The "Later" Column Is Deferred Accountability

Open your team's roadmap. Look at the "Later" column. Count the items. Now check when the oldest one was added. That number — in days — is your Context Collapse score.

Nobody moves a ticket to "Later" as an act of sabotage. They move it because there's genuine uncertainty, a competing priority, or a dependency that isn't resolved. The intention is: we'll get to this when the time is right.

The problem is that the time never arrives. The ticket ages. The person who wrote it changes teams. The business context shifts. The technical constraints that made it hard evolve — but nobody updates the ticket. Eventually, "Later" isn't a column. It's a graveyard where context goes to decompose.

The Memento Problem

In Memento, the protagonist makes decisions based on polaroids and tattoos — fragments of context captured at a moment in time, increasingly disconnected from reality. He trusts them because he has to. He has nothing else.

Your "Later" column works the same way. The ticket is a polaroid. It captured someone's understanding at the moment it was written. By the time it resurfaces, the person reading it is reconstructing intent from fragmentary evidence. They'll either make confident decisions on stale context, or they'll spend engineering time reconstructing what was already known.

Both outcomes are waste. The first is invisible — you don't know what you don't know about why that ticket was deferred. The second is visible but feels like due diligence.

Context Has a Half-Life

Fossilized Backlog — Q4 Sample Age in Days

Auth v2 Migration

Deferred — original engineer no longer on team

640d

Observability Revamp

Deferred — "after the platform migration"

412d

Technical Debt Sprint

Deferred 3rd time — no context in ticket

—

Mobile Parity

Deferred — blocked on design system decision

289d

We've analyzed ticket repositories across dozens of engineering organizations. The pattern is consistent: context degrades in roughly two phases. In the first 60–90 days, a deferred ticket can be re-evaluated by almost anyone on the team. After 180 days, only the original author can reconstruct the full reasoning — and they're often gone. After 365 days, the ticket is effectively fiction. It looks like a plan. It isn't.

Where Accountability Goes to Die

The "Later" column creates a specific organizational dysfunction: accountability without ownership. The ticket exists. It's on the roadmap. Someone, notionally, is responsible for it. But nobody is actively thinking about it, updating it, or making a decision about whether it should ever be done.

This is the Looper problem. In the film, characters make a deal to defer a problem — they'll handle it when it arrives. In the meantime, the deferred problem compounds, its scale growing in ways nobody anticipated, until the reckoning is far more costly than the original intervention would have been.

Engineering organizations do this constantly with deferred work. The Auth v2 Migration that would have taken two weeks in Q2 now requires three months because the codebase evolved around the old architecture. The Observability Revamp that was blocked on the platform migration is now blocked on a completely different set of constraints that the ticket doesn't mention.

The Roadmap Tells a Story. Check If It's True.

Now

Context is fresh

Specific and actionable. The author can be found. The constraints are understood. Estimates are grounded.

Context is plausible

Rough shape, some gaps. Probably still accurate. Worth a brief re-evaluation before picking up.

Later

Context has collapsed

Fiction dressed as a plan. The intent survives. The reasoning doesn't. Act on it and you're guessing.

The column doesn't describe priority. It describes how much of the original decision-making context has survived.

Most teams treat the roadmap as a snapshot of priority. It's actually a snapshot of knowledge at the moment each ticket was written. The NOW column reflects current understanding. The LATER column reflects someone's understanding from months or years ago — under different constraints, with a different team, for a different business context.

The Diagnostic Question

"If we picked up the oldest item in our Later column tomorrow, what would we need to re-learn before we could start?"

If the answer is "we'd need to talk to four people, read two Confluence pages, and check what changed in the platform since this was written" — that's your Context Collapse cost. That's the invisible tax sitting in your backlog, accruing daily.

The BTA diagnostic quantifies this cost directly. We analyze your ticket repository for age distribution, context density, and deferral patterns. We produce a Context Collapse score and a prioritized list of backlog items — ranked not by age or priority, but by reconstruction cost. The items that will cost the most to revive are surfaced first.

Some tickets should be closed, not deferred. Some should be escalated. Some have been superseded by changes that nobody noticed. The diagnostic tells you which is which — so you can stop guessing and start deciding.