Most creative timelines slip on days nobody scheduled: the four days waiting for consolidated feedback, the week where the decider was travelling, the round that arrived as six contradictory emails. Suppliers plan their own work carefully and leave the client side as a gap called "review." Put client dates in the timeline with a stated consequence and most slippage disappears, because the cause was never the making.
Build the timeline backwards with both sides in it
Start at the live date and work back, and give every client step a date the same way you give your own steps dates.
| Step | Owner | Duration |
|---|---|---|
| Live | Client | fixed |
| Final delivery | You | 2 days before live |
| Round 2 feedback | Client | 2 days |
| Fine cut | You | 3 days |
| Round 1 feedback | Client | 2 days |
| Rough cut | You | 4 days |
| Direction approved | Client | 1 day |
| Concepts | You | 3 days |
Add it up before agreeing to a live date. That example is 17 working days, and seven of them belong to the client. Timelines that count only your seven working days are the ones that fail.
The consequence line
One sentence in the scope: late feedback moves delivery by the same number of days. Not as a threat, as arithmetic, stated at the start when nobody is defensive.
Without it, the supplier absorbs client delay silently until the last step, then compresses their own work and ships something worse. With it, everyone can see the trade in advance.
Where to put the buffer
Not at the end. A buffer at the end gets consumed by earlier slippage and disappears before it protects anything.
Put buffer before the first client review, where discoveries happen and where your own estimate is least reliable. Put a smaller one before final delivery for the fixes that always appear. Two buffers of two days each survive contact with reality better than one four-day buffer at the end.
What changed, and what did not
Production estimates got more reliable. Generation is fast and predictable in a way shoots are not: no weather, no talent scheduling, no location availability. A day that used to be "three to five days depending" is now a day.
Which means the variable moved to review. When production takes an afternoon, the timeline is almost entirely made of decisions. That should make timelines shorter and mostly does not, because review capacity did not change. The bottleneck is now the number of people who can judge work, and no amount of generation speed helps it.
Do not shrink the review windows to look faster. Two days for consolidated feedback is not padding, it is how long it takes to get six people to agree on one list. Promising a one-day turnaround creates the delay it was meant to prevent.
The three habits that hold a date
Name the decider before the timeline. An unnamed approver adds an unscheduled round.
Book the review meetings when the timeline is agreed, not when the work is ready. Calendars fill, and a review that waits four days for a slot is four days of slippage nobody planned.
Send work at the start of the day, not the end. Friday-at-six delivery buys you nothing and costs the client a weekend of not looking at it.
FAQ
How long should a video project take? Count both sides. A short-form campaign with two review rounds is realistically three to four working weeks, of which roughly half is client time.
What do I do when the client is late but the live date is fixed? Say what has to give: a round, a deliverable, or the date. Choosing silently, which usually means compressing your own work, produces worse work and the same date.
Should I pad estimates? Buffer explicitly rather than padding invisibly. Padded estimates get negotiated down because everyone knows they contain slack; named buffers survive because they have a stated purpose.
Half the timeline belongs to the client. Put their dates in writing and the schedule stops being fiction. The 8frame canvas is free and unlimited, and generation is paid from $19/month.