How to Get More Customers on Slow Restaurant Nights: Fill Monday Through Wednesday Without Discounting Every Table

Pick one underfilled Monday-through-Wednesday daypart, set an incremental-cover target your kitchen and floor can serve, and give nearby guests one trackable, brand-fitting reason to visit inside that window. Judge the campaign on incremental completed covers and contribution from settled checks—not responses, claims, reservations, or gross sales.

That is the narrow version of a restaurant marketing plan: one window, one reason, one action, and one decision. It lets you see whether the target window gained enough profitable traffic to repeat the campaign instead of moving a regular from Friday to Wednesday and paying for the privilege.

“We have zero problems filling in seats on Friday and Saturday. We do have a problem filling seats during the weekday.”

Source: An all-you-can-eat restaurant operator

Write this one-page campaign brief before making an ad or sending a message:

1. Choose one window where you can actually serve more covers

Choose the daypart from completed covers and settled checks for the same weekday and hours, not from a feeling that “Tuesdays are slow.” If the whole restaurant has slowed, find the traffic, conversion, check, or repeat-visit leak first; a one-night campaign should not hide a broader operating problem.

Pull comparable POS and reservation records, then write down:

1. the day and exact hours;
2. completed covers in that window;
3. settled checks and settled sales under one POS definition;
4. average covers per check and average settled check;
5. empty seats on a normal service;
6. the most covers the kitchen, bar, and floor can add without hurting service; and
7. the incremental-cover target for this campaign.

Use the smallest operational limit for the target. Forty empty seats do not create room for forty extra covers if the kitchen can only handle twelve.

Compare like with like. A Feast measurement review found covers down while spend per person rose and called for year-over-year baselines, holiday adjustments, and event logs before giving marketing credit. If lunch is the window, keep the offer and scorecard there; the lunch-sales workflow covers that daypart.

2. Make the first check work before choosing the promotion

Calculate conservative contribution per settled check before promising a discount, free item, or experience. Start with settled food-and-beverage sales under the POS definition you will use after launch, then subtract the costs that rise when that check happens:

Check contribution = net settled sales after discounts and refunds − sold-item food and beverage cost − incremental labor − payment-processing, packaging, fulfillment, and incentive costs.

Use one incentive treatment: actual incremental food, packaging, and fulfillment cost, not face value. If those dollars already appear elsewhere, label them memo-only; count them once.

Keep campaign costs separate: media; production; applicable food-influencer compensation and usage rights; hosted meals at actual incremental food, packaging, paid labor, and fulfillment cost; software; agency; management. Count a hosted meal once, never again inside influencer compensation or check costs. Allocate shared charges by a written basis that reconciles to the full charge; do not repeat bundled components.

Now choose one reason to visit. It might be a limited menu item, a midweek experience, early access, a group occasion, a value bundle, or a carefully bounded offer. Pick the one that shows why your restaurant is worth the trip, not simply what you are willing to mark down.

“Because the experience is going to attract the user, not the offer.”

Source: A restaurant owner discussing a slow-night event

That does not mean an offer is always wrong. It means the offer should support the reason rather than replace it. If brand fit is the concern, use the substance-first approach to restaurant promotions: show the food or experience first and keep the incentive bounded.

The output of this step is one written reason to visit and a conservative check-contribution number it can support.

3. Put rules around the window so the promotion cannot eat the whole week

Make the promotion redeemable only under rules your guest, host, server, and manager can all repeat. Write the rules before launch, then place the same version on the campaign page, reservation note, staff sheet, and POS setup.

State:

• the eligible day and hours;
• whether booking, claiming, or arriving before a cutoff is required;
• eligible party sizes and menu items;
• a minimum spend only when it fits the experience and check math;
• whether the offer can stack with other discounts or rewards;
• how many times one guest or party can redeem;
• the honest expiration date;
• the maximum number of claims, reservations, or redemptions the restaurant will accept; and
• what happens when a guest arrives outside the rules.

Protect capacity as well as price. Stop taking reservations or claims before the target window exceeds what the team can serve. If the same item or deal is available on Friday, existing Friday guests can shift nights without creating net-new covers; make the reason meaningfully tied to the slow window.

The output of this step is an approved rule sheet and a capacity limit. If the manager on duty cannot explain both without improvising, the campaign is not ready.

4. Use one channel and ask for one trackable action

Choose one primary channel that can reach nearby guests and one action you can identify later. That might be an existing email or text list, a local paid-social campaign, or the restaurant's organic audience; no channel is universally best.

If you use email or text, apply your permission and suppression process before selecting recipients; an existing record is not a send decision. Track unique eligible, suppressed-before-send, attempted, sent or provider-accepted, delivered, failed, pending or unknown, and opted-out people separately. Delivery does not prove a read.

Message-action rate = the subset of unique eligible delivered recipients in the denominator who acted inside the fixed window ÷ unique eligible delivered recipients whose full window elapsed and whose action outcome was reconciled and known. Report open, immature, ineligible, and unresolved records separately.

Ask the guest to do one thing: reserve through a campaign-specific path, claim the visit with a permitted identifier, or reply with a clear keyword. Do not split the first test among several offers and several actions. You need one path you can diagnose from message to check.

Food influencer content is optional. Use it only when seeing the item or experience makes the reason easier to understand and you have the permission and workflow to use the footage. A promotion does not need an influencer merely because it runs on a slow night. If you need more possible concepts before choosing one, rank restaurant marketing ideas by the result you can track, then return to one campaign.

The output of this step is one approved message draft, one selected primary channel, and one identifiable action. Do not publish or send the message until the full response-to-check path passes the acceptance test in the next step.

5. Carry the response through arrival and the settled check

Build and test the response-to-check path before the campaign message goes live. A claim or reservation is interest; the restaurant earns a completed visit only when the party arrives, and it earns a settled check only when payment is complete.

Use this chain:

1. Response: the guest completes the chosen action.
2. Identification: the campaign saves the permitted identifier needed for follow-up and matching.
3. Claim or reservation: the guest receives the eligible date, hours, rules, and arrival instructions.
4. Reminder: the guest gets the time, location, cancellation path, and what to tell the host.
5. Arrival: the host finds the reservation or claim and records whether the party arrived, canceled, rescheduled, or did not show.
6. Checkout: the server applies the correct item or discount under the rule sheet and attaches the campaign code or guest record to the POS check.
7. Settlement: the paid check closes under the same sales definition used for the baseline.

Give the shift one plain instruction: “Confirm the eligible window and rules, attach the campaign record before tender, and record the final visit outcome.” If the guest pays cash or another person pays, the campaign identifier may not match automatically; use the approved POS code or record that check as unmatched rather than inventing a connection.

Before sending the campaign, test the response, reservation or claim, arrival, and POS attachment. In test mode or with a controlled refundable transaction, settle a paid, non-voided check and confirm the campaign, amount, and settled-sales field in reporting. Then void or refund it and confirm exclusion. Send only after the host can find the test guest, the server can apply the rule, and the manager can verify both inclusion and cleanup without a spreadsheet.

6. Treat seven-day expiration and extension as one test design

Use a seven-day expire-then-extend sequence only as a prewritten test, not as a universal deadline. One recorded campaign comparison found a seven-day expiration produced a higher redemption rate than a 14-day expiration. A separate account note described reminding guests before expiration and offering an extension the next day; that sequence helped drive more redemptions in that case.

To test it honestly:

1. give the first group a real seven-day redemption window;
2. send a factual reminder before it closes;
3. close that window as stated;
4. invite eligible non-visitors the next day with a clearly stated new expiration; and
5. label original-window and extension redemptions separately.

Do not call the first message a “final chance” when an extension is already planned. A different restaurant may need a different duration because of booking lead time, service schedule, or how often the audience reads messages. The output is not a rule for all future promotions; it is a clean comparison between the original window and the extension.

7. Score completed covers, checks, contribution, and displaced traffic separately

Use separate counts and denominators so a full reservation list cannot masquerade as a profitable night. Use eligible claim or reservation records as the single unit for visit conversion; keep unique people, covers, and checks as separate counts. Freeze the definitions before launch and keep pending or future records out of outcome rates.

Matched settled sales show a recorded connection between the campaign guest and a check. They are not profit, and they do not prove the campaign caused the sale.

Measure the target window against its like-for-like baseline, then check nearby windows for displacement:

Target-window observed cover lift = completed covers in the campaign window − baseline covers for that same window.

Net observed cover change = target-window observed lift − documented cover losses in nearby windows the campaign may have displaced.

Both are observed comparisons, not causal proof. Note holidays, events, closures, staffing changes, and other obvious differences beside the numbers. Convert check contribution to a conservative per-cover figure using average covers per settled check, then estimate:

Conservative campaign contribution estimate = net observed cover change × conservative contribution per cover − all campaign costs in the marketing ledger.

If net observed cover change is zero or negative, do not rescue the campaign with attributed gross sales. If the comparison is too messy to support an incremental estimate, report matched sales, matched check contribution, campaign costs, and the uncertainty separately.

8. Continue, repair, or stop from the first broken step

Continue only when the full path is working and the conservative campaign contribution estimate clears the minimum you wrote before launch. More responses are not a reason to continue if completed covers, matched checks, or contribution are still missing.

• Continue: the decision window has matured, tracking is intact, net observed cover change and conservative contribution meet the written threshold, and the restaurant still has service capacity.
• Repair: find the first weak handoff. Fix message delivery before the response, response before the reservation or claim, reminder and rules before attendance, staff execution before check matching, and check economics before buying more reach.
• Stop: end the campaign when it reaches the spending cap or decision date with negative conservative contribution, unresolved tracking, harmful service pressure, or documented movement from another night without net gain.

Keep the original brief, rule sheet, staff instructions, exported outcomes, POS definition, and final decision together. That turns a slow-night promotion into a reusable test instead of a story about whether last Wednesday “felt busier.”

Feast can connect the campaign response to the settled check

Feast becomes useful when the manual handoffs are the problem: its campaign pages can capture a guest's phone number, follow-up SMS can bring the guest back, and a payment using the same phone identity can be matched to the campaign in the POS. Feast reports POS-linked campaign sales; that is not profit or proof of causal lift, so the contribution and baseline work above still applies.

Book a demo of Feast to build a slow-night campaign with guest follow-up and POS-linked sales reporting.