Last-Mile Delivery: In-House Fleet vs 3PL vs Gig Platforms
A practical decision guide for e-commerce and DTC brands. The cost math, the service-level tradeoffs, and when to switch from one to the other.
Last-mile delivery is the most expensive part of your shipping cost — typically 41-53% of the total. It's also the part your customer actually experiences, so it disproportionately drives your brand reputation, your repeat purchase rate, and your customer support volume.
For most e-commerce and DTC brands, the question isn't "should we offer fast delivery?" — it's "what's the right model for our scale, our customer expectations, and our economics?"
There are three main options, and most successful brands end up using a mix. Here's how to think about it.
The three models, in plain English
1. In-house delivery fleet
You own the vehicles. You employ the drivers. You wear the brand. Think: branded sprinter vans with your logo, drivers in your uniform, your dispatch software.
Who it works for: Brands with daily order volume above ~150-200 in a single metro, customers who expect a specific branded experience, or products that need specialized handling (large/bulky, cold chain, high-value).
Pros: Total control of customer experience, brand visibility, no per-order margin paid to a third party, direct feedback loop from drivers.
Cons: High fixed cost, complex management, scales linearly (not step-function), hard to cover multiple metros efficiently.
2. Third-party logistics (3PL) with delivery
You hand the package to a fulfillment provider (often a national brand like ShipBob, Deliverr, or a regional 3PL) who picks, packs, stores, and delivers from their warehouse using their own drivers or a delivery partner.
Who it works for: Brands with 50-200 orders/day, businesses that don't have warehouse space, companies that need to expand into new metros quickly.
Pros: Predictable per-order cost, scales without capex, multi-metro coverage out of the box, less operational headache.
Cons: Lower brand control, the "3PL driver in a stranger's car" experience, harder to handle special cases, minimums and surcharges on small packages.
3. Gig platforms (DoorDash Drive, Uber Direct, Roadie)
You use a marketplace of independent drivers to fulfill same-day or next-day orders. Pay per delivery, no fleet, no fixed cost.
Who it works for: Brands testing same-day delivery in a new market, businesses with spiky volume, low-stakes items.
Pros: Zero capex, instant scale, good for pilot markets.
Cons: Wildly inconsistent service quality, no brand control, per-order cost is high ($8-15+ per delivery), drivers don't know your product, no chain-of-custody, problematic for high-value items.
The cost math, simplified
For a typical e-commerce brand shipping ~100 orders/day from a single metro, here's what each model costs per delivery (excluding packaging and inventory holding):
| Model | Per delivery | Monthly fixed | Service quality |
|---|---|---|---|
| In-house fleet | $3-6 | $15,000-25,000 | ★★★★★ |
| 3PL with delivery | $6-12 | $0-2,000 (storage) | ★★★★ |
| Gig platforms | $8-15+ | $0 | ★★ |
Break-even math: In-house fleet breaks even around 100-150 deliveries/day in a single metro (depending on the metro). Below that, 3PL is cheaper. Above 250-300/day, in-house wins by a lot — but only if you can keep utilization high.
The model that most brands should actually run
Hybrid. The smartest brands use all three, but with intentional routing logic:
Tier 1: In-house for your home metro
Where 60-80% of your orders come from. Brand experience matters most here. Drivers know your product. Customers see your vans. This is your flagship experience.
Tier 2: 3PL for adjacent metros
Where you have 10-100 orders/day. Don't build a fleet there — use a 3PL with delivery. Pick a 3PL whose service level matches your brand promise.
Tier 3: Gig platforms for off-hours and overflow
For orders that come in after your in-house drivers are off-shift, or for spiky volume days (Black Friday, promotions). The service quality is lower, but it's better than missing the delivery window.
When to switch models
Most brands don't switch cleanly from one model to another. They grow out of a model, then back-fill. Here's the typical progression:
Stage 1: 0-50 orders/day
Use: 3PL for fulfillment + delivery. Maybe a gig platform for premium same-day in your home metro.
Why: Capex is too much for low volume. 3PL absorbs the operational complexity.
Stage 2: 50-150 orders/day
Use: 3PL for fulfillment, in-house for delivery in your home metro.
Why: Delivery is your brand differentiator at this scale; 3PL drivers don't know your product. In-house delivery gives you the customer experience.
Stage 3: 150-400 orders/day
Use: In-house fulfillment + delivery in home metro. 3PL for adjacent metros. Gig for off-hours.
Why: At this scale, fulfillment in-house starts to pay off. You have the volume to justify the operations team.
Stage 4: 400+ orders/day
Use: Multiple fulfillment centers, regional in-house delivery fleets, 3PL for edge markets.
Why: You're past the threshold where capex on regional DCs pays off. Build the network that matches your customer geography.
The hidden cost most brands miss
It's not the per-delivery cost. It's the cost of bad deliveries.
Every wrong-address delivery, every late package, every damaged item triggers:
- A customer service ticket (avg cost: $15-30 to resolve)
- A potential refund or reship ($20-50 in direct cost)
- A negative review or social post (unrecoverable brand impact)
- A churned customer (the $200-1,000 lifetime value hit)
The brand that ships at $4 with 1% error rate is often more expensive than the brand that ships at $7 with 0.1% error rate. Focus on the total cost of bad delivery, not just the per-order cost.
What we do at DSS
We build and operate last-mile delivery for e-commerce and DTC brands in South Florida. Some clients use us as their home-metro delivery arm — we wrap their packaging, we wear their uniform, we drive their vans. Others use us as overflow capacity for the 3PL they already have. A few use us as their full-stack ops partner for a single region.
The right model depends on where you are in the growth curve, what your customer expects, and how much operational complexity you want to own.
Free last-mile cost analysis
Tell us your order volume, your metros, and your service-level goals. We'll model the cost across in-house, 3PL, and gig — and tell you which combination makes sense for your stage.
Request Free Analysis