Project Management for Startup Operations

Ops at a startup is a catch-all, not a department

At a company with a real ops function, operations means specific things: procurement, facilities, vendor management, compliance. At a startup, "ops" usually means one person, sometimes a founder, sometimes the first ops hire, who's fielding a vendor payment, a laptop that needs replacing, a compliance filing deadline, a landlord question and an ad hoc request from engineering, often within the same hour. It's genuinely a different shape of work from a specialist function: less depth per task, far more variety, and almost no slack to absorb a surprise.

That variety is exactly what breaks most process. A tool built around one workflow, tickets, or campaigns, or sprints, assumes the requests it handles look alike. Startup ops requests don't. A board that can hold "renew the office wifi contract" next to "get the new hire's laptop shipped" next to "file this quarter's GST return" without forcing any of them into an unnatural shape is doing something a specialist tool usually can't.

The other constant is that there's rarely a backup. If the one person running ops is out for two days, whatever's in their head stays there. That's why the goal here isn't more process, it's process light enough to actually get used, and visible enough that it doesn't all depend on one person being reachable. A ten-person company usually can't afford a week of "we'll sort it out when they're back" on something like payroll or a lapsed vendor contract, so the bar for the ops workflow isn't thoroughness, it's that the essentials survive one person's absence without anyone having to reconstruct them from scratch. That's what the ShipSprint triage inbox and wiki are actually for here: not process for its own sake, but a place the essentials live that isn't one person's head.

Low process overhead is the actual requirement

The honest test for any tool here isn't "does it have enough features," it's "does using it correctly take less effort than not using it." A triage inbox only works if logging a request into it is faster than firing off a DM. A wiki only survives if writing the SOP down took no more effort than explaining it verbally once. That's the bar the whole ops workflow has to clear at this stage, because there's no second person to enforce compliance with a heavier process. Whoever's doing the work will simply route around anything that's slower than the shortcut.

The WIP limit does something specific for a stretched team: it doesn't add a rule so much as it makes an existing problem visible. When a column fills up, that's not a policy being enforced. It's a fact that was already true (this person has taken on too much) finally showing up somewhere instead of staying invisible until something drops.

The founder-does-everything phase has a shelf life, but it's real while it lasts

In the first several months, "ops" is frequently whichever founder isn't in a sales call or writing code that hour, handling whatever came up between the two: a vendor invoice, a co-working space booking, a question from a new hire about expense reimbursement. It's not a role anyone applied for, and it's rarely the founder's strength, which means the instinct is to spend as little time on it as possible and hope nothing falls through in the gaps.

The risk in that phase isn't that any single task is hard. None of them are. It's that there's no record of what's already been handled and what hasn't, so the same question gets re-answered from scratch, or worse, something genuinely time-sensitive (a compliance deadline, a contract renewal) slips because it was never written down anywhere durable in the first place. A lightweight board that just holds "what's outstanding" does more good here than any formal process would, because the alternative isn't a better process. It's no record at all.

When ops stops being one person's job

The transition from "whoever's free handles it" to an actual first ops hire is usually gradual and often happens without anyone declaring it. One day there's simply enough recurring work that a founder can't keep absorbing it between other things, and someone gets hired or reassigned to own it properly. What makes that transition smoother is whether there's already something for the new person to inherit: a board with the recurring tasks on it, a wiki page with the vendor list and the filing deadlines, rather than a founder trying to transfer two years of undocumented context out of their head in a single handover week.

That's really the argument for starting with structure this light rather than no structure at all. Not because a five-person company needs process for its own sake, but because whatever exists on day one is what a first real ops hire inherits on day two hundred, and starting from a board with history beats starting from a blank page and a founder's fading memory of how things have been done so far.

Compliance deadlines don't care how busy the week is

A GST filing, a labor-law registration, an annual compliance renewal: these dates don't move because the week got busy, and they're exactly the kind of task that gets silently deprioritized when there's no external pressure reminding anyone until a penalty notice arrives. Unlike a client request, nobody's chasing an ops team about a filing deadline until it's already missed, which makes it the easiest category of task to lose in the noise of everything else landing in the same week.

Putting recurring compliance dates on a board with their own reminders, rather than trusting a calendar entry set up once and never revisited, means they compete on equal footing with the louder, more visible requests instead of losing by default because nobody's actively pushing for them.

What it costs

• Free: 5 users, 2 projects, no time limit, no card required. The right tier for most startup ops functions in year one.
• Team: ₹299 per user per month, or ₹2,899 per year, up to 40 users. For once ops has grown past one person into a small team.
• Business: ₹599 per user per month, or ₹6,499 per year. Adds forecasts and the owner command center, generally not needed until there's more than one function to coordinate.
• Every paid plan opens with a 14-day full-access Business trial, sample project preloaded, no card required.