We've been looking at our development cycles recently because the cadence that we used to expect has started to change. It’s not that the process is broken exactly, but we’re seeing developers run out of tickets to work on, while at the same time things are getting backed up during testing.
Here is a breakdown of what we found across the stages
Product planning: faster to produce, no faster to agree on. Research, synthesis, user stories, and identifying edge cases come together much faster now. However, getting everyone to a shared understanding before locking things down hasn't changed.
Generating more material often creates confusion rather than clarity; you need enough text to reach consensus, not just look smart.
UX and design: faster to get started, no faster to be right. It's easy now for a designer (or even a non-designer) to put several plausible-looking design directions together in a matter of hours, but identifying what users actually need takes just as long as before.
Designing a cohesive experience means balancing user needs, business goals, and technical constraints. Crucially, testing solutions in context with real users to identify root problems remains a step that cannot be shortcut.
Development: the stage that changed the most. Tasks that once took two weeks now land in a few days. Five developers can move like fifteen. While exponential efficiency claims are often exaggerated - and overlook code quality, token costs, and long-term maintenance - producing code is unquestionably the primary area where the product development process is genuinely faster.
QA and testing: the new bottleneck. Testing hasn't gotten harder, but with development accelerating, QA is struggling to keep up. While generating test cases is faster, actually verifying complex changes still requires substantial human judgment. Because development output surged, releases are backing up at verification.
Worth a shout-out: our long-time partners at QA Camp bring a depth of testing experience that's hard to match, and we've relied on them for years.
What We're Doing About It
Upstream (Planning & Alignment):
Consumable artifacts: Favor short briefs, clickable prototypes, or quick walkthroughs over lengthy docs nobody reads. Create something memorable that will outlast the meeting.
Human-first story drafting: Write the core user value manually before asking AI to flesh out details. Also have AI work inside the codebase, so it has full context for completeness.
Verbal alignment: Confirm agreement out loud by asking team members to state the plan back before locking scope.
Clear ownership: Assign individuals clear end-to-end ownership of specific product outcomes rather than split responsibilities.
Downstream (Testing & QA)
Limit the blast radius during dev: Instruct AI agents to stay narrowly focused to avoid touching unrelated code.
Diff tickets against code: Ask AI to verify what changed beyond ticket scope before sending work to testing.
Cut smaller releases: Ship smaller updates more frequently so the test surface remains manageable.
Involve QA early: Ensure acceptance criteria are testable at the ticket stage rather than discovering issues during builds.
No matter how fast development runs, user adoption operates at human speed. Shipping in two days instead of two weeks doesn't change how quickly users form habits or decide your product adds value. If you want to read more about that, check out you can check out this post: User Adoption Is the Real Speed Limit.
If one stage of your cycle got fast and something else is now suffering under the weight of it, we'd love to compare notes.
