The applied science of holding many deadlines in mind.

Your deadlines are a shape, not a list. Past a certain number, your mind stops tracking them as separate line items and starts feeling them as one uneven mass: where they cluster, where they thin out, where the next wall is. That distribution is the shape. Deadlinewatch renders it, so you stop carrying it in your head. You decide what to do with it. That's the whole idea, and everything below is how I got there.

[01]What I felt

The problem.

You may never have put it this way. For many years I didn't. But past a certain number of commitments held in your head, you stop treating your due dates as line items and start treating them as a lump: uneven weights, sharp edges, bumps you brace for without quite looking. Your mind adapts to that overhanging shape on its own, predicting, shifting, negotiating, all of it running in the background while you try to do the actual work.

I spent over 20 years in deadline-heavy professional work, across four continents. Formal submissions, presentations, reports, reviews: managing them was the daily job, and I tried every aid and system I could find. Nothing quite held. Circumstances changed too fast, too unpredictably, in too many directions for a flat list to keep up. So I started working the problem the way I'd work any other one: take inventory, analyze, test, solve. This tool is the product of five years of loose research and two years of building. What follows is what I found, and how it became the thing in front of you. If the idea interests you, the cornerstone essay Your deadlines are a shape, not a list argues it in full.

[02]What I read

What the research says.

When I went looking for the work behind the experience, it was older and better established than I expected. None of it surprised me. It gave names to things I'd felt for years.

Working memory holds only a handful of things at once, the famous "seven, plus or minus two" from George Miller's 1956 paper, and fewer still for anything you can't group into a chunk. A dozen live deadlines don't chunk. Past that handful you stop holding them and start spending attention to keep them from slipping. You're also a poor judge of the load. The planning fallacy, named by Kahneman and Tversky, is our reliable habit of underestimating how long our own work will take, which is why an open-looking Monday turns into a crunch by Thursday. And what helps isn't willpower. Ariely and Wertenbroch found that people do better against deadlines they can see than against the ones they carry in their heads. Structure you can see beats discipline you have to summon.

The rest of the literature, on procrastination and the cost of carrying open loops, rhymes with all three. None of it is news to anyone who has done deadline-heavy work for long. It only puts names to what you already know in your body by Wednesday afternoon.

[03]From research to design

The four principles.

The research above doesn't tell you what to build. It tells you what to respect. Four rules came out of trying to respect it.

Externalize the weight

You hold most of this in your head by default: what's active, what's deferred, what matters more than its date suggests. The controls on a deadline (Status, Snooze, Priority, an email reminder) put each of those on the screen instead, where you can read them without effort. Move a deadline up or down, hand off the remembering, check where things stand at a glance. You feel the relief in proportion to how much you stop carrying. Two later controls answer 2 more questions you used to answer from memory. The padlock records that a date isn't yours to change, so you never re-derive which deadlines you can negotiate. The working calendar records the days you don't work, so the tool can tell you when a deadline falls on one. That leads straight to the next principle, the shape itself.

Render the shape

The second cost is the topography: which days are dense, which weeks are quiet, where the work stacks up. A list can't show that. A shape can. The tool renders it two ways. The Dashboard reads everything you're carrying as a single ordered list, soonest first. The Forecast reads the weeks ahead as terrain, through two instruments. The Curve traces the rise and fall of the load across whichever outlook you pick, 90 days, six months, or a year. The Heatmap shades the same load as a grid, darker where deadlines bunch, and pages back and forward through the seasons either side of today. See a crunch coming and you can move the deadlines you control, then watch the shape redraw, instead of running the arithmetic in your head.

Describe, don't direct

A tool that tells you what to work on next is claiming an expertise it doesn't have, and taking the part of the job you're actually paid for. The tool doesn't know which client is hot this week, which deadline your reputation rides on, which one has slack you can borrow against. You do. So the tool reports facts and you decide priorities. Everywhere it could have used a directive ('do this next'), it uses a plain label instead: Priority, Overdue, Today.

Tell the truth, even when it's uncomfortable

The fourth principle is the one I almost didn't write down, because it should be obvious. The tool never hides a fact to make you feel better. A snoozed deadline that's overdue says so on its chip ('snoozed until Thursday, was due 3 days ago'). The Dashboard never suppresses an overdue row. There are no completion celebrations beyond a brief card-fade. The only thing a deadline tool has to offer is the truth about your week. Bend that truth in any direction, even kindly, and you stop trusting it, at which point the tool is worse than no tool at all.

A few of the numbers running inside the tool are worth showing plainly. Not because you need them to use it, but because the math is real and I want it on the table.

Per-day intensity

The Heatmap shades each day a deeper red the more deadlines cluster around it. The intensity at any date is the exponential-falloff sum of every nearby deadline:

I(d) = Σ exp(-distance(d, deadline) / τ)
       where τ = 2.5 days, distance cutoff = 7 days

A lone deadline contributes 1.0 to its own day. A second one a day later adds about 0.67. Three on the same day stack to 3.0. That number sets the cell's red: a single deadline a light wash, a pair a mid-tone, a dense cluster the darkest red, with the scale anchored near a three-deadline cluster. The shade ceiling floors at 2.4 times a single deadline, so a quiet stretch still reads in proportion. The math is symmetrical in time, bounded (anything past 7 days contributes nothing), and deterministic. Same inputs, same shade, every time.

Curve smoothing

The Curve reads the same field as a single rolling line. Where the Heatmap uses a sharp exponential kernel, a crisp cusp on each deadline's day, the Curve sums a gaussian bump per deadline, a rounded hill instead of a spike, so adjacent deadlines blend into gentle terrain rather than an EKG trace:

C(d) = Σ exp(-distance(d, deadline)² / 2σ²)
       where σ = 2.5 days at the 90-day outlook,
       widening as the outlook widens

The two instruments scale against the same reference and use it differently, on purpose. The Heatmap saturates at that reference, so a busy fortnight always reads darker than a quiet one and the shading means the same thing wherever you page to. The Curve takes the same figure as a floor and grows past it when a day is heavier than that, so a genuine crunch still crests above everything around it. Same load, two readings.

Templated, not predicted

The observations on the Dashboard come from fixed detector functions, not a language model. One reads: "3 deadlines next week, Wednesday is heaviest." Each detector checks a specific condition and emits a templated sentence when it's met. Same data in, same observation out. You can audit any of them by checking the numbers underneath. Nothing happens that isn't shown.

[06]Boundaries as principle

Where I hold the line.

The choices below are deliberate. Each maps back to a principle above and stays firm for the same reason.

No rankings or scores

The tool never tells you which deadline to work on next, and no deadline carries a number that ranks it against the others. No 'top priority,' no suggested order, no value-weighted index. The math you see is days-until-due, and the order is yours to choose. Anything else would be the tool pretending to know your work better than you do.

No AI claims

The pattern observations come from deterministic templates, not a language model. Nothing here is 'smart,' 'intelligent,' or 'AI-powered.' When the Dashboard surfaces '3 deadlines next week, Wednesday is heaviest,' it's because a function counted 3 deadlines and emitted a templated sentence. That's all it is, and I'd rather say so than dress it up.

No theater, no game

No 'act now,' no countdown timers, no red dots demanding attention. No badges, no leaderboards, no confetti when you complete a deadline. There's a small streak counter on your profile, off to the side, easy to ignore. Everything else a game would add, the tool leaves out. It's an instrument, not a game, and you already feel enough pressure from the deadlines themselves.

No dates worked out from rules

Deadlinewatch holds the dates you give it. It never works one out for you. It ships no holiday data in any form, it makes no claim about your jurisdiction, and when a deadline falls on a day you marked off it tells you and offers to move it earlier, never later. Rolling a date forward past a closed day is a real rule for certain filings in certain places, and it is flatly wrong for a great many contracts, where a date that falls on a Sunday is due on the Sunday. To offer that move the tool would have to know which of those you are holding. It doesn't, and it won't pretend to.

No unlimited claims

Your account holds up to 1,000 active deadlines and up to 100 active streams, and repeating series are created 2 years ahead at a time. Completed work and anything you set aside sit outside those numbers. I would rather publish a ceiling you can check than an unlimited one nobody has tested, because an untested claim fails the same way every time: a view quietly leaves something out, and looks composed doing it. Testing found where the scope ended, which is what turned the ceiling into a stated feature instead of something you would meet by surprise.

No mass reminders

Email reminders are on by default, but calm by default. Each deadline carries its own reminder, 3 days out by default and adjustable per deadline, never a blanket all-at-once send across your whole list. You can dial any single deadline up or off, and one master switch turns the whole channel off. The default is 'remind me,' not 'flood my inbox.' Forgetting a deadline is the worse mistake, so the friction sits on opting out, not opting in.

[07]The interview note

How I asked.

Before building, I talked with a few dozen people who carry deadline-heavy work. I asked how they keep track, what they've tried, what they discarded, what they wished worked. I wasn't running a formal study. I was checking my own experience against theirs. The themes were consistent: lists work until they don't, the cost shows up as fatigue, and the available tools either over-promise or under-deliver. 20 years of doing the work, a reading of the research, and those conversations: three sources, weighed against each other, became one tool.

[08]Signed, Conrad

A note from the maker.

I built Deadlinewatch the way I built everything else in my career: as an instrument I would use myself before asking anyone else to use it. The math, the principles, and the lines I won't cross aren't a marketing position. They're how I actually work, transferred into software. And I still answer the email myself.

If you want the rest of that story, why I built it, what I drew on, and what it cost, I told it by the fireside.

If any of this resonated, the demo walks the mechanics in practice, and the pricing is one number and a trial. If you'd rather read than click, the rest of the thinking lives on the blog. Thank you for reading this far.

Conrad

Maker of Deadlinewatch