 2Agentic 5 172.9 0 Not availableNo merged PR sample
Cohort 3Multi-agent 15 46.0 24.3 15.4h

Lines added is average daily code volume for users in the cohort. It does not measure quality or mean Copilot wrote the lines.

PRs merged is the average merged pull-request activity observed for users in each cohort during the window.

Median merge time is the midpoint elapsed time from opening a PR to merging it. Small cohorts can move sharply.

SIX-MONTH TREND

Monthly pull requests and cohort mix

August recovered to 673 PRs after the July low of 513; this is a delivery trend, not proof of Copilot causality.

CURRENT CONSUMPTION

Copilot usage snapshot

AI units / Futurice / as of 13 Sep 2026

Explore the current enterprise-level Copilot consumption snapshot.

Current snapshot 527,554 AI units Consumption units, not invoice dollars
METRIC SHARE OF TOTAL
The current licensing source does not provide a dated 30-day range or channel-level allocation. AI units are shown as a current snapshot, not invoice dollars.

PRODUCT FOOTPRINT

Platform scale and utilization

Licensing snapshot as of 13 Sep 2026

Current AI units snapshot527,554
Coding Agent AI units750
AI budget$200 alert-only
Budget view used$0 reported

The budget view reports $0 used while AI-unit consumption is present. These are separate measures and do not imply zero AI usage.