DEEPER DIVES: Europe Is Crushing Last Mile Delivery With APMs And PUDOs

Good Morning,

First, as always, thank you for joining.

Last mile is one of those areas that new-age providers all try to pretend is a math problem.

While your numbers have to work, the reality is there are also winning and losing conditions when it comes to delivery. While the US is obsessed with trying to drive costs down through low cost gigwork labour, companies in Europe are looking at things different.

There were a lot of new subscribers who joined DEEPER DIVES to see this take on InPost and APMs. Welcome. This is edition 55 of DEEPER DIVES, now that you are subscribed you have access to everything I have every published.

Here’s what this issue brings:

• Out of home delivery isn’t something that gets a ton of attention in North America, but the most profitable logistics providers in Europe are doubling down on the benefits of APMs and PUDO points. Here’s everything I’m sharing after InPost shared huge results last week
  

• Looking to drive more insights from your silo’ed system files? Try DuckDB + Metabase

Will Record Profitability Finally Make Lockers The Default Delivery Strategy?

InPost concluded their F24 fiscal year and shared the results on Friday.

Impressive is a bit of an understatement.

For a company doing last mile logistics, they absolutely crushed it.

Here’s a summary of what they achieved:

• Parcel Volumes:

  • 1.1 billion parcels delivered (+22% YoY), outpacing e-commerce market growth in all geographies​​

• Revenue:

  • PLN 10.9 billion (+23.5% YoY) - $2.66B USD

• Adjusted EBITDA:

  • PLN 3.6 billion (+33.5% YoY), with 33.3% EBITDA margin (+250bps YoY)​ - $878M USD

• EPS:

  • Nearly doubled, +92% YoY​.

• Free Cash Flow (FCF):

  • PLN 934.5 million (+22.3% YoY) at Group level - $228M USD

  • Poland: PLN 1.6 billion (+32.5% YoY) - $390M USD

  • International: -PLN 661.6 million (reflecting ongoing investments)​​ - ($161M USD)

• Net Leverage:

  • 1.9x, down from 2.2x in 2023, despite heavy investment and M&A (PLN 400 million, $98M USD)​

• CAPEX:

  • PLN 1.4 billion, +37% YoY, primarily for network expansion​ - $341M USD

What caught my attention was the profitability.

33.3% … in an industry that is traditionally MUCH lower margins.

For comparison, this is what is typically seen from different providers in the market:

The current level of performance that is being achieved is pretty much double what we are seeing from ‘traditional’ vehicle-to-door competitors.

Two major questions hit me when looking at the results.

First, what is InPost doing differently than its competition in Europe.

And second, if parcel lockers are dramatically outperforming vehicle-to-door delivery, why are countries like the US and Canada so far behind.

What’s InPost Doing To Thrive

Out of home (OOH) delivery is nothing new in Europe.

Customer expectations have shifted. There is more emphasis on flexible delivery options, sustainability, and reliability when consumers make purchases online.

This has paved the way for out-of-home delivery methods to gain traction, as they can offer efficiency benefits.

In 2023 alone, the number of parcel lockers in Europe jumped by 29% (year-over-year), while the count of PUDO service point locations grew about 6%.

InPost (which has one of the most robust locker networks), is highly focused on their APMs over increasing the number of PUDO points.

APMs allow companies like InPost to achieve near perfect first time delivery success (over 99% SLAs) while simultaneously providing customer with convenience and security when it comes to collection.

Another large benefit for locker networks is the fact that many of them are able to be accessed 24/7. This provides huge ranges of time for customers to collect their goods, but also gives providers like InPost greater flexibility when it comes to delivering and collecting from parcel locker banks.

Because of its focus on APMs, InPost has the largest locker (not OOH) network in Europe.

Traditionally, other providers have focused more on PUDO networks in order to leverage the benefits of OOH delivery and drive efficiency.

While this approach does make expansion and service point density building easier (because you are leaning on partnerships and avoiding the CAPEX cost of a locker infrastructure), it comes with its own challenges.

• More limited opportunity to customize and tailor the user experience
  

• Managing and maintaining the service hours from different PUDO locations
  

• Managing and working with different capacities at each PUDO location (since their are often partner businesses, the number of packages each location can hold varies)
  

• Potential impact to the overall network as partners leave and substitutes need to be found

InPost has taken the right view of the costs associated with building out their infrastructure. Instead of seeing lockers as the alternative to partnership based PUDO networks, they are looking at it compared to traditional vehicle-to-door models.

And because of this, they see the cost benefits of the CAPEX.

In addition, they are leveraging the operational activity to build out a more robust digital presence.

InPost has launched a number of initiatives to create a great digital ecosystem and easy to use consumer app.

• InPost Pay: A fully integrated payment, delivery, and return solution that simplifies online shopping. Enables 1-click checkout, positioning InPost as a fintech-meets-logistics player
  

• Loyalty Rewards: A gamified loyalty program launched in 2024. Offers "InCoins", perks, and rewards to drive more frequent use of InPost services
  

• App Quality: App users order 40% more than non-users. The UK and France apps have exploded in usage. For the UK, 2 million downloads, doubled YoY while the app has hit #1 in France for app store ranking in Q4 2024

This flywheel is elevating their brand presence and significantly increasing their B2C activity and success with major retailers throughout Europe (most retailers won’t invest into parcel networks until they can cover ~90% of a target geography).

InPost has also been using their data to drive better utilization of the locker network in every country they operate in.

Lockers in the UK (for example) ran at over 100% utilization, and InPost is doubling down on APM deployment due to demand outstripping capacity.

InPost sees APMs as its long-term moat and margin driver, while PUDOs are treated as a complementary channel — useful for geographic coverage but not the focal point of the growth strategy.

If Lockers Are So Good, Why Isn’t The US Onboard?

First, APMs and PUDO is growing in countries like the US.

Amazon has over 40,000 lockers in the US with another 60,000+ counter (PUDO) locations from different partnerships.

UPS and FedEx both have extensive PUDO networks driven by their own stores/locations and external retailers. Over 90% of the US population is located within 5 miles of a UPS pickup/dropoff location for example.

Resistance from markets like US comes more from challenges with density. In Europe and Asia, these systems thrive in dense urban environments where consolidation of deliveries offers clear cost savings, while in North America, suburban sprawl and cultural preferences for doorstep delivery have hindered widespread adoption.

The above graphs show some of InPost’s largest markets compared to the top 20 US cities for population density.

The constant line in both graphs is set to 6,550 people per square mile.

On average, you can see that the European markets are a better overall population density to support the required infrastructure for deep APM coverage.

One popular frame for the unite economics of having an effective OOH network is that you generally need 10 OOH delivery points per 10,000 people (InPost’s research specifically calls out 12/10k).

Generally, consumers won’t regularly use OOH services if collection points require >7-10 minutes of walking.

This is a primary reason why PUDO activity has been the option of choice in order to provide enough options for consumers.

But, by taking on this location point requirement with APMs, InPost is showing the enhanced profitability and customer loyalty it can drive.

In Poland for example, InPost has 20 OOH points for every 10k people and more than 80% of delivery volume goes to an APM.

Given the differences in city planning and construction, it’s fair to say that a successful locker strategy for North America will need to take on a different twist from the success that company’s like InPost are seeing in Europe.

For APM popularity to grow in NA, here’s what needs to happen:

• Networks need to be designed for top consumer convenience
  This means having lockers strategically placed for where consumers work and play
  

• Lockers need to become the execution path for rapid delivery. Tying faster delivery services to APMs would drive user adoption
  

• APMs need to be positioned by retailers as a cheaper delivery option. While free delivery has long been a staple of NA eCommerce, we are seeing shifting trends where consumers are paying for expedited service. Pushing free delivery to slow 2-3 day options and offering an APM at a reduced cost versus a vehicle-to-door alternative is required
  

• Open/Carrier-Agnostic parcel network (or agreements to share access)

The Future Of Lockers

While lockers are driving amazing results for InPost and their ability to drive down last mile costs while also reducing CO2 emmissions has been demonstrated by multiple organizations, they aren’t without their challenges.

First, they are “hardware”. An APM network is an investment and requires significant funds in order to build the lockers and deploy them. The faster that you want to scale, the more money an organization will spend over a shorter period of time.

Second, location matters. For lockers to be highly effective, it means that the placement and the community it serves are extremely important. There is only so much prime real estate to go around, and every provider essentially needs the same geographic benefits.

Third, lockers require quite a bit of regular maintenance to keep them working perfectly for consumers. This means that by owning the hardware for the network, an organization needs to also have a decent sized maintenance activity to support it. These maintenance operations only continue to increase as the size of your network does.

City push back. While everyone appreciates the benefits of APMs and the benefit they provide to consumers, more cities are implementing regulations when it comes to installing them. Paris, Barcelona, Amsterdam, Florence, Berlin, and San Francisco are all examples of cities that have or are looking to implement specific guidelines for how and where lockers can be placed (the growing trend now is for them to be installed underground in transit stations and other connection points).

None of these challenges themselves are deal breakers, but the writing is on the wall when it comes to access, or better thought of as “shared” access.

Delivery providers will have to get a to position where they are sharing lockers and have decent agreements in place for how they are to use each other’s network and what capacity limits they will have to respect.

Those organizations that build their own infrastructure will have advantages to those who simply build their offering on using everyone else’s hardware.

Lockers also need to be a strategic offering where different types are put into different communities based on volume and consumer expectations.

While InPost (for example) allows up to 48 hours for a package to be picked up today, that length of time will have to drop down to 24h and likely to 12h in order to keep up with volume increases without having to add more physical capacity to the network (throughput is already something they are managing very well).

Appending - Additional Facts & Figures

• Preference for home delivery among European online shoppers fell by 9% in one year (from ~73% to 64%)

• Out-of-home methods comprise 20–30% of parcel deliveries in Western Europe (the rest being to-door)

• InPost, DHL, DPDgroup together operate over 70,000 lockers across Europe

• Parcel lockers can reduce last-mile delivery costs by up to 78%

• Lockers can move 300-400 parcels per day where as the most effective and efficient truck routes do 150-250

• APM locations can be visited up to 30 days a day by different carriers

Well All Know Your Data Is Everywhere. Get Deeper Insights From The Information You Already Have

It’s super trendy in logistics to talk about tools these days.

Everyone wants you to be on their platform and get the benefits they claim they can deliver.

While there are huge advantages for your business when you systems are well connected, the reality of a lot of brands and smaller service providers is that you have data in a lot of different systems.

And most of the time, those systems don’t speak well to each other.

One area that I have been working in a lot more with my clients is creating workflows to take data from different systems and make them easier to work with.

That’s where DuckDB comes in.

Basically DuckDB is an open-source column-oriented Relational Database Management System. It is designed to provide high performance on complex queries against large databases in embedded configuration.

On top of being able to query from a variety of database sources, it also gives you the ability to read in CSV files (think data dumps from your independent systems) and quickly create your own local database file.

Different files can be loaded to their own databases, or you can simply load all of the files (as separate tables) into one consolidated database file.

For developing insights and analytics from your system transactions, I’ve been dumping all of these files into one consolidated database.

When you have all of your separate files as different tables all in the same database file, you can now use tools like Metabase (or other BI platforms) to read the DuckDB file and start performing different types of analysis on the information.

You can use Metabase to join data from different tables as well as do all of the aggregations and groupings to support the work you are doing.

This type of flow can handle huge amounts of data, is much faster (and safer) than trying to do anything in something like Excel and, since you are reading the data into another database container, your original information stays completely untouched - so you always have a record of your original extracts and can always validate totals independently for validation.

That’s it for this week. Thanks for being here.