Delivery App Promotions: Which Ones Convert and Which Ones Burn Your Margin
Introduction
If you have ever launched a "20% off everything" on Rappi or PedidosYa and then looked at your numbers at the end of the month wondering where the money went, that is not a coincidence. Promotions on delivery apps are the most misused tool in the LATAM restaurant playbook.
The problem is not running promotions. The problem is running promotions without understanding what changes the algorithm and what changes the customer — and confusing order volume with profitability.
In this article you will see which promotions genuinely convert, which ones only generate volume without margin, and how to build a promotion strategy that adds money rather than giving it away.
👉 If you want to check your numbers before launching any promotion, request your free Growth Delivery App audit.
---
Why Most Delivery Promotions Fail
Before talking about what works, let us understand why what most restaurants are already doing does not.
The typical restaurant logic is: "if I lower prices, I get into more visibility packages and get more orders." The reasoning is not entirely wrong, but it ignores three critical variables:
1. The algorithm does not reward promotions — it rewards conversions. Rappi and PedidosYa measure how many users visit your profile and of those, how many place an order. A promotion that attracts curious browsers with no order history does not improve your ranking. A promotion that activates a customer who was on the fence does.
2. The promotional customer is not the same as your regular customer. When you run "20% off everything," you attract a segment that only orders when there is a discount. When the discount ends, that segment disappears. Your base of recurring customers (those who come back without a promotion) has not grown at all.
3. The margin burns before you see any ROI. A 20% promotion on an item with a 18% net margin is already in the red before you count the app's commission. You would need at least 3× your usual order volume to break even. That rarely happens.
---
The 5 Promotions That Actually Convert in Delivery
1. Free Delivery with a Minimum Order Value
How it works: the customer pays no delivery fee if their order exceeds a certain amount (typically 10% to 20% above your current average order value).
Why it converts: customers perceive the delivery cost as a concrete loss. Removing that friction activates orders that were dropping off at checkout. The app shows it as a highlighted benefit in the grid, which also improves your CTR.
How to calculate it:
If your average order value is USD 12, set the threshold at USD 14.
The customer adds an extra item to reach the minimum.
Result: higher order value, the same or proportionally lower commission.
When to use it: as a permanent promotion during low-demand hours (Monday, Tuesday, mornings). Do not limit it to the weekend — that only acts during the hours you are already selling well.
---
2. 2-for-1 on a High-Margin, Low-Food-Cost Item
This is the most efficient promotion for restaurants with products that have a low ingredient cost (drinks, desserts, sides, premium sauces, garlic bread, etc.).
The maths:
A dessert with a USD 2 food cost, selling at USD 7.
In a 2-for-1, you sell two for USD 7. Gross margin: USD 3 (versus USD 5 individually).
But the order includes an extra item that in 70% of cases was not the reason for the original order. You are upselling without the customer noticing it as an additional expense.
What this is not: a 2-for-1 on your signature dish with tight margins. That will hurt you.
Why it converts: the perception of a gift is very powerful. "2-for-1" is visible in the grid before the customer even enters your profile. It typically lifts CTR by 15% to 30%, which directly impacts the algorithm's ranking signal.
---
3. "Express Lunch" Combo During a Defined Time Slot
What it is: a special combo available only within a 2 to 3 hour window at lunchtime (typically 12pm to 3pm or 1pm to 4pm), at a fixed price slightly lower than the regular item prices combined but above the individual food cost.
Example:
Sandwich + drink + dessert at USD 15 (individual prices add up to USD 19).
Margin is trimmed, but: preparation speed is maximised because it is always the same combination; order volume in that time slot typically rises 20% to 40%.
Why it works better than a blanket discount:
It is scarce (only in that time window). Scarcity activates purchase decisions.
It standardises the operation: the kitchen knows exactly what to prepare, times fall, rating improves.
The app shows it as featured content during that time slot — you occupy more visual space in the grid.
---
4. Reactivation Promotion for Lapsed Customers (Via App or Platform)
Rappi and PedidosYa have native tools to send discounts to customers who ordered from you at least once in the past 90 days but have not ordered in more than 30 days.
How it works: the platform filters the segment and shows a push notification or personalised discount to that user. You pay the discount cost; the platform provides the channel.
Why it is different from a general discount: you are speaking to someone who has already tried your product, already knows who you are, and simply forgot about you. A reactivation discount costs 8% to 15% less than acquiring a brand-new customer.
When to use it: once a month, with a 10% to 15% discount on the first returning order — not across the entire catalogue, but on one of your top 3 items.
---
5. Star Product with a Visibility Boost (Surgical Sponsorship)
This is not exactly a "promotion" in the discounting sense, but it is the most profitable use of platform visibility investment.
What it is: instead of paying for a general banner, you pay for a specific item — your highest-margin, best-rated product — to appear as featured in its category or in relevant searches.
Why it works better than sponsoring your entire profile: a well-presented item with good photos and reviews in a featured position converts between 2× and 4× more than a profile without any menu work done. Visibility without conversion is wasted money.
Condition for it to work: the sponsored item must have at least a 4.5 rating and a professional photo. Otherwise the additional traffic does not convert and you are paying for eyeballs without orders.
---
The 3 Promotions That Burn Margin Without Building Anything
"20% Off the Entire Menu"
It is the most used promotion and the most destructive to profitability. Why:
It subsidises purchases that would have happened anyway (recurring customers who would have ordered without a discount).
It does not differentiate by margin: you give 20% off your highest-margin dish just as readily as your lowest.
The perceived value of your brand drops: the customer learns that "you always have a discount" and starts waiting before ordering.
It does not improve your ranking because it is not a quality signal — it is just a price signal.
---
Free Delivery Without a Minimum Order Threshold
Offering free delivery with no conditions means absorbing the logistics cost of the platform with your margin. If the typical delivery cost is USD 3 to USD 5 per order and your average order value is USD 10, you are surrendering 30% to 50% of margin before paying the app's commission.
What does work: free delivery from a minimum order value. The threshold becomes an upselling engine.
---
Long-Running Promotions With No Visible End Date
A promotion running for more than 30 days ceases to be a promotion and becomes your listed price. The customer loses any sense of urgency and the perception of getting a deal.
Furthermore, the platform tends to give less visibility to promotions that have been running for many weeks, because the algorithm detects they are no longer generating new incremental orders.
Practical rule: no promotion should run for more than 3 consecutive weeks in the same format. Rotate it or restructure it.
---
How to Build Your Monthly Promotions Calendar
A promotions calendar is not "launch something when sales drop." It is a structured strategy with three variables: objective, segment, and channel.
This rhythm gives you four distinct platform signals per month without cross-contaminating discounts. Each week has only one active promotion so they do not cannibalise each other.
---
How to Measure Whether a Promotion Actually Worked (Beyond Order Volume)
The number that should matter to you is not how many orders you received during the promotion. It is how many of those customers came back to order again — without a promotion — in the following 30 days.
Key metrics:
Post-promotion retention rate: how many customers from the promotion ordered a second time?
Average order value during vs. outside the promotion: if the order value falls, the promotion is not lifting orders — it is only changing the price.
Rating variation during the promotion: if volume spikes suddenly and the kitchen cannot keep up, your rating will drop and cancel out the ranking benefit.
Net margin per order during the promotion: gross promotion sales minus food cost, minus platform commission, minus the discount cost.
If the net margin during the promotion is more than 20% below your normal net margin, the promotion was not profitable — even if you had "a great weekend" in terms of order count.
👉 If you want to see these numbers without calculating them manually, audit your account for free with Growth Delivery App. We show you which promotions are costing you more than they are generating.
---
What the Algorithm Says About Promotions
Neither Rappi nor PedidosYa publish their algorithms, but behaviour observed across many audited accounts shows clear patterns:
Promotions do not replace a good rating. A restaurant with a 4.2 rating and a 30% discount loses to one with a 4.7 and no discount in the organic grid.
Conversion is the king KPI. If your promotion lifts visits but not orders, the algorithm interprets it as your offer not being attractive and may lower your position.
Consistency beats intensity. Three weeks of small, well-calibrated promotions have more impact on ranking than one weekend of massive discounts.
Platform campaign promotions add a signal. Participating in the platform's own campaigns (sale days, Valentine's Day, Father's Day, etc.) gives you additional free visibility. It is worth evaluating item by item what to include — not your entire catalogue.
---
Frequently Asked Questions (FAQ)
How much of a discount do I need to offer for a promotion to be effective on Rappi?
There is no universal number, but the practical reference is: at least 10% for the customer to perceive a real benefit, and no more than 20% before it stops being profitable for most categories. The sweet spot for conversion without margin destruction is between 12% and 18%, applied to specific items with sufficient margin.
Is it worth joining Rappi and PedidosYa discount campaigns?
Yes, with selectivity. Platform campaigns (like delivery weeks, seasonal deals) provide free visibility and real incremental traffic. The key is not to enter with your entire catalogue: choose 2 or 3 items with high margin and optimised photos, and keep the rest out of the campaign.
How long does a promotion need to run for the algorithm to detect it?
The algorithm needs at least 5 to 7 days of consistent data for a promotion to have an impact on ranking. Promotions lasting only 2 to 3 days have short-term volume effects only. If you want organic visibility impact, the promotion needs to run for at least 10 to 14 consecutive days.
Can I run more than one promotion at the same time?
Technically yes, but it is not recommended. Multiple simultaneous promotions cannibalise the message, confuse the customer about what the main offer is, and make it impossible to measure what worked and what did not. One promotion at a time, measured on its own merits.
Do promotions on PedidosYa work the same as on Rappi?
The mechanism is similar but the customer base and grid structure differ by country and category. Generally, PedidosYa places more weight on rating in ranking than Rappi in Argentina. Free delivery promotions work particularly well on PedidosYa because its customer base is more sensitive to delivery cost than to product discounts.
---
Conclusion
Promotions on delivery apps are not inherently bad. They are a precision tool that, when misused, destroys margin without building a base of recurring customers.
The difference between restaurants that grow and those that complain "promotions do not work for us" is simple: the former use surgical discounts on high-margin items, during specific time slots, targeting specific customer segments, and always measuring post-promotion retention. The latter run "20% off everything" every couple of weeks and then wonder where the margin went.
If you want to build your promotions strategy based on real data from your account, start by knowing where you stand:
👉 [Audit your restaurant for free at Growth Delivery App](https://growthdeliveryapp.com/demo)
In under 5 minutes you have a diagnosis of which promotions to activate first, based on your margin, your rating, and your current grid position.