Introduction
If you manage delivery operations in Latin America—whether e-commerce, restaurant, supermarket or drugstore—you know that the last mile can be your best ally or your worst enemy. I have seen businesses in Ecuador, Mexico, Colombia and Argentina who sell well but lose money on each delivery without realizing it. And others that, with the same volumes, are profitable because they measure what matters.
In this article I share with you The indicators we use in Picker to evaluate delivery operations, why they are critical in the Latin American context (where things work differently), and how you can apply them without the need for a data science team.
We explain it to you with real data
The last-mile market in LATAM is booming. According to IMARC Group, we went from USD $11.85 billion in 2024 to a projection of USD 43.66 billion for 2033 — an annual growth of 15.6%.
Brazil and Mexico lead the volume, but the fastest growth is in markets such as Colombia, Chile, Ecuador and Peru, where e-commerce still has a lot of room to grow.
But the beautiful numbers hide a harsh operational reality: the last mile represents 53% of the total logistics cost. More than half of what you spend on logistics goes to the last stretch.
This means that if you don't optimize the last mile, you're leaving money on the table — or worse, losing money on every delivery.
The 5 KPIs that really matter
1. Cost per Delivery (CPD) — The Key Financial Metric
This is the number. If you don't know your real CPD, you're operating blindly.
What to include in the calculation:
- Direct costs (fuel, salaries, vehicle maintenance)
- 3PL commissions or delivery platforms
- Incident costs (reattempts, returns, compensations)
- Packaging and shipping materials
- Customer service costs related to deliveries
Formula:
CPD = (Fixed Allocated Costs + Variable Costs + 3PL Commissions + Incident Costs)/Number of Deliveries
How we apply it:
We calculate CPD by zone, by time zone and by provider. We often find that certain zones or fringes destroy the margin without the business knowing it. We also systematically compare our own fleet vs 3PL — the results vary depending on the context.
Practical goal:
The CPD should be below 15% of the average ticket. If it exceeds 20%, there is a structural problem to solve.
2. First Attempt Delivery Rate (FADR)
Every retry is money that is lost. In LATAM, this KPI is especially critical because of the number of variables out of control.
Half of failed deliveries are due to incorrect or incomplete addresses. This is solvable.
In LATAM, urban variability (non-standardized addresses, security, availability of delivery people) reduces FADR compared to mature markets; data from platforms and route analysis show notable differences by city and supplier (Parcel Perform).
(FADR chart by sector —restaurants, supermarkets, pharmacies— with estimated minimum/average/maximum ranges for LATAM)
How to improve it:
- Confirmation via WhatsApp before the delivery person leaves (free and highly effective in LATAM)
- Required field of reference and between streets at checkout
- Real-time zip code validation
- Show location on a map for the customer to confirm the pin
Goals by type of operation:
- With external 3PL: minimum 90%
- With own fleet: minimum 95%
- If you're below 85%, this must be your problem #1
3. Average Delivery Time (TME) and Promise Fulfilment
The most common mistake is to promise times that cannot be consistently met. In LATAM, it's better to communicate “60-90 minutes” and comply, than to promise “30 minutes” and fail.
What to measure:
- Time from when the order is ready to effective delivery
- % of deliveries within the promised SLA
- Variation by zone and time
Practical recommendation:
Segment your delivery promises:
- Zone A (high density, close to warehouse): 30-45 min
- Zone B (medium density): 45-60 min
- Zone C (periphery or complicated areas): 60-90 min
The reality in LATAM: Traffic in cities such as CDMX, São Paulo, Bogotá or Lima can triple delivery times during peak hours. Don't promise the same thing at 10am as at 6pm.
4. Incidence Rate
Damage, missing, wrong deliveries, product in poor condition. Each incident has a direct cost (compensation, reshipment, lost product) and an indirect cost (loss of trust, negative reviews).
Incidents continue to be a major source of cost and loss of trust in LATAM. A significant percentage of these incidents are linked to incorrect or incorrectly formatted address data, which generates reattempts and additional costs; address quality studies estimate that up to 10— 25% of delivery problems are due to erroneous data, and the average cost per failed last‑mile analysis attempt is around US$17.20 in retail/logistics studies (Veho).
The graph shows how the average cost per failed attempt (USD $17.20) is multiplied according to the percentage of incidents due to incorrect addresses. In a scenario of 1,000 orders:
- With 2% of incidents, the total expense is USD $344
- With 3%, it rises to USD $516
- With 4%, it reaches USD $688
- With 5%, it reaches USD $860
The visualization shows that Each additional percentage point of incidents significantly increases last-mile costs, showing the importance of validating addresses and reducing errors to protect operating margin.
What to register:
- Type of incident (damage, missing, error, extreme delay)
- Average resolution cost by type
- Responsible supplier or deliverer
- Frequency and trend
How to reduce it:
- Mandatory departure checklist for delivery people
- Delivery photo as proof (resolves 70% of disputes)
- Packaging protocols by product type
- Contractual KPIs with 3PLs with real consequences
Prioritization:
Classify by severity and frequency. It attacks the ones that cost the most, not the ones that generate the most noise.
5. Delivery Density
This is the operational efficiency KPI. Higher density = lower CPD.
What to measure:
- Hourly deliveries per delivery person
- Deliveries per kilometer traveled
- Average delivery time (including waiting)
- Route occupancy
How to optimize it:
- Group orders by zone and time window
- Give the customer delivery window options
- Implement pickup points in high-demand areas
- Encourage customers to choose convenient times (5-10% discount for flexibility)
Case Study:
Operations in LATAM that implement 2-hour delivery windows instead of “same day” can reduce their CPD by 15-25%.
Why measuring this matters to your business
The data is clear:
- 89% of customers buy back after a good delivery experience
- 65% prefer to pay more for fast and reliable shipping
- 70% they don't come back after a failed delivery
In LATAM, where competition in e-commerce intensifies every year, the delivery experience is a real differentiator. It's no longer enough to have a good product and a good price — if the delivery fails, you lose the customer.
What do you gain when you measure these indicators
Actual profitability per order:
You stop trading with misleading averages and you can identify exactly which zones, schedules or suppliers destroy margin.
Visibility into hidden costs:
Reattempts, incidents and compensations cease to be “part of the business” and become quantifiable problems.
Data-based decisions:
You can compare your own fleet vs 3PL, evaluate new suppliers and negotiate contracts with real information.
Ability to scale with control:
Grow volume without costs skyrocketing proportionately.
What do you lose when you DON'T measure them
Margin without realizing it:
You may be selling orders that seem profitable but that, when you add retries and incidents, end up in loss.
Customers for broken promises:
Without measuring TME, delays become chronic and the customer leaves without explaining to you why.
Money in avoidable retries:
A low FADR means logistics costs that multiply without structural correction.
Bargaining power:
If you don't compare KPIs by 3PL, you don't know who complies and who doesn't.
Conclusion
The last mile is the lever that decides if your delivery operation is profitable and if the customer returns. It doesn't require sophisticated technology or huge equipment — it requires discipline to measure the right thing.
My recommendation: start with CPD, FADR and TME. These are the three most impactful indicators. Implement a 3-4 week measurement pilot and you'll discover things about your operation that you didn't know.
Don't look for the perfect entry solution. Test, measure, adjust. With a well-orchestrated mix of your own fleet where density justifies it and 3PL for peaks or areas of lower volume, you'll gain control over the experience, reduce costs and increase recurrence.
And that, in the end, is what makes a delivery business in Latin America sustainable.
