Plain, senior, hype-free answers for Canadian small businesses. 2026-06-10T00:00:00Z https://quieteye.ca/ Kirsten Weagle hello@quieteye.ca 2026-05-20T00:00:00Z https://quieteye.ca/blog/what-can-ai-do/ <p>For a small service business today, AI reliably does about six things: it answers leads instantly, follows up until people reply, drafts your routine writing, schedules and dispatches, asks for reviews, and handles booking. That's it. Not &quot;transform your business.&quot; Each one quietly replaces a job someone is doing by hand, usually badly, usually at the wrong hour.</p> <p>I run a brand practice, not an AI startup, so I have no reason to oversell this. Here's the honest version.</p> <h2>So what does it actually do, in plain terms?</h2> <p>Here's the deflated inventory. What it does, roughly what it costs to keep running, and the job it replaces.</p> <table> <thead> <tr> <th>What it does</th> <th>What it costs to run</th> <th>What it replaces</th> </tr> </thead> <tbody> <tr> <td>Answers every lead in seconds, day or night</td> <td>Low, monthly</td> <td>You, checking your phone at 9pm</td> </tr> <tr> <td>Follows up by text and email until they reply</td> <td>Low, monthly</td> <td>The follow-ups you meant to send</td> </tr> <tr> <td>Drafts newsletters, posts, replies for approval</td> <td>Low, monthly</td> <td>An afternoon of writing a week</td> </tr> <tr> <td>Builds and texts the crew schedule</td> <td>Low, monthly</td> <td>The 6am whiteboard ritual</td> </tr> <tr> <td>Asks for a Google review after each job</td> <td>Very low</td> <td>Nagging customers, or not asking</td> </tr> <tr> <td>Handles routine booking and questions</td> <td>Low, monthly</td> <td>Phone tag and the back-and-forth</td> </tr> </tbody> </table> <p>None of these is exciting on a stage. All of them are quietly worth more than the marketing tactic you were about to buy.</p> <h2>What can it not do?</h2> <p>It can't decide what your business should say, or which of these is worth doing first. It can't fix a weak position or a service nobody wants. AI is leverage on a decision that's already right. Point it at the wrong thing and it just gets you to the wrong place faster.</p> <h2>Where should I start?</h2> <p>One workflow. The single job that eats your evenings or your mornings. Not a strategy, not a &quot;roadmap.&quot; You automate one thing, it earns its keep in a month, and that pays for the next one. Anyone selling you a transformation is selling you a project, not a result.</p> <h2>Is it expensive?</h2> <p>Far less than people expect, and usually less than the one tactic that isn't working. For most of the businesses I see, this isn't new spend at all. It's the same budget, redirected out of something that leaks and into something that runs.</p> 2026-06-03T00:00:00Z https://quieteye.ca/blog/marketing-leak/ <p>If your marketing stopped working, the fix is almost never more marketing. Money is leaking somewhere between your message, your website, and what happens after a lead arrives. Until someone finds the leak, every new dollar follows the old ones out.</p> <h2>Leak one: a message anyone could claim</h2> <p>Read your homepage's first line. Could a competitor put their logo on it? &quot;Quality work. Fair prices. Family owned.&quot; If yes, you're marketing the category, not your business, and the cheapest competitor collects.</p> <h2>Leak two: the gap after the click</h2> <p>A lead arrives Friday 9pm. You reply Monday. They called three others Saturday. You paid for that lead and it leaked out the bottom. Answering instantly, well, at 3am, now costs less than one underperforming tactic. That wasn't true five years ago.</p> <h2>Leak three: tactics on no diagnosis</h2> <p>Website from one vendor, ads from another, a CRM nobody set up. Treatment without diagnosis. When marketing &quot;stops working,&quot; usually nothing was working together to begin with.</p> <h2>What to do this week</h2> <p>Read your homepage like a competitor. Time your own enquiry response on a Saturday. List every tool you pay for and the question it was meant to answer. Then resist the next tactic until someone looks at the whole system.</p> 2026-06-10T00:00:00Z https://quieteye.ca/blog/6am-schedule/ <p>If you write the crew schedule by hand every morning, you don't have a scheduling chore. You have a system, and you are the server it runs on. It boots at 6am, takes the best half-hour of your day, and never gets cheaper. The good news: this is the easiest thing on the whole list to hand off.</p> <h2>Why is this worse than it looks?</h2> <p>Add it up. Half an hour a day is roughly a full working week every quarter, taken from the front of your sharpest hours. But the real cost isn't the time. It's that the schedule lives only in your head and on a whiteboard, so nothing else can use it. The crew texts you to ask where they're going. A job moves and three people need telling. You're the integration layer, by hand.</p> <h2>What would replace it actually do?</h2> <p>It holds the jobs, the people, and the rules, then builds the day and texts each person their stops. A change in one place updates everyone. No app for the crew to learn, it arrives as a text. You approve, or you let it run. The morning ritual becomes a thirty-second glance.</p> <h2>Isn't my scheduling too specific to automate?</h2> <p>Almost everyone says this, and almost everyone is wrong about the part that matters. The judgement calls that are genuinely yours stay yours. What gets automated is the copying-out: turning a decision you've already made into twelve texts. That's not your expertise. That's the tax you pay every morning for not having a system.</p> <h2>What's the smallest first step?</h2> <p>Pick one crew, one week. Get the schedule sending itself by text. If it saves the half-hour and the phone stops buzzing, expand it. If it doesn't, you've lost almost nothing. That's the whole point of starting with one workflow instead of a transformation: the bet is small and the relief is fast.</p>