Shippo and Amazon-Like Logistics For the Rest of The Internet
Background:
Amazon has spent the better part of the last two decades perfecting logistics. While they have established dominance on non-branded goods and continue to expand exponentially, the long-tail of e-commerce continues to thrive (as seen by the growth of Shopify, Wish, and even Jet, RIP) for a couple of reasons. In very, very simplified terms:
Lack of New Item Discovery Amazon can technically hold as many items on the site but for any given search term, new product discovery becomes harder over time. Amazon’s algorithm favors items with more reviews and customer engagement, which leads to even more traffic, create a self-reinforcing pattern favoring incumbents . The TLDR on this is that there will always be small businesses who in addition to Amazon, are looking for less-mature platforms where they can be a first-mover.
Branding: Amazon doesn’t work well for branded items — people rarely search for specific branded items on Amazon or descriptive keywords that might lead to new brand discovery (i.e. “all-natural, Shamrock-colored, gluten-free Yogurt). Brands who are able to create memorable experiences on other channels (via paid ads, social media following, WoM, etc.) have no reason to give up a % of profit that they can capture by bringing users right on their site or a marketplace with better profit-sharing economics
Private Label: While private label makes up an incremental part of Amazon’s existing sales, anecdotally it seems that there is long-term fear of Amazon cannibalizing certain categories. This probably fits into point (1) as well.
For primarily off-Amazon sellers, the sellers tend to be active across many different sales channels (their own Shopify site, Faire, eBay, whatever) since there’s no dominant alternative. This creates the need for one-size fits all solutions for Amazon-like capability across platforms. The best example of this might be fulfillment; in a tool-less world you’d have to predict demand and send branded packages to the fulfillment service of each platform you were on.
Most importantly, Amazon has re-conditioned customers to the degree that not having some of their features are actually growth levers, which is where plug and play startups are very profitably filling individual features. Have to enter too much payment information or sign up on a new site -> churn (Bolt), can’t get 2-day shipping -> churn (Deliverr), tacking on a $2.99 shipping charge at the end of checkout or no free returns-> churn (Shippo).
What is Shippo?
At the highest level, Shippo greatly simplifies what I call pre-carrier logistics. This is roughly anything in the shipping process from the time an order is placed before something gets on a truck to a fulfillment center or directly to the customer.
In practice, they do a couple of different things (primarily through an API-based approach, though they’ve built out a UX recently).
Order Aggregation: Shippo aggregates orders across different sales channels, which can be viewed/edited in a single neat dashboard. Common API logic can also be used to create shipping labels for orders from different sales channels.
Label Creation/Validation: When creating a label, Shippo’s endpoint automatically validates that the address exists (for US and Australia customers). Based on the carrier/shipping parameters sent, Shippo can generate an end label that can be printed and thrown on a package.
Rate Comparison/Shipping Scale: Shippo integrates with all the carrier services (DHL, USPS, etc.) to return a rate comparison for any specific package based on a variety of factors (route, zones, type of shipping, customs, package size, etc.). Because Shippo serves as a front for the rolled-up long-tail of customers, they negotiate wholesale prices on rates with some carriers, bringing down costs by up to 90%
International Shipping: Most impressively, Shippo provides the same services for label generation for international routes. This includes the ability to programmatically fill out custom declaration forms and rate-comparison (which is harder than domestic because of import taxes/tariffs, carrier route specialization, etc.)
Tracking: Based on historical route data/integration with the carrier’s APIs, Shippo can get the carrier event-data (packaged dropped off at facility, delivered, etc.), which can be used to generate a customized tracking page and/or whatever the merchant chooses to do with the event data (for example, send email or text alerts to the user).
Returns: Shippo also has an endpoint for generating return labels, though it doesn’t backwards integrate into the order and handle the event-tracking and financials behind reverse logistics.
The “why Shippo” here is rather simple: a) carrier APIs suck and are hard to integrate and analyze data (i..e you can get the zone to zone price, but how is that actionable), b) transparent pricing on label generation, c) e-commerce is booming and d) the little things done well (international shipping, package insurance, etc.)
Shippo fundamentally serves two types of customers. :
Small Businesses: Shippo primarily caters to small businesses who are doing their own shipping to take — logically, any company using a 3PL wouldn’t use it and enterprise-size companies managing their own fulfillment likely already have wholesale carrier rates, different customs requirements, and tracking systems. Speculatively, a large % of these SMBs are probably DTC brands who are in a position to own the entire buyer experience. The company claims to be serving warehouses/3PLs, but I’m skeptical this is a large % of the customer base having seen the thought put into optimizing warehouse operations.
Asset-Exchange Marketplaces: Marketplaces that require a seller to ship directly to customers without holding inventory (eBay) can integrate Shippo into their marketplaces. This gives the seller the easiest shipping experience possible and passes lower prices onto the end-buyer without the marketplace having to hold any of the complexity. Shippo counts a couple of huge marketplaces (GOAT, Mercari) as customers and there’s a pretty logical pipeline of customers here as well (Faire, Ritual, etc.) if they can perfect the financial side of reverse logistics as well.
Where is Shippo Going?
Shippo’s future direction is much more interesting than their current product; strategically they have a couple of things to think about.
Scale Ceiling: Shippo’s advantage on pricing negotiating with carriers has decreasing marginal returns. At a certain point carriers, even with wholesale shipping volume, need to break-even and Shippo’s customer base growing isn’t going to give them anymore leverage.
Competition: The competition in the programmatic shipping space is extremely tight (EasyPost, ShipHawk, ShipStation, Shippo, AfterShip, Stamps.com). And this ties into point 1; you just have to get to “enough scale” to be competitive on price, after that all advantages are value derived from extra product features. Different players in the space are trying different things — EasyPost unsuccessfully tried to move into fulfillment (actually managing the warehouse instead of utilizing excess capacity like Deliverr), Shiphawk actually tried multi-warehouse optimization, Shippo is leading the way on returns etc.
Does Shippo Scale With Customers: One of the more interesting things about Shopify (which caters to a similar demo as Shippo) is that their revenue has a Pareto-like breakdown. The top 20% of customers who manage to scale up make up a majority of their revenue — through payment costs, app rev-sharing, etc. If a DTC customer scales, they have 2 options — move to a 3PL or build out their own fulfillment infrastructure. With a 3PL, Shippo can still serve the merchant but the decision-maker becomes the 3rd-party not the merchant, which potentially weakens the relationship. For companies building out their own fulfillment at scale, there’s a big opportunity for Shippo to provide some sort of service that helps choose the right infrastructure (best warehouses/fulfillment centers) based on a combination of carrier insights and the company’s existing shipping data (merchant’s geographic concentration of customers).
Platform Consolidation: Platforms working together (see Walmart<> Shopify’s deal) is a potential threat way in the distance. One of the reasons that Shippo exists is because of the fragmentation of non-Amazon sales channels (and to be clear this isn’t novel to Shippo, it’s true for any company building e-commerce infrastructure). While Shopify can never build “all-for-one” tools with excellence, they can get pretty close over time.
Potential Roadmap:
Looking from the outside in, it seems Shippo directionally has three different options:
move upstream to packaging/shipping supplies: (package design, package sizing, marketplace with package manufacturers /generic shipping supplies)
move downstream the shipping process (enabling fulfillment/becoming a 3PL, wholesale management, warehouse logistics)
build tools for SMBs outside of shipping: using order data and large number of small businesses on the platform to enable advantages in different areas (i.e. bringing scale to buying non-programmatic ads inventory, cross-site recommendation engine, )
I’ll dive into a couple of those sub-options:1. Become a Full-Scale Shipping Marketplace: Shippo is already a marketplace at core — they fundamentally connect carriers with e-commerce businesses. One of Shippo’s big challenges (as previously mentioned) is helping customers fill in logistics gaps as they scale, preferably in an asset-light way. The challenge for merchants here isn’t actually the integration with a custom branding packaging manufacturer, warehouse management system, or 3PL, but actually finding the right service for their specific needs (all three of these 3rd parties are hyper-fragmented and have huge SEM spends).
What if Shippo could help bridge this gap? The TLDR design here would be to have a discovery section for each type of 3rd party provider (small 3PLs, big custom packager, etc.). Each section would link to a specific 3rd party provider, who would have a detailed page about the services they provide, transparent pricing, and the type of business they catered too. There would then be a built-in method of communication between merchant and 3rd party they select (asynchronous email or live chat, etc.). Reasoning:The Incentives Line Up: In a hyper-fragmented industry, service suppliers are always looking for a new way to reach incremental customers. Buyers already on the platform (merchants) can streamline the amount of extensive research that they have to do across for a given service. Plus, in weird times (holidays), this could be the ideal way of filling inefficiencies quickly. Since every provider/customer is vetted/customer-reviewed, if a merchant needs an extra warehouse or a fulfillment center has extra capacity, they can move quickly to connect with a large group of customers.
Free Distribution : There is no centralized review platform for logistics services. Merchants on Shippo will leave verified reviews for each of these services and good services will likely tout their customer satisfaction (i.e. 98% approval!) by linking to their Shippo reviews, driving some organic traffic back. Given how heavily these specific-types of 3rd parties invest in SEM, this is likely to be high-intent traffic in Shippo’s core demo. This extra traffic would lead to more conversions on Shippo’s core product.
Post-Shipping Analytics: Shippo has mentioned “post-shipping analytics” a couple times on their site. In practice, all downstream logistics (a carrier, 3PL, warehouse, etc. etc.) are quantitatively evaluated on only one vector: price. But this could be an interesting way to measure how effective certain parties are on service (3PL on fulfillment time or Package Designer on Customer Satisfaction). The simple way of measuring this would just be through user-feedback (i.e. “this took too long”), but part of me wonders if there’s a quantitative way. If Shippo already knows the average carrier time for a certain type of package/carrier, plus knows the order date — can you somehow subtract them away to find the average service time (even if they aren’t integrated with the carrier)?
Rev Model: The revenue model here would be similar to something like G2, have affiliate deals with each provider that comes on the platform and earn a fixed fee when a customer converts. Another G2-like model would be to sell leads to providers (i.e. the company info of a customer who views the provider page). If this really scales, a possibility would be to offer an ads program for prime page ranking.
What Does This Look Like? The big challenges here are a) supply acquisition and b) how to evaluate quality of given service and present it in a nice way to merchants. Supply acquisition is a growth/BD problem — quite frankly, solving the cold start problem might require cold-calling/building relationships with a bunch of the target customers in whichever 3rd partyvertical is most logical. Aggregators with big audiences (i.e. Deliverr, Lumi) might be a good first step.
The intuitive way of measuring quality is to build out a reviews system (1-10, five star, etc.). A better way might be to make users evaluate the third-party. on different dimensions (customer service, quality of service, price, deliverables, etc.). Each additional category brings more friction, lowering the chance of completion (which is essential to such a system).
The second question would be how to actually present customer reviews. Having a customer’s identity (i.e. Sue’s Flower Shop) is a great trust mechanism that most sites use. However, for something like custom packaging or 3PL, a merchant might want to see reviews from customers with a similar business profile (size, industry, etc). An alternative would be to ask a merchant to enter basic information about their business, but stay anonymous (Business A, with 500 employees and $20 mil of revenue rated 5 stars).
In terms of visual design, think doing something like Faire does for sellers to help showcase their goods (in this case services) might be a good look!
Trade-offs:Massive use of sales/dev teams, large opportunity cost
No short-term revenue benefit
Potential moderation and vetting of customers needed.
Integrated Package Pricing: Shippo currently provides by value by allowing users to optimize their price once they’ve decided on the type of package they want. But what if the opposite was true; what if shipping prices could help decide what product/packages should be used? A good example of this would be a feature that allowed users to define the absolute specs of their product (not their packaging). Shippo would then have a dynamic recommendation engine showing what type of packaging other customers with similar-sized products were using and what you’d save by switching to the most popular choices.
The extension of this would be to allow a direct linkage system to a Shippo customer’s manufacturer, who could then potentially change the product — this would be especially beneficial for “volume based products” (i.e. something like a protein-powder or a drink), where per/oz prices could become cheaper. Why would something like this be needed?Carrier prices fluctuate pretty wildly year over year and there aren’t linear increases across package sizes. Choosing a different type of packaging might substantially reduce costs or if costs fall, gives the opportunity to create additional packaging inserts or product volume.
For many smaller businesses this might be something a) be something they are even thinking about or b) aren’t aware of the different package sizes available for each carrier
What Does This Look Like? The easiest way of implementing this might just be to create a separate endpoint on Shippo’s core
API (Recommended Package Size by Carrier, Sample Size, Existing Package Type (to measure savings)). Adding a parameter to the core “Shipping Object” would be a little odd since it requires the parcel dimensions. There’s a data science aspect to this as well: pre-loading all the recommended package types, how each update changes the recommendation, etc. but going to just gloss over the details here (: