How to get more customers on slow days at a small shop

The fastest way to get more customers on slow days is to quit promoting to "slow days" as a category and aim at one window. Tuesday, 2 to 5pm. Take one offer to the people who have already paid you once, then judge it by a single number four weeks later. That's the method. Everything below is how to make each decision without spending a month learning it the hard way.

Most owners get the offer wrong. The reflex is a percentage off everything, which is the most expensive way there is to buy a customer who was coming anyway. Cheaper shapes exist. An add-on that costs you forty cents and reads as three dollars of value. A bundle priced so the average ticket goes up while the customer still walks out feeling like they got something. I'll run one example the whole way through: Marlow's, a fourteen-seat sandwich counter that does most of its money before 1:30.

Prove the window is dead before you promote to it

Owners guess wrong about their own slow hours constantly. I once sat with a guy who was certain Mondays were killing him. His POS said Mondays were ordinary and Wednesday afternoons were the wound. Pull the last ten Tuesdays by hour and lay them side by side before you spend a dollar.

At Marlow's, 11am to 1pm runs 84 tickets. The 2 to 5pm block runs 11 tickets at an average of $9.40. Call it $104 of revenue across three hours with two people on the clock at $18 an hour. That's the hole. It also isn't all of "afternoons," because business picks back up around 4:30 with school pickup, so the dead stretch is really 2:00 to 4:15.

Then check one more thing. Has the window been slow every Tuesday for a full quarter, or only the last six weeks? A six-week dip is usually weather, or a grand opening down the block borrowing your customers for a while, and it tends to fix itself. A quarter of flat Tuesdays is structural. Structural is what a campaign is for.

Give the campaign exactly one job

Write the job as one sentence with a number in it. "Get Tuesday 2 to 5pm from 11 tickets to 25 by September 30." That's the brief. Awareness is not a job. Neither is filling the afternoon while also growing the email list, because a campaign carrying two jobs sells nothing; every decision downstream gets settled by splitting the difference, and split differences don't move tickets.

Keep the sentence somewhere you'll see it. It settles arguments later. When the deals-site rep calls with a citywide coupon he'll blast to 40,000 people, the sentence answers for you before you've finished being flattered.

Shape an offer that protects your margin

The offer has to be sayable in one sentence. If it needs a paragraph of conditions, cut conditions until it fits. "Dine-in only, excludes catering, while supplies last" is three exceptions nobody reads and your weekend counter kid gets wrong in front of a line of eight people.

It also needs a day or an end date attached. "Come see us sometime" is a pleasant thought. "Every Tuesday through September 30" goes in someone's week.

And it should add value rather than cut price. Here's the math at Marlow's. Food cost runs about 31%, so a $9.40 ticket leaves roughly $6.50 of margin. Twenty percent off hands back $1.88 of that, nearly a third, and hands it to the regular who was coming Tuesday no matter what. Now the other shape: any sandwich after 2pm comes with a drip coffee or a house iced tea. The coffee costs 38 cents to pour and reads as $3.25 on the board. Margin goes $6.50 to $6.12, the sandwich price never moved, and come October you aren't un-training anyone from a lower number.

Bundles work from the other direction. "Tuesday afternoon board: half sandwich, cup of soup, cookie, $12" prices out at $14.75 à la carte, and soup is the cheapest thing Marlow's makes. The ticket climbs from $9.40 and nothing on the regular menu got repriced. I like the bundle better than the free coffee, honestly, though it's harder to say fast across a counter, which is the trade.

Then redemption. It has to be a word said out loud. "Just say early bird." The moment there's an app to download or a coupon to print, redemptions fall, and the people who fall off first are the marginal customers the whole thing was built for.

Use a channel you already own

Customers who have paid you before are the cheapest audience you will ever reach, and that list is already sitting in your email tool, your text list, or the signup line on your POS receipts. Social comes after that, then signage. Paid ads go last, because $200 of Instagram spend will teach you something and fill nothing.

Pick one channel to carry the message and one to echo it. Marlow's has 1,140 email addresses and opens run around 38%, so a Sunday send puts the offer in front of roughly 430 people who have eaten there. The echo is a chalkboard line by the register on Friday and again Monday, so the regular who deleted the email hears it anyway from whoever hands over the sandwich.

Don't pick a channel you can't maintain. If your last Instagram post is seven months old, this is a bad week to become an Instagram person.

Write it the way you'd say it across the counter

Lead with what the customer gets. Nobody has ever once been excited by a business's news, so "we're excited to announce" shouldn't survive your first draft.

Subject lines I'd test: "Tuesday afternoon, your coffee's on us" and "Free coffee with any sandwich, 2 to 5 on Tuesdays." A third one, "The quiet hours at Marlow's," is the one I'd bet on losing, which is exactly why it's worth running against the other two.

The body, about 70 words:

"Tuesdays get quiet here after two, so we're doing something about it. Any sandwich between 2 and 5 comes with a drip coffee or a house iced tea, on us, every Tuesday through September 30. There's nothing to print and no code to type. Just say 'early bird' when you order. The window table is almost always free at 3."

Then hand your staff one line for the register: "If you're around on a Tuesday afternoon, sandwiches come with a free coffee until 5." One thing to do, said once, in words a person would use out loud.

Decide what "it worked" means before you launch

Name the number now. For Marlow's it's tickets rung between 2 and 5 on Tuesdays. Baseline 11, target 25. Attribution can be a tally sheet taped by the espresso machine or a $0 modifier in the POS, so counting "early bird" mentions costs nobody any thought at the end of a shift.

Set the date you'll call it: four Tuesdays in, September 30. If the last two average 20 or better, run it again in January. If you're sitting at 13, look at reach before you blame the offer. 430 opens should have produced more than two extra sandwiches, and when they don't, it's usually because the email went out Sunday at 9pm or the subject line buried the actual offer under a paragraph about the new patio furniture.

One thing people forget to check: make sure the gain wasn't stolen from noon. If 2 to 5 climbs to 25 and 11 to 1 drops by 9, you rearranged your customers and added none. I've watched an owner celebrate that result for a week before anyone pulled the noon numbers.

Then write down what you ran. Offer, dates, channel, the number you hit. Six months from now you won't remember whether the free coffee beat the soup bundle, and you'll spend next winter arguing with yourself about it.

Scaffle is an AI advisor for businesses. It learns how yours runs and keeps an eye on the tools you connect, so last quarter's Tuesday numbers and the campaign you ran in September are already in the room when you sit down to pick what to try next. It's at scaffle.ai.