Why Campaign Launches Slip: The Three Dependencies That Cause Almost Every Delay
Campaign delays are rarely caused by the work taking longer than expected. They're caused by three specific dependencies that sit outside the campaign team's control and are almost never sequenced in the plan.
By Robin Deane — Founder & Marketing Strategist, RD
Campaigns rarely slip because the creative or the copy took longer than estimated — that work is usually estimated fairly well by the people who do it. They slip on three dependencies that sit outside the campaign team's control and are almost never given a place in the plan: approval from someone who has other priorities, a technical prerequisite owned by another team, and an input from a third party with no obligation to your deadline. All three share a property that makes them uniquely damaging: they are invisible until they are already late, because nothing signals a delay until the thing you were waiting for fails to arrive. The fix is not more buffer. It is sequencing these three earlier than feels necessary, giving each a named owner and a date, and treating a missed dependency date as an escalation the day it happens rather than a problem discovered at the end.
Ask why a campaign launched two weeks late and you will usually get an answer about the work: creative went through more rounds than expected, the copy needed a rewrite, the build was more complicated than it looked. These answers feel right and are almost always wrong — or at least, they describe the last visible symptom rather than the thing that caused it.
Track a slipped launch backwards through the actual record and the pattern is remarkably consistent. The creative rounds were not unusually numerous; they started nine days late because the brief was waiting on a positioning decision. The copy rewrite was not the problem; the copy was written twice because the offer changed after legal reviewed it in week five instead of week two. The build was not complicated; it sat finished for a week waiting on a tracking parameter from an agency that had never been told there was a deadline.
None of those are estimation failures. They are dependency failures, and they behave differently.
Why Don't Buffers Fix This?
Because a buffer protects against work taking longer than planned, and a dependency delay is not work taking longer. It is work not starting.
If a design task is estimated at five days and takes seven, a two-day buffer absorbs it. That is what buffers are for and they work. But if the design task cannot start until an approval arrives, and the approval arrives eleven days late, no realistic buffer covers it — and worse, the buffer was consumed silently while everything looked fine. The plan showed a task not yet started, which is exactly what a plan shows for a task that is not due to start yet. Nothing flagged.
This is the property that makes dependencies uniquely damaging: they generate no signal until they are already late. A slow task is visibly slow while it is happening. A waiting task looks identical to a task that is comfortably on schedule, right up until the moment it isn't.
What Are the Three Dependencies?
Definition: An external dependency is any input a campaign needs that is produced by someone outside the campaign team, on a timeline the campaign team does not control. The defining characteristic is not difficulty but ownership: the campaign is accountable for the launch date while somebody else is accountable for the input, and that somebody has their own priorities against which your deadline may rank low or not appear at all.
| Dependency | Typical form | Why it slips | When to sequence it |
|---|---|---|---|
| Decision | Positioning sign-off, offer approval, budget release, legal or brand review | The decision-maker has other priorities and no visibility of what is blocked behind them | Before any dependent work is briefed — this is the earliest item in the plan, not a gate near the end |
| Technical prerequisite | Landing page, tracking setup, CRM field, audience segment, integration, DNS or domain change | Owned by a different team with a separate backlog and sprint cadence that your campaign does not appear in | At brief stage, entered into that team's actual planning process rather than requested informally |
| Third-party input | Agency asset, partner co-marketing approval, publisher slot, translation, influencer or spokesperson availability | No contractual obligation to your internal deadline, and often no awareness one exists | First, with a stated date and a written confirmation that the date is accepted |
The right-hand column is the whole argument. Every one of these needs to be sequenced earlier than it feels necessary, because the cost of asking early is a few minutes and the cost of asking late is the launch date.
Why Does the Decision Dependency Cause the Most Damage?
Because everything else is downstream of it, and because it is the one most often left out of the plan entirely.
Technical prerequisites and third-party inputs at least tend to appear as tasks somewhere. Decisions frequently do not — they are assumed to happen naturally, in the course of things, because they take five minutes of someone's time. That five minutes is real. The problem is the queue in front of it: a decision that takes five minutes to make can take three weeks to obtain, and the gap between those two numbers is where campaigns die.
The failure mode is specific and worth recognising. Someone is asked for a positioning decision by email. They read it, judge it non-urgent because nothing in the message says otherwise, and file it behind things with visible deadlines. Meanwhile four downstream tasks are quietly blocked, none of which is visible to the person holding the decision. Two weeks later the campaign manager escalates and the decision is made in five minutes, exactly as predicted — the estimate was never wrong, the queue was.
The correction is cheap: when a decision is requested, state what is blocked behind it and the date it becomes a problem. "We need positioning confirmed by Tuesday the 9th, because creative is briefed on the 10th and a later decision moves launch" converts an invisible request into a visible one. Most decision-holders respond to that immediately, because most of them were never refusing — they simply had no way to know.
How Should a Plan Actually Handle These?
Walk the campaign forward and write down everything that has to come from outside the team. Most plans skip this and start with the work, which builds a schedule on top of assumptions nobody has checked.
"Legal" cannot be chased and does not miss a date; a named person can and does. A dependency assigned to a department is a dependency with no owner, and it will be the one that slips.
A date you assigned to someone else is a wish. A date they accepted is a commitment. The difference shows up entirely in week four, and the confirmation costs one message.
Decision-holders and other teams deprioritise what looks non-urgent. Naming the downstream tasks and the date the delay becomes a problem is the single most effective thing a campaign manager can do, and it costs a sentence.
Standups review what people are doing. Nobody is doing a dependency, so it never comes up until it fails. A short standing item — "what are we waiting on, and is it still on track" — surfaces the silence before it becomes a delay.
Not the week after, not once it is clearly a problem. A dependency one day late is a conversation; a dependency nine days late is a launch date. The instinct to wait a few days before chasing is what converts the first into the second.
Step six is the one teams find hardest, because chasing on day one feels impatient. It is worth being explicit with the team that it isn't: escalating a missed date immediately is a process behaviour, not a judgement about the person who missed it, and the alternative is a delay that costs far more to absorb later.
What About Genuinely Unforeseeable Delays?
They exist and they are rarer than they are claimed to be.
The honest test is the one from any decent post-mortem: could this have been known earlier by someone, and if so, why didn't it reach the campaign team? A key person's planned leave is not unforeseeable — it was in a calendar. An agency's capacity constraint in December is not unforeseeable. A legal review taking two weeks is not unforeseeable if it has taken two weeks every previous time.
What genuinely qualifies is a small set: illness, a platform outage, an unexpected regulatory intervention, a partner withdrawing. For those, the mitigation isn't better planning — it's knowing in advance which parts of the campaign can ship without them, so a delay in one input doesn't hold the entire launch. That question is worth asking at brief stage: if this one input is late, what still goes out? Most campaigns have more independence between their parts than the plan assumes, and finding it early converts a total delay into a partial one.
What Changes When a Team Gets This Right?
- Campaigns rarely slip because work took longer than estimated — they slip because work started late, waiting on something outside the team
- Buffers protect against slow work and do nothing for waiting work, which is why adding buffer never fixes a dependency problem
- Dependencies generate no signal until they are already late: a blocked task looks identical to a task not yet due to start
- The three are decisions, technical prerequisites, and third-party inputs — all need sequencing earlier than feels necessary
- Decisions cause the most damage because they're often left out of the plan entirely; a five-minute decision can take three weeks to obtain
- Stating what's blocked behind a request, and the date it becomes a problem, converts an invisible ask into a visible one — most delays are queue position, not refusal
- Assign dependencies to named people, not departments, and get the date confirmed by the person who owns it
- Escalate the day a dependency date is missed; waiting a few days out of politeness is what turns a conversation into a slipped launch
- Ask at brief stage what still ships if a given input is late — most campaigns have more independence between parts than the plan assumes
The visible change is not that launches stop slipping entirely. It is that slippage becomes early and small rather than late and large. A team managing dependencies well discovers a problem in week two, when there are still options — resequence, descope, move a date deliberately. A team managing them badly discovers the same problem in week six, when the only remaining option is to launch late and explain why.
That difference is almost entirely about when you find out, which is why the practices above are all forms of the same thing: making silence visible earlier. Nothing here requires more resource, a better tool, or a more disciplined team. It requires treating "we are waiting on something" as a status worth tracking, rather than the absence of one.
Our campaign management service runs this sequencing as standard, and our use case on launching against a fixed deadline covers what it looks like when the date genuinely cannot move. For the review that catches recurring dependency failures across campaigns, see the campaign post-mortem nobody runs. Where the delay is between marketing and sales rather than inside the campaign, sales and marketing SLAs covers the same problem across a team boundary. And for the checks that stop a rushed launch shipping broken, see campaign pre-flight.
Frequently Asked Questions
Why do marketing campaigns usually launch late?
Not because the creative or build work took longer than estimated, but because that work started late while waiting on something outside the campaign team's control — a decision, a technical prerequisite owned by another team, or an input from a third party. Tracked backwards, most delays that look like execution overruns began as a wait several weeks earlier.
Why doesn't adding buffer time prevent campaign delays?
Because a buffer absorbs work that takes longer than planned, and dependency delays are not slow work — they are work that has not started. A task waiting on an approval looks identical in a plan to a task that is simply not due yet, so the buffer is consumed silently with nothing flagging until the input fails to arrive.
What are the three dependencies that cause campaign delays?
Decisions (positioning sign-off, offer approval, budget release, legal or brand review), technical prerequisites (landing pages, tracking, CRM fields, audience segments, integrations), and third-party inputs (agency assets, partner approvals, publisher slots, translation). All three sit outside the campaign team's control and all three need sequencing earlier than the plan usually allows.
Why do approval delays cause the most damage?
Because everything else is downstream of them and they are most often left out of the plan entirely, being assumed to happen naturally since they take only minutes of someone's time. The estimate is accurate but the queue is not: a five-minute decision can take three weeks to obtain when the decision-holder has no visibility of what is blocked behind it.
How do you get faster decisions from busy stakeholders?
State what is blocked behind the request and the date the delay becomes a problem — for example, that creative is briefed on the 10th and a later decision moves launch. Most decision-holders are not refusing; they judged the request non-urgent because nothing indicated otherwise. Making the downstream cost visible usually produces an immediate response.
When should you chase a missed dependency?
The day it is missed. A dependency one day late is a conversation; the same dependency nine days late is a slipped launch date. The common instinct to wait a few days before chasing, out of politeness, is what converts the first situation into the second — and immediate escalation is a process behaviour rather than a judgement about the person.
How do you plan for genuinely unforeseeable delays?
By identifying at brief stage which parts of the campaign can ship without each input, so a single late dependency does not hold the whole launch. Most campaigns have more independence between their components than the plan assumes. Genuinely unforeseeable events are a short list — illness, outages, a partner withdrawing — and most delays described that way were knowable by someone earlier.
Keep Reading
Want this handled properly?
If this is the kind of problem you're wrestling with, a short conversation is usually enough to tell whether there's a real opportunity here.