How Work Actually Moves Across Departments

Sep 10, 2026 | Blog, Deep Essays

The weekly “alignment” call has twelve people, a shared deck, and a polite vibe. Sales says the deal is closed. Customer Success says they never got a real handoff. Delivery says the kickoff is next Tuesday if someone sends scope. Finance asks when billing starts. Everyone nods. The meeting ends on time. Three days later, nothing has moved except the calendar invites for next week’s alignment call.

If that scene feels familiar, you do not have a collaboration problem in the soft sense. You have an interface design problem. The work is stalling at the seams between departments, and meetings are being used as a substitute for the thing that actually moves work: a named owner carrying a named artifact against a named clock.

This essay is for managers through VPs who live in matrix or multi-team work: sales to delivery, product to ops, HR to finance partners, regional to HQ. It is not another plea to “communicate more.” Communication without an interface is theater. The goal here is cross functional collaboration best practices that treat collaboration as flow design.

Reframe: collaboration is work-flow design across interfaces

Most advice on how to collaborate across departments starts with personality. Build trust. Have coffee. Assume good intent. Those things help after the interface is clear. They rarely fix an undefined handoff by themselves.

Cross-functional teamwork at work fails in a predictable pattern. Two teams leave a meeting believing they agreed. Each team returns to its own system of record. Each updates its own stakeholders. The baton never becomes an object anyone can point to. When someone asks “where is this?”, the answer is a story about who said what, not a link to an artifact with an owner and a due state.

Here is the reframe that changes how you lead:

Cross-functional collaboration is work-flow design across interfaces, not chemistry between personalities.

An interface is the place where work leaves one team’s control and enters another’s. Interfaces need four design choices: how work enters, who owns the baton now, what artifact proves progress, and what clock forces movement. If you cannot answer those four, you do not have a collaboration model. You have hope plus meetings.

This sits next to the literacy on the How Departments Work Together hub. That hub starts with a one-page interface card. This essay expands the operating logic behind that card into a full framework you can run across a company journey, not just one handoff.

Named framework: INTAKE → OWNER → ARTIFACT → CLOCK

Call the framework INTAKE → OWNER → ARTIFACT → CLOCK. Use it every time work crosses a department boundary.

INTAKE

Intake answers: how does work legitimately enter this team?

Without intake rules, work arrives as Slack pings, hallway asks, forwarded emails, and “quick favors.” Teams then look busy while their real queue is chaos. Good intake is not bureaucracy for its own sake. It is a definition of ready that both sides can see.

Intake design questions:

What triggers acceptance of work (form, ticket state, signed order, approved brief)?
What information must be present before the receiving team starts?
What gets rejected or returned, and who decides that?
Where does rejected work go (back to sender with a reason, not into a void)?

A definition of ready is kinder than a surprise. It prevents the receiving team from starting half-informed work that will be reworked later.

OWNER

Owner answers: who holds the baton now?

Not “who cares.” Not “the working group.” One human is Accountable for the current seam. Others may be Responsible for tasks. Many can be Informed. Few should be Consulted on every micro-move.

This is where Who Owns This: RACI Lite matters. RACI fails when twelve people are marked Consulted and two leaders claim Accountable. Dual-A is shared inbox cosplay. Force one A per outcome, or split the outcome into two named results.

Owner design questions:

Who will leadership call if this handoff is late?
Who can unblock without a committee?
Who updates the artifact when state changes?
Who escalates when the clock expires?

ARTIFACT

Artifact answers: what object proves the work moved?

Conversation is not an artifact. A meeting recording is weak evidence. Strong artifacts are durable, findable, and specific enough that a stranger on the next team can act.

Examples of real artifacts:

Closed-won handoff packet with scope, success criteria, contacts, and known risks
Kickoff brief with out-of-scope list and decision log
Billing readiness checklist tied to contract milestones
Async status update with decisions needed, risks, and next owner

If you cannot point to the artifact that moved and the owner of the next baton, you collaborated in conversation only. That sentence is the whole essay in one line.

The Async Update Instead of a Meeting play exists because status theater loves live meetings. Async works when the artifact contains decisions needed, blockers, and the next owner. Async fails when it is a novel nobody reads.

CLOCK

Clock answers: what timebox or service expectation forces movement?

Without a clock, stuck work becomes ambient. People feel busy. The queue ages in silence. A clock is not always a legal SLA. Treat workplace clocks as operating norms: agreed expectations teams hold each other to. Sample clocks in this essay are practice examples, not industry standards.

Clock design questions:

How long can a handoff sit before it is “stuck”?
Who gets pinged when the clock expires?
What is the escalation path after one nudge fails?
Does the clock reset when work is returned for incomplete intake?

Intake without a clock creates polite queues. Clocks without owners create blame. Owners without artifacts create heroics. Artifacts without intake create junk in the system of record. You need all four.

Designing an interface between two teams

Before the worked example, make the interface design concrete. For any pair of teams, write a short interface card.

Field	Design choice
Handoff	What state or packet moves from A to B?
Definition of ready	Minimum contents before B starts
Definition of done	Acceptance checks both sides use
Owner	One Accountable for the seam
Artifact home	System of record and link pattern
Clock	Stuck threshold and escalation
Exceptions	Who can waive ready/done, and how

Use a Stakeholder Map (30-Minute Version) when the room keeps growing. Map interest and needed involvement. Do not confuse “wants to be copied” with “must decide.” CC culture is how interfaces die in email.

Escalation should be boring and early. Escalate on a named path after N days or X failed attempts, not after people are already frustrated. Escalation is not failure. Silent aging is failure.

Worked example: sales closed-won → CS → delivery → finance (sample)

Labeled sample. Fictional company: Midscale B2B services firm. Roles are composites. The point is the seams, not a universal org chart.

Seam 1: Sales closed-won → Customer Success handoff

INTAKE: CS will not schedule kickoff until a handoff packet exists in the CRM: primary contacts, purchased scope in plain language, success criteria the customer said aloud, known risks, discount/exception notes, and contract link. “Closed-won email forward” is not intake.

OWNER: The AE owns completing the packet. The CS lead owns accepting or returning it within two business days. One Accountable for packet completeness: AE. One Accountable for acceptance decision: CS lead. Not “sales and CS together.”

ARTIFACT: Handoff checklist packet in CRM, plus a short customer-facing kickoff agenda draft. Use the Sales → CS Handoff Checklist as the operational template. Pair with Help This Deal by Role earlier in the cycle so non-sales partners are not surprised at signature.

CLOCK: Packet due within 24 hours of closed-won stage change. CS accept/return within 2 business days. If returned, AE has 2 business days to remediate. Stuck after that goes to sales manager + CS manager on a shared escalation channel.

Without this seam, CS inherits rumor. Delivery inherits rumor later. Finance inherits a billing surprise after that.

Seam 2: CS handoff accepted → Delivery kickoff

INTAKE: Delivery needs a kickoff brief: scope in/out, timeline hypothesis, customer constraints, dependencies on other teams, and “definition of done” for phase one. CS owns assembling it from the accepted handoff plus discovery notes.

OWNER: Delivery PM owns scheduling kickoff once the brief meets ready. CS lead owns briefing quality. Delivery lead owns whether capacity exists this cycle. Capacity is a decision, not a vibe. If capacity is no, the owner must say no to the date, not quietly hope.

ARTIFACT: Kickoff brief in the project system, linked from CRM. Decision log starts at kickoff: what was promised, what was deferred, who decided.

CLOCK: Kickoff scheduled within 5 business days of CS acceptance, or a written date exception with sponsor name. If delivery cannot staff, the clock forces an explicit freeze or re-sequence, not a soft maybe.

Seam 3: Delivery milestone → Finance billing readiness

INTAKE: Finance bills from contract milestones and delivery confirmation, not from optimism. Billing readiness requires: milestone definition from contract, evidence of completion, customer contact for invoice questions, and any credit/exception flags logged before invoice.

OWNER: Delivery PM owns milestone evidence. Finance ops owns invoice creation once evidence is complete. Sales owns explaining commercial exceptions before billing, not after the customer disputes the invoice.

ARTIFACT: Billing readiness checklist attached to the opportunity or order record. No checklist, no invoice. This is how you stop “we’ll figure billing out later” from becoming a collections incident.

CLOCK: Evidence due within 3 business days of milestone claim. Finance queues invoice within the company’s normal billing cycle after ready. Disputes escalate to a named commercial owner within 2 business days of customer pushback.

What the full path teaches

Every seam uses the same four letters. The content changes. The discipline does not. Directors and VPs who want systemic improvement should inventory recurring seams, not only firefight one project. Interface debt compounds the same way technical debt does: each undefined handoff taxes every future deal.

If you lead a portfolio, map the five most expensive seams in your world. Run INTAKE → OWNER → ARTIFACT → CLOCK on each. That is how you fix systemic interface debt instead of coaching one project at a time. The Director / VP hub’s portfolio map (own / influence / ignore) pairs well here: you own some interfaces, influence others, and should stop attending meetings on interfaces you neither own nor can unblock.

Managers running a single team still benefit. Your job is to refuse work that arrives without intake, name the owner on your side, keep the artifact honest, and honor the clock. The New Manager hub is useful when you are still learning to stop being the human router for every cross-team ping.

Sales literacy helps partners understand what “closed-won” actually means in How Sales Works. Money seams get clearer with Finance for Non-Finance Managers. Cross-functional collaboration is not one hub. It is the interfaces between hubs.

Meetings, async, and operating rhythm without sprawl

Cross functional project success does not require more standups. It requires fewer meetings with better artifacts, plus a small set of rituals that inspect stuck handoffs.

When async beats sync

Prefer async when:

The goal is status, not a decision that needs live negotiation
Participants span time zones or calendar chaos
You need a durable record more than a vibe check

Sync when:

Decision rights are contested and need a live call
Trust is damaged and repair needs voice
Ambiguity is high and a 20-minute whiteboard will prevent a week of thrash

An async artifact should contain at least:

Current state in one paragraph
Decisions needed (with options if possible)
Blockers and who can clear them
Next owner and next date
Link to the system-of-record home

Without those five, async is just a longer email.

How managers run rhythm without meeting sprawl

Use three layers:

Daily/ongoing: artifact updates in the system of record. No meeting required.
Weekly: stuck-handoff review (see Monday cadence below). Short. Exception-based.
Monthly: interface retrospective on one painful seam. Improve intake or clocks. Do not relitigate personalities first.

Leadership meetings should not become dump grounds for every cross-team status. Leadership Team Meeting Design helps keep LT time on decisions and risks, not oral status novels.

RACI-lite and stakeholder maps against “everyone / no one”

Two failure patterns dominate ownership:

Everyone owns it: twelve Consulted, dual Accountable, endless loops.
No one owns it: “the process” owns it, which means the loudest person owns it under stress.

RACI-lite prevents both if you discipline the A. Stakeholder maps prevent meeting inflation if you separate Informed from Consulted. Working across teams best practices are mostly about shrinking the decision set and making the baton visible.

A matrix collaboration at work environment makes this harder because people report to one boss and deliver for another. That is exactly why interfaces need written design. Informal “we’ll sync” collapses under matrix load.

Failure modes: alignment theater and friends

Failure mode 1: Meetings as substitute for artifacts

Symptoms: recurring alignment meetings, no single link anyone trusts, status that resets every week.

Repair: ban status-only meetings unless the artifact is updated first. Make the meeting about decisions and stuck clocks.

Failure mode 2: Liaison heroes as single points of failure

Symptoms: one person who “knows how to get things through” every department. When they are out, work freezes. When they stay, they burn out and become a political node.

Repair: turn hero knowledge into intake checklists and owner maps. Celebrate the person who made themselves unnecessary at the seam.

Failure mode 3: RACI with twelve Cs

Symptoms: large matrices, slow decisions, meetings that exist so nobody feels left out.

Repair: trim Consulted ruthlessly. Move people to Informed. Use stakeholder maps to justify the trim.

Failure mode 4: Tool sprawl without ownership

Symptoms: tickets in one system, truth in Slack, decisions in email, “source of truth” debated weekly.

Repair: pick an artifact home per seam. The tool is not the strategy. Ownership of the artifact is the strategy. Migrating tools without owners just moves the mess.

Failure mode 5: Polite CC culture

Symptoms: everyone is copied; nobody is Accountable; risk is diffused until it explodes.

Repair: require an owner line in every cross-team ask: “Owner: Name. Ask: Decision or action. By: Date.”

Failure mode 6: Definition-of-done drift

Symptoms: Team A marks done; Team B starts rework; trust erodes.

Repair: write acceptance checks both sides sign at interface design time. Audit them monthly.

These are the dominant failure modes in hand offs between teams. Personality conflict is real, but it is often a costume worn by undefined interfaces.

Monday operating cadence: keep interfaces healthy

Here is a practical cadence you can install without adding a bureaucracy department.

Weekly: interface board for stuck handoffs (30–45 minutes)

Who: owners of the top seams (not spectators).

Agenda:

List handoffs past clock (only exceptions).
For each: missing intake, unclear owner, missing artifact, or clock ignore?
One repair action with owner and date.
Escalate what the room cannot unblock.

Board columns can be simple: Ready / In handoff / Stuck / Accepted. The point is visibility of batons, not pretty project theater.

Weekly: definition-of-done spot audit (10 minutes)

Pick one recent handoff. Did the receiving team actually get what “done” promised? If not, fix the checklist, not the people first.

Monthly: interface retrospective (45–60 minutes)

Pick one painful seam. Ask:

Where did intake fail?
Where did ownership blur?
Which artifact was fake?
Which clock was ignored or unrealistic?

Change one design element. Ship the change as a written interface update.

Quarterly (directors/VPs): interface debt review

Inventory recurring seams. Rank by customer pain and internal thrash. Fund redesign of the top two. Ignore the fantasy that culture talks alone will clear structural debt. Pair with portfolio clarity on the Director / VP hub.

Managers can run a lighter version inside Weekly 1:1 Operating Cadence: ask each report “what baton are you holding across teams, and is the artifact clear?”

What directors and VPs uniquely own

Managers can enforce local interfaces. Directors and VPs own the system of interfaces:

Which seams are strategic enough to standardize
Where dual-A conflicts need executive resolution
Whether incentives punish good handoffs (for example, celebrating signature with no attach-quality check)
Whether LT meetings inspect stuck handoffs or only revenue anecdotes

If you only coach project by project, you will relearn the same lesson forever. Treat interface design as operating infrastructure. Cross departmental collaboration framework work is dull in the best way: checklists, owners, clocks. Dull systems beat charismatic chaos.

Related deep practice on the sales-to-delivery pain specifically lives in the companion theme of closed-won quality (when a sales win becomes a company loss). The handoff checklist plays above are the tactical spine. This essay is the operating philosophy behind them.

A note on “aligned” meetings that still stall

People leave rooms saying they are aligned because disagreement felt socially expensive. Alignment without an artifact is agreement about tone, not about work. Before you end a cross-team meeting, force four sentences into the notes:

Intake status: ready / not ready (why).
Owner of the next baton: Name.
Artifact link: URL.
Clock: next visible state by Date.

If those four cannot be written, you are not done. You are postponing the stall. This is also how you teach teams that hand offs between teams are designed objects, not vibes.

For contested decisions inside the meeting, name the decision right live: who decides, by when, with what input. If nobody can name the decision right, schedule a short decision meeting with the actual decider. Do not schedule another “alignment” meeting that invites the same twelve people to reenact uncertainty.

Capacity is part of the interface

Interfaces fail when the receiving team has no right to say “not this week.” Fake yes dates destroy trust faster than honest capacity limits. Build capacity checks into OWNER and CLOCK:

Who can refuse a start date?
What alternate outcomes exist (queue, reduce scope, buy help, escalate tradeoff)?
How is the refusal documented so sales or sponsors see reality early?

A company that celebrates every yes and punishes timed nos will invent shadow work and quiet failure. Cross functional project success includes the courage of a dated no.

Putting it to work this week

Pick one recurring seam that already hurts.
Write INTAKE → OWNER → ARTIFACT → CLOCK in one page.
Socialize it with the two team leads in a working session, not a broadcast email.
Run the weekly stuck-handoff review for two weeks.
Adjust clocks that were fantasy. Keep owners singular.

You will feel resistance. Some people prefer meetings because meetings feel like progress. Artifacts feel like accountability. That discomfort is the point.

Closing

Cross functional operating rhythm is not a personality upgrade. It is the discipline of making work move as objects with owners under time. When leaders ask for better collaboration and then schedule another alignment meeting without an artifact owner, they recreate the stall.

If you remember one test, remember this: If you cannot point to the artifact that moved and the owner of the next baton, you collaborated in conversation only.

Design the intake. Name the owner. Keep the artifact honest. Honor the clock. That is how work actually moves across departments.