A few of the things you check every morning sit outside your mail. Point can run that round for you. You say what to check and how often, in plain words. Point goes out and looks, then comes back with the answer in one line. On the days there’s nothing to report, Point stays quiet.
- The trigger here is a clock. You ask for a round to be run, and Point goes out and looks.
- The sentence has two halves: what to check, and how often. The interval is the half people get wrong. It comes from how fast the answer can change and how fast you could act on it.
- What comes back is a finding. The reading is already done, and that’s the whole reason this is worth handing over.
- A quiet check is free, so you can run more of these than you can run watches on your mail. What they ask of you instead is a small habit of confirming, once, that the answer is true.
The round you do before you open your mail
Nearly every owner has one, and it rarely gets counted as work.
Four or five things you look at before the day properly starts, every one of them outside your mail. The state board’s notices page, because the continuing education rules moved once and you’ve been half-watching ever since. The status page for the practice management system, because when that’s down your morning goes somewhere else entirely. The page where a supplier posts current rates. A docket, or a filing status, for the one matter your quarter is stacked behind. The agency page that will eventually say when this season’s forms actually open, which you’ve checked most mornings since July, and which has said the same thing every time.
Six minutes, maybe. It feels like less, because it’s spread across the first hour and mixed in with coffee. The minutes are the cheap part. The cost is that the round has to live in your head every morning. The morning it slips your mind is the morning it mattered.
That round is what this page is about. You describe it once, in ordinary words, with an interval attached. It becomes a standing check that Point runs for you. The check lands in your task list, in the same place as every other watch and everything else you’re tracking. You read it back there, and you stop it there.
One question decides whether you’re in the right place. It’s worth asking before you write anything.
Is this thing ever going to email you about itself? Plenty of them will. Portals send a notice when a status changes. Vendors mail their customers when they publish new terms. When a message is going to arrive, waiting for it is faster and surer than going and looking. That’s the other arrangement: one sentence describing the arrival, checked against every message that lands. Telling your inbox to notify you when something arrives is that, in full.
When the only way you’d ever know is by going and looking, you’re in the right place. That’s most of the morning round. You check those pages by hand because nobody has undertaken to tell you.
What to check, and how often
Describe it out loud, in the words you’d use with somebody sitting in for you next week. One line of English, with an interval in it. The whole setup is that sentence. Point states the instruction back before the check is saved.
The first half of that line is the same discipline as any other watch. Name the thing precisely. Describe what you’re actually looking for, rather than the words you expect to see. That part is covered properly next door, and it works the same way here.
The second half is new, and it’s where people go wrong. The interval is the answer to two duller questions, and you take the slower of the two. Everything on your morning round matters, which is why it’s on the round.
How fast can the answer actually change? A fee schedule gets revised twice a year, so it suits a slow check. An hourly one asks a question that comes back with the same answer a few thousand times in a row.
How fast could you act on a change? This is the one people skip. Say a change would only ever be dealt with on Monday morning, because that’s when you have the file open and the client on the phone. An hourly check then buys you six days of knowing something you can act on only on Monday.
Take the slower answer, and that’s your interval. It’s almost always daily, and daily is a good default for a reason worth noticing. A daily check on something that changes twice a year stays silent on the 363 days it finds nothing. What a check costs you is only ever what reaches you.
Four that are worth writing:
- Check the state board’s notices page every morning and tell me if anything new goes up about license renewal. Daily, and silent for months at a time, which is exactly right.
- Once a week, check whether Brightside has published new rates, and tell me what changed. Weekly, because their rates move on a quarterly cycle and you renegotiate in one sitting anyway.
- Every morning, check the status page for our practice management system and tell me if there is an active incident. Daily, and this one earns its place because the answer changes your morning rather than your quarter.
- Check the county’s docket for the Ruiz matter each day and tell me when it moves. A named source, a named matter, and an end that arrives on its own when the matter closes.
And two that will disappoint you.
- Check the news every hour for anything about our industry. No source, no question, and an interval chosen out of anxiety. What comes back on an hour where something happened is a summary of an industry. That’s a thing to read, rather than a thing to act on.
- Check whether the client has paid. That answer almost always lives in your accounting system or your bank, behind a login. The honest fix is usually that the system already emails you, and a watch on that message is the tool for it.
Start with one, from the round you actually do, rather than the five you could list if you sat down to it. You want it running long enough to find out whether the answer it gives you is the answer you’d have written yourself. That takes a week or two of looking at both.
A finding, not a pile of links
Here’s the part that decides whether the whole arrangement pays.
When the check runs, Point searches the live web and reads what’s there. What comes back is one line saying what was found. One line, in your own terms. It’s the answer to the question you asked, in the form you’d have written it down in after doing the looking yourself.
That’s the whole return. The looking was always the cheap part. The expensive part is the reading and the deciding. Opening the page, finding the paragraph that changed, working out whether it applies to you. A check that came back with links would move the work six inches and leave all of it with you.
The finding goes where the check lives, in your task list. The record of what it has found builds up in the same place, so the history of the check reads as one list. Your mail stays out of it, which matters more than it sounds if your inbox is already full of software telling you about other software.
Most days there’s nothing to say, and Point says nothing. The silence here has its own shape, and it’s worth knowing. A watch on your mail sits dormant until a message comes. A scheduled check works every morning. Point goes out, looks, reads, and concludes that today matches yesterday. The work happened and you heard about none of it, which is the right outcome. It’s also why a quiet check earns a habit of its own.
The full range of what Point can hold on your behalf sits on the benefits page. What a check has found, how often it has run, and how to end it all live in the record behind it. That’s one click, and seeing and stopping every watch covers it.
What a scheduled check can reach
Point searches the live web and reads what’s there. So the natural fit is anything you could look up in a browser without signing in. A notices page, a status page, a rate card, a public docket. That covers more of your morning round than you’d guess. Most of what you check is published on purpose, by somebody who wants it read.
Three cases sit outside that, and they’re worth knowing now, before you write a check that quietly answers the wrong question.
A page behind a login. A client portal, your bank, a payroll system. Those sit off the open web, however well you describe them. Portals almost always send mail when something changes, so the answer for those is a watch on the arrival rather than a check on the clock. When six systems are already mailing you about the same event, and the one that matters is buried among them, that’s its own job. It belongs to turning ten alerts into one.
Something you need to the second. A check runs on your interval. The gap between the world changing and you knowing is at most that interval, and on average half of it. For a fee schedule, an interval of a day is invisible. For a live system where an outage costs you money by the minute, a daily check reads as a calendar reminder, and what you want is a proper monitoring service that pages somebody.
A judgment you’d want to make with the page in front of you. A check is good at what does this say now and what changed. It’s weaker on whether this is worth doing, because that answer sits with you rather than on the page. When the finding would only ever be the start of you reading the thing properly, ask for the finding and keep reading the thing properly.
A check shares one thing with every other watch. It can hold a next step as well as a question. Tell me when the docket moves, and put the call to the client on my list for that afternoon. Each kind of action carries its own position on the autonomy dial. Out of the box, each one starts at review, which means the work gets prepared and waits for your go. When a watch lines up the next piece of work is where that lives, and setting how much your inbox does on its own is where each position on the dial begins and ends.
How you know a quiet check is still working
A watch on your mail hands you the message. You click through from the fire and read the words the other side actually wrote. Whether Point read them correctly is something you settle yourself in four seconds.
A scheduled check hands you a sentence Point wrote after reading a page. That’s a different thing to trust, and it earns one specific habit. You’d do the same with a new colleague’s first week of work.
The habit is small, and you do it once. In the first week of a new check, go and look at the source yourself on two or three of the days. See whether you’d have written the same line. That’s five minutes in total, spent while the stakes are low, and it settles the question properly. Either the finding matches what you’d have said, and you stop looking, because the check is doing what you set it up to do. Or the page turns out messier than you remembered, and the line thinner than you need. Either way, you learned it in a week rather than in October.
After that, one rule is worth keeping. On the day a finding would make you do something expensive, look yourself anyway. That’s what you’d do with a line from a colleague too, and it costs you ninety seconds on the handful of days it applies.
Then there’s the failure that arrives quietly. Pages move. Publishers stop publishing. The matter closes in June, and the check is still running in October, asking a docket about a case that finished. A check that has been silent for four months looks the same either way: the world was quiet, or the page has been returning an error since the spring.
The defense is a date rather than vigilance. Once a quarter, open the list and read the checks you have running. For each one that has stayed quiet, ask the question that matters: does this still point at something, and do I still want the answer. Ending one takes a click. The log picks up a timestamped line for it, with a link back to the task, so a check you retired in October is still accounted for in January.
This is a different discipline from the one that governs watches on your mail. There, interruption limits the number you can run. Every fire costs you attention, so the careless ones tax the careful ones, and the answer is to keep the list short. That arithmetic is set out properly next door, and it stays there. A quiet check is free while it’s quiet, so you can reasonably run more of them than you can run watches on your mail. The limit here is a different one: run the checks you’ve kept confirming. A line you act on while wondering where it came from costs you more than the six minutes you were trying to save.
Where an alert or a feed still fits
Most people arrive here having already tried something, and the something is usually a Google Alert. It’s worth being exact about what that does, because the two tools answer different questions.
Google’s own help page puts it plainly: “You can get emails when new results for a topic show up in Google Search” (checked September 6, 2026). The page also says you can change “How often you get notifications” and “What accounts get the alert” (checked September 6, 2026). So the mail can be slowed down, and pointed somewhere other than your main address.
Two differences follow. Only one of them is about delivery.
The first is that an alert arrives as mail. Sent elsewhere and batched, it’s still a message landing in a mailbox. The thing you set up to watch the outside world has added to the pile inside it. A scheduled check leaves your mail alone.
The second is deeper, and it’s about the question being asked. An alert answers has anything new been published about this term. A scheduled check answers what does this source say this morning. Those two come apart on exactly the cases that tend to matter. A status column flips from pending to accepted. A fee is edited in place on a page that has existed for three years. A date moves quietly. In each one the page simply says something different than it did yesterday, and a search for new results has no reason to notice.
An RSS feed is the same shape with better manners. It stays out of your mail, and it’s free. If the source you care about keeps a real feed, and you already have somewhere you read feeds, that’s a perfectly good answer and cheaper than anything on this page. It carries the same limit. It tells you when the publisher publishes. What the page says this morning is a different question.
And there’s a case where an alert is genuinely the better tool, worth saying rather than skating past. Broad, open-ended watching across the whole web, mentions of your firm’s name anywhere at all, is a job where you want breadth and you want the raw links. You’re going to judge each one yourself, and a single-line finding would be the wrong shape for that. Use the alert there. Use a check for the four or five specific pages you actually open every morning, which is a different job and always was.
Common questions
Can Point check a website for me every day?
Yes, and it takes one sentence rather than a setup. You say what to check and how often, in plain words. Point saves that as a standing check and runs it on your schedule, searching the live web, reading what’s there, and coming back with what was found. The check sits in your task list, clear of your inbox, where you can read it back and stop it.
How often should the check run?
Take the slower of two answers: how fast the thing can actually change, and how fast you could act on a change if it came. Whichever is slower is your interval. Daily suits most of a morning round, including things that change twice a year. A check that finds nothing says nothing, so a quiet day costs you no attention.
Does the finding come to me as an email?
No. The finding goes where the check lives, in your task list, with the record of previous runs beside it. An alert service delivers by mail even when you send it to a second address, and that’s the difference. If your reason for reading this page is that too much software already writes to you, that’s the part worth knowing.
What if the thing I check is behind a login?
Then a watch on the arrival is the better tool, because the open web has no page there for Point to read. A portal will almost always mail you when something changes. You write one sentence describing the message you expect, and Point checks it against everything that lands. Telling your inbox to notify you when something arrives covers how to phrase it.
How do I know a check that never finds anything is still working?
You confirm it once rather than worrying about it forever. In the first week, look at the source yourself on two or three days, and see whether you’d have written the same line. After that, put a quarterly note in the calendar to read the list. For any check that has stayed quiet throughout, ask whether it still points at something. This matters more here than for a watch on your mail, where a message you can open is its own proof.
Is this the same as a Google Alert?
No, on both halves. Google describes Alerts as sending you emails “when new results for a topic show up in Google Search” (checked September 6, 2026), which is a question about publishing. A scheduled check asks what a specific source says today, so it catches the page edited in place that never produced a new result. The other difference is delivery: the alert arrives in your inbox, and the finding waits in your task list.
The short version
Part of your morning round lives outside your mail, and always has. You describe that round once, with an interval attached, and Point runs it. Out to the live web, reading what’s there. Back with one line saying what was found. Quiet on the days there’s nothing. Pick the interval from how fast the answer can change and how fast you could act on it. Keep it to sources that are actually published, because a portal behind a login is a watch on the mail it sends instead. Spend five minutes in the first week checking that the line Point gives you is the line you’d have written. That’s the one thing this arrangement asks of you. After it, you get the six minutes back every morning, and the round can leave your head.