What "Blocked By" Actually Buys You: Task Dependencies as a Small Agency''s Cheapest Risk Management
Picture a small dev agency mid-sprint on a client site rebuild. The frontend dev is ready to wire up the new component library — except the design system components aren't finished yet. Everybody on the team sort of knows this. It came up in standup. But "sort of knows" is exactly the kind of fact that survives fine in a four-person team's heads for about a week and then quietly stops being true the moment someone's on vacation or a new task gets added to the pile.
That's the actual case for task dependencies, and it's a much smaller, cheaper thing than "risk management" usually sounds like. It isn't a certification or a discipline. It's writing down "this can't start until that's done" in a place the whole team can see, instead of relying on memory.
What it looks like in practice
If your plan includes Task Dependencies, every task's detail page has a "Blocked by" field. Mark a task as depending on another task in the same project, and that relationship is now visible to anyone who opens either task — not just the two people who discussed it in a hallway conversation. Remove the dependency the same way, once it's no longer true.
That's genuinely the whole feature. It's not a critical-path calculator, it's not automated scheduling, it's not a Gantt chart that reshuffles your whole project when one task slips. It's a single fact, attached to the task it's actually about, that stays there whether or not the person who set it up remembers to mention it again.
Why that's enough to count as risk management
The expensive version of risk management is a formal process — regular risk reviews, a register, someone whose job includes tracking what could go wrong. A four-person agency doesn't have the headcount for that, and honestly doesn't need it at that scale. What it does need is for the one blocking relationship that actually matters this week to not get lost.
"Blocked by" data does something specific once it exists: it turns from a static fact into a compounding record. Every dependency you set is another data point about how this particular project's work actually depends on itself — and that record is the exact input a Gantt view or a risk scan uses to tell you something you didn't have to notice yourself.
Where the dependency data shows up again
Set dependencies once, and they don't just sit on individual task pages. A project's Timeline (Gantt) view, if your plan includes it, lays out tasks colored by status rather than by person — a deliberate choice that keeps the view answering "what's the state of the work" instead of turning into an implicit leaderboard of who's behind. Hovering over a task bar on that timeline shows its blockers, which means the one piece of information that actually explains a schedule risk — "this is stuck, and here's why" — is available at a glance instead of requiring someone to click into every task individually.
On Max plans, there's a further step: AI Risk Analysis reads the dependency graph you've already built through ordinary "Blocked by" entries and turns it into an early warning rather than a static diagram — surfacing where a blocked chain is likely to cause a real problem before it actually does. It's worth being precise about what that is: it's not guessing at risk from nothing, it's reading data your team already entered while doing normal task management, and pointing out a pattern in it that's easy to miss by eye across a project with dozens of tasks.
The honest limits
Dependencies only exist between tasks in the same project — there's no cross-project "this client's launch is blocked by that other client's unrelated task" linking, which is fine, because that's rarely a real relationship anyway. And a dependency is only as good as someone remembering to set it. It doesn't automatically detect that two tasks are related; a person still has to notice and mark it. The feature's value is in making that judgment durable once it's made, not in making the judgment for you.
How to do this in SparkyProjects
Open any task's detail page and use "Blocked by" to mark it as depending on another task in the same project — remove it the same way if the dependency stops being true. If your plan includes Timeline view, switch a project into Timeline to see tasks colored by status with blockers visible on hover. On Max plans, AI Risk Analysis reads this same dependency data to flag likely trouble spots automatically.
Task Dependencies, Timeline, and AI Risk Analysis availability by plan is listed on the features page.
None of this requires a small agency to start acting like a bigger one. It just requires writing down the one thing everybody already half-knows, in the one place it'll actually still be true next week.