Discounts vs added-value offers for restaurants: protecting your margin

A percentage discount pays back when it fills covers you would otherwise lose, such as a quiet Tuesday. The extra business must also outrun the margin you give away. To bring back guests who already like you, a free extra usually wins. A $7 dessert costing $2 to plate rewards a return visit without cutting your menu price.

Should your restaurant offer a discount or added value?

Choose by the seat you need to fill and the guest you are sending to. A discount suits empty tables and diners who have stopped coming. A free extra or an experience suits regulars, birthdays and anyone you want back without teaching them to wait for a deal.

The difference sits in what each offer takes out of the check. Take a party of two spending $60. A 20% discount removes $12, and every dollar of it comes off your profit. The food, drink and staff time cost the same whether the guest pays full price or not.

A free dessert with a $7 menu price might cost $2 in ingredients. The guest sees $7 of value, and you pay $2. The discount is worth more to the guest, but it costs you six times as much.

That gap is why added value protects margin. It does not make discounts wrong. An empty seat on a Tuesday earns nothing, so a discounted cover that fills it can still add profit. The expensive mistake is sending the same discount to everyone, including regulars who would have booked at full price.

One term runs through this guide. Contribution margin is what remains from each dollar of sales after the costs that rise with every cover: food, drink, card fees and packaging. Fixed costs such as rent sit outside it. Take your own figure from your costings. The examples below assume 65%.

Where do discounts and added-value offers differ?

Both offers can bring a guest back. They differ in what they cost, who they attract and what they do to your prices over time.

Factor	Percentage discount	Added-value offer
Cost on a $60 check for two	$12 at 20% off, all from profit	About $2 food cost for a $7 dessert
Value the guest sees	$12, a clear saving	$7, a treat rather than a saving
Who it attracts	Price-sensitive diners and deal-seekers	Diners who already value your food and service
Cost when regulars redeem it	Full discount on visits that would happen anyway	Food cost of the extra only
Effect on menu prices	Repeated discounts lower the price guests expect	Menu prices stay intact
Kitchen and floor impact	POS change only	One extra plate or pour, staff need briefing
Best use	Empty seats and lapsed guests	Regulars, birthdays, new menu items
How to measure it	Extra covers against a holdout group	Return visits against a holdout group

Discounts reset the price guests expect

Guests remember what they paid last time. Run a discount every month and part of your list learns to wait for the next one. Your menu price becomes the price nobody pays. Added value leaves that price untouched, because the guest still pays for the meal in full.

Discounts reward visits you already had

Regulars already plan to come back. Send them 20% off and they will use it, often on a visit they were going to make anyway. That is margin given away for no new trade. A free extra still goes to those guests, but it costs the food cost of a dessert, not a fifth of the bill.

When does a percentage discount pay back for a restaurant?

A discount pays back when the extra covers it brings earn more than the margin it gives away. You can work out the tipping point before you send anything.

The break-even sum

Divide the discount by your contribution margin minus the discount. The answer is the rise in covers you need just to stand still. At a 65% margin, a 20% discount gives 20 ÷ 45, so you need 44% more covers. A seating of 100 covers must become 144 to break even.

The sum assumes every guest in that seating uses the offer. That is the worst case, and the usual case for an open discount posted on social media.

Discount	Extra covers needed at 65% contribution margin	Extra covers needed at 70% contribution margin
10%	18%	17%
15%	30%	27%
20%	44%	40%
25%	63%	56%
30%	86%	75%

The curve steepens fast. A 30% offer needs nearly twice the extra trade of a 20% one at a 65% margin. Every extra five points of discount demands a much larger lift in covers.

Conditions where a discount earns its keep

• The seat would otherwise be empty. Early-week evenings, early seatings, and January are the usual candidates. Empty seats earn nothing, so a discount there costs margin only on covers you would not have had.
• Only lapsed guests can redeem it. If only new covers use the offer, any discount below your contribution margin adds profit on each one.
• It is single-use, named, and time-boxed. A code tied to one guest and one two-week period caps your exposure.
• A minimum spend or prix fixe menu protects the bill. "20% off a three-course prix fixe menu" limits the discount to dishes you have costed.
• The guest needs a clear reason to return. Someone who has not visited in three months may ignore a free coffee but notice a fifth off.

Worked scenario: filling Tuesday at a 90-cover neighborhood restaurant

The figures below are a worked model, not a reported result. Swap in your own.

Situation. An independent Italian restaurant seats 90. Tuesday evenings average 30 covers at $30 a head, with a 65% contribution margin. Tuesday earns $585 in contribution (30 × $30 × 65%).

What was done. The owner modeled two versions of a 20% Tuesday offer.

• Open offer. 20% off for everyone on Tuesdays, promoted on social media and to the full list. Covers rise to 48, and all 48 pay the discounted price. Revenue is $1,152 (48 × $24) and variable costs are $504 (48 × $10.50). Contribution is $648, a gain of $63.
• Targeted offer. A single-use code sent only to guests whose last WiFi login was more than 60 days ago. Covers again rise to 48, but the 30 regulars pay full price and earn $585. The 18 returning guests earn $243 (18 × $13.50). Contribution is $828, a gain of $243.

Outcome. The same 18 extra covers produced nearly four times the profit when the offer was targeted. Across 13 Tuesdays in a quarter, that is $3,159 against $819. The discount was not the problem. Who received it was.

When does a free extra drive the same return visit for less?

A free extra works when the guest already wants to come back and needs a nudge, not a price cut. It costs you its food cost, while the guest values it at menu price.

Conditions where added value wins

• The guest is a regular or a recent visitor. They do not need persuading on price. They need a reason to book this week rather than next month.
• The extra has a high menu price and a low food cost. Desserts, coffee, a glass of house wine, bread and olives or a drink upgrade are common choices. Cost each one from your own recipe cards before you choose.
• An occasion brings a group. A birthday dessert for one guest brings a table of four who pay full price.
• You are launching a dish. A free taster puts a new plate in front of guests who may order it at full price next time.
• Your brand depends on price integrity. Fine dining and premium venues rarely want their name next to "25% off".

Experiences often cost less than they look

Experiences are the other form of added value. Think of a chef's table seat, a tasting of the new seasonal menu or a cocktail masterclass on a quiet Sunday afternoon. Their main cost is staff time in hours when your team is already on the schedule, plus ingredients.

Some cost almost nothing. Early booking access for a new menu, or first refusal on Christmas tables, uses inventory you were going to sell anyway. For a loyal guest, being asked first can matter more than a saving.

Worked scenario: a 12-site bar and grill group in January

Again, the figures are a worked model, not a reported result.

Situation. A 12-site bar and grill group holds 9,000 consented regulars, defined as three or more WiFi logins in 90 days. It wants a January return visit. A party of two spends $55 on average, at a 65% contribution margin. The group weighed 15% off the bill against one free dessert per table, with a $6.50 menu price and $1.80 food cost.

What was done. The list was split at random into three groups of 3,000: discount, dessert and a holdout that received nothing. Return visits were measured by WiFi logins within 30 days. Redemptions were counted from single-use codes at the register.

Modeled results. The holdout returned at 20%, the discount group at 26% and the dessert group at 24%.

• Discount. 780 redemptions at $8.25 each cost $6,435. Extra visits over the holdout were 180, so each extra visit cost $35.75.
• Dessert. 720 redemptions at $1.80 each cost $1,296. Extra visits were 120, so each extra visit cost $10.80.

Outcome. Each extra visit is worth $35.75 in contribution at full price ($55 × 65%). The discount's 180 extra visits earned $6,435, exactly the amount given away, so the offer broke even. The dessert's 120 extra visits earned $4,290 against $1,296 in desserts, a gain of $2,994. The discount brought more guests back, but the dessert made the money.

What does each offer cost to run?

Count the cost of every offer per extra visit, not per redemption. Five costs decide the answer.

Margin given away on each redemption

This is the visible cost. A discount gives away its full value from profit, and an extra costs its food or beverage cost. On a $60 bill for two, 10% off costs $6, 20% off costs $12 and a $7 dessert costs about $2.

Paying for visits that would have happened anyway

This is the largest hidden cost, and it is why the open offer above barely moved profit. Every regular who redeems a discount on a planned visit is a pure loss. Only a holdout group shows you the size of it. Without one, every redemption looks like a win.

Leakage and code sharing

Open codes spread to deal sites and group chats, and reach diners who were never on your list. Named, single-use codes limit leakage for both offer types. They also make every redemption attributable to one send.

Kitchen and front-of-house load

A discount is a register setting. A free extra means an extra plate or pour, a briefing for the floor team and a clear rule on what is included. Experiences need a host and a date. Budget the staff time, especially on busy sites.

Long-term effect on price

A one-off discount to lapsed guests does little damage. A monthly discount to the whole list trains guests to wait. Added value carries little of that risk because the guest still pays your menu price.

How do you prove which offer brought guests back?

Measure return visits against a group that received nothing. Without a holdout, you cannot separate the visits an offer created from the ones it merely rewarded.

Build the list from consented WiFi logins

Purple Engage builds your list from guest WiFi logins. A diner connects, makes a conscious choice about marketing, and becomes a consented, first-party contact. Each later login adds a visit to that guest's record.

That visit history does two jobs. Before you send, it separates regulars from lapsed guests, so the discount goes only where it creates new trade. After you send, a later login shows who came back.

Mailchimp and Klaviyo send email well. Neither captures consent from diners sitting at your tables, so you supply contacts and visit data from somewhere else. Purple runs at 80,000+ live venues and recorded 440 million logins in 2024, according to Purple's own data. Hospitality brands including Pizza Express, Stonegate Pubs and Sweetgreen use Purple.

Keep consent clean under CCPA/CPRA

CCPA/CPRA defines consent as freely given, specific, informed and unambiguous. In the US, the TCPA and CAN-SPAM also govern marketing email, as the FTC and state attorneys general guidance sets out.

In practice, give guests WiFi either way and ask for marketing consent as a separate choice. That is how Purple's conscious-choice opt-ins work. Every offer email should carry a working unsubscribe link.

Set up the test

1. Define the segment from visit history. For example, regulars with three or more visits in 90 days, or guests lapsed for 60 days or more.
2. Split at random and hold back 10-20%. The holdout receives no offer.
3. Issue a single-use code per guest. Redemption at the register then ties revenue to the send.
4. Fix the window. Thirty days suits most restaurant offers.
5. Compare cost per extra visit. Divide total redemption cost by visits above the holdout rate.

For redemption mechanics and ROI reporting, see Digital vouchers and offers for pubs: redemption tracking and ROI. To check whether added value lifts satisfaction as well as visits, pair your test with NPS surveys for multi-site restaurant groups: comparing loyalty across sites.

How should you choose an offer for your venues?

Match the offer to the guest and the sitting. This matrix covers the situations most restaurant and bar groups face.

Your situation	Recommended offer	Why
Seats empty on Monday to Wednesday evenings	Time-boxed 15-20% discount, lapsed guests only	Empty seats earn nothing, so new covers add profit
Regulars with three or more visits a quarter	Free dessert, coffee or drink upgrade	They come anyway, so protect the bill
Guest birthday or anniversary	Free dessert or glass of fizz for the guest	Brings a table that pays full price
Guests lapsed for 90 days or more	Single-use discount with a set expiry	A clear saving gives a reason to return
Launching a new dish or menu	Free taster for recent visitors	Sampling builds future full-price orders
Friday and Saturday peak	Early booking access, no price offer	Tables already sell, so any discount is lost margin
Fine dining or premium brand	Chef's table or tasting evening invitation	Rewards loyalty without cutting price
Surplus of a perishable ingredient	Free side or extra featuring that ingredient	Turns likely waste into perceived value

Start with one test, not a new loyalty program. Pick your largest segment, run a discount against an added-value offer and a holdout for 30 days, and compare cost per extra visit. Most groups find the answer differs by segment. That is the point: the choice is per guest, not per brand.

Frequently asked questions

Is a discount or a free extra better for bringing diners back?

A free extra is usually better for diners who already visit, because it costs only its food cost while the guest values it at menu price. A discount is better for filling empty seats and winning back lapsed guests, where most covers it brings are new trade. Test both against a holdout group. Then compare the cost of each extra visit rather than the redemption count.

How much extra trade does a 20% discount need to break even?

At a 65% contribution margin, a 20% discount needs about 44% more covers to break even if every guest in that sitting uses it. Divide the discount by your contribution margin minus the discount. If only new or lapsed guests can redeem it, through a single-use code, any discount below your contribution margin adds profit on each extra cover.

Can I send restaurant offers to guests who joined my WiFi under CCPA/CPRA?

Yes, if the guest gave marketing consent as a separate, informed choice. Privacy regulations require consent to be freely given, specific, informed and unambiguous. Laws warn against making a service conditional on it. In the US, the TCPA and CAN-SPAM also govern marketing messaging. Purple uses conscious-choice opt-ins: guests get WiFi either way and choose marketing separately.

How does Purple Engage compare with Mailchimp or Klaviyo for restaurant offers?

Purple Engage builds the list for you, while Mailchimp and Klaviyo send to a list you already have. Each guest WiFi login adds a consented, first-party contact and records the visit. You can then target lapsed guests or regulars and see who returned after an offer. With an email-only platform, you need another source for contacts and visit history.

Does Purple work with the WiFi hardware my venues already have?

Yes. Purple is hardware-agnostic and runs as a cloud overlay on Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet. You keep your existing access points and add Purple on top. That lets you start building a consented guest list across your locations without replacing equipment first.

How do I prove an offer created a return visit?

Hold back a random 10 - 20% of the segment and send them nothing. Compare the return-visit rate of each offer group with that holdout over the same window, such as 30 days. The difference is the visits the offer created. Divide the total cost of redemptions by those extra visits to get the cost per incremental visit.

Should I send the same offer to every guest on my list?

No. A single offer to the whole list pays regulars for visits they would have made anyway. Segment by visit history first. Send discounts to lapsed guests and to seatings with empty seats. Send free extras or experiences to regulars and birthday guests. In the worked Tuesday model above, targeting alone turned a $63 nightly gain into $243.