From burnt-out at 6am to a week of marketing done — how a Mount Eden café owner uses Marlaio

6:32am, Monday — “I haven’t posted in two weeks”

Sarah is dialling in the second grind of the morning when she realises the last time she posted on Instagram was on a Tuesday she can’t quite place. Maybe the long weekend before Queen’s Birthday. Maybe earlier. The cafe’s feed has three latte photos in a row from May, and one mediocre shot of the cabinet that she took because she felt guilty.

She opens Marlaio on her phone while she steams milk. Instead of a blank text box, Marla — the AI co-pilot — is already waiting: “Hey Sarah. Long time. Want me to look at this week and suggest what to post?”

Sarah taps yes.

How Marla figures out what’s actually worth posting

Marla doesn’t guess generically. She quietly checks a handful of signals before she suggests anything:

• The NZ calendar. Matariki is coming. School holidays. Long weekends people actually take.
• The weather. Auckland forecast is cold and wet. Hot drinks sell better. Customers’ intent shifts.
• Sarah’s last few packs. What worked? The flat white reel got 230 likes. The cabinet shot got nine.
• Sarah’s reviews. Three people in the last fortnight mentioned the cinnamon brioche. Two mentioned how warm the corner table felt on a wet day.

Marla suggests a focus for the week: a quiet promotion of the weekend brunch menu, leaning into the warm-corner-on-a-cold-day angle that customers are already telling her matters.

6:41am — Sarah uploads four photos from her camera roll

This is where most AI marketing tools quietly fail. They have two options for images: either you spend an hour finding the right photos and writing the right caption, or the tool generates a beautiful AI fake — a glossy “café” that doesn’t exist, with a barista who isn’t Sarah, in a building that isn’t hers.

Sarah opens /account/photos and drags four images off her camera roll into the upload zone:

• A latte she pulled five minutes ago, foam still sitting clean.
• The cinnamon brioche on a slate plate.
• The corner table from last Saturday, sun coming through.
• A wide shot of the cabinet, taken at 7am before the rush.

Marlaio runs each through an automated content check, marks them ready to use, and saves them privately to her account. No one else sees them. They’re not used to train any model. They’ll get used when she generates her next pack — as the actual source images, not as “inspiration” for a fabricated version.

This part matters more than it sounds. New Zealand’s Fair Trading Act 1986 section 9, and the AANA Code of Ethics 2.6 in Australia, both treat misleading imagery as a real problem. An AI-generated café that doesn’t look like Sarah’s, captioned as if it does, is squarely the kind of thing both laws care about. Sarah doesn’t have to think about any of that — Marlaio just defaults to her actual photos.

6:48am — One pack, four channels, in 60 seconds

Sarah picks the intent: “promote this weekend’s brunch menu, lean into the warm-corner-on-a-cold-day angle.” She taps generate. A status bar moves. About a minute later Marlaio has drafted:

• An Instagram caption — warm, specific to a Mount Eden cold morning, no scarcity language that could trip the FTA.
• A Facebook post — slightly longer, fewer hashtags, a clearer call to come in.
• A short email for her regulars list — neighbourly, not salesy.
• A Google Search ad — three headlines and two descriptions, ready to drop into Google Ads.

Each draft is paired with Sarah’s own photo — the latte for Instagram (square crop), the corner table for Facebook (landscape), the brioche for the email header.

6:52am — She reads, edits one line, taps approve

Sarah doesn’t love the Instagram caption’s opening line — “Cold morning. Warm corner.” — she thinks it’s a touch too poetic. She taps the caption, deletes the full stop, types “Cold morning, warm corner — and a flat white that knows it’s your day off.”

She approves each draft individually. Marlaio does not auto-post anywhere. The Instagram caption gets copied to her clipboard with an “Open Instagram” button; she opens the app, pastes, picks her photo from the ones Marlaio prepared, and posts. Same for Facebook. The email goes out to her list with one more confirmation. The Google ad copy gets saved to her account so her ads manager (or her, if she’s brave) can paste it into Google Ads later.

Total time: 15 minutes between the first grind and her first customer walking through the door at 7am.

Saturday — “Was it actually worth it?”

Saturday morning Sarah opens Marlaio again, this time on her laptop with the second cup of coffee. The activity dashboard shows her week. In an illustrative version of how this might go, the dashboard could surface something like:

• Instagram post performance (likes, saves, comments)
• Facebook post performance (reactions, shares)
• Email open rate against her regulars list
• Google ad clicks and calls to the café

What those numbers actually look like depends entirely on Sarah’s list size, her location, her hours, the season, and a hundred other things Marlaio can’t control. We won’t pretend otherwise. What Marlaio does control is putting the dashboard in front of her at all — so she can see what worked, and skip what didn’t, next week.

In our hypothetical version of Sarah’s Saturday, the cafe is a little busier than usual. Whether that’s because of the Instagram post, the weather, or just Saturday being Saturday is the kind of question marketing has always struggled to answer cleanly. Marlaio doesn’t pretend it can answer it for her either.

What this is meant to show

This isn’t a sales pitch. It’s a walk through the actual shape of using Marlaio as a small-business owner. A few things to take from it if you’re evaluating:

• 15 minutes a week, not 4 hours. The unit of work is "open the app while milk steams", not “block out Sunday afternoon”.
• Your photos, not AI fakes. The images on your channels are actually you. Customers can tell. The law cares.
• You approve everything. Marlaio drafts; you publish. Your voice, your brand, your call.
• Local by default. Matariki, EOFY, Auckland winter, kiwi tone — Marlaio doesn’t write “summer sale” in June.

If you’re a café, tradie, boutique, or any small operator in NZ or AU, the version of you running this same routine is who Marlaio is for. We’re building it one customer at a time, and if you want to be one of those customers, marlaio.com is where to start.