Integration-as-a-Service: The modern day railroad
“I was a little boy
Back in West Virginia
We lived very close to a railroad track
And I remember I used to dream about them trains
And where they might go
And the people that worked on the trains
And the train made music
It made you feel good
I can remember sometimes
I could just stand off all by myself
And just get funky list’nin’ to the train”
Integration-as-a-Service: A theme I first thought about 2-3 years ago that has increasingly come to light since then is that of Integration-as-a-Service; software products whose main value is solving integration pain points once themselves and then leveraging this across their customer base. I first saw this dynamic in businesses tackling product feed management who would help online brands sell different products across different marketplaces/channels in different countries. These companies would integrate with a PIM (Product Information System) to gather product data (could be different photos of a sofa in .jpeg form or data on dimensions of the sofa) and then sort it/create rules for how this data was represented and pushed to different end channels (could be Ebay in the UK, or Zalando in Germany). For large online retail brands selling cross Europe, this saves a ton of cost and resource vs building one’s own ‘rules engine’ for how each channel works. I think of the complexity as this: number of products in catalogue x number of marketplaces you want to sell across x number of countries x volume. Accordingly this is how pricing tends to work (more value, higher price) for this software. For large online retail brands this problem is significant. The other problem this category of software cracks is analytics, as data is returned to 1 place from different channels on how products/marketplaces/geographies perform which will inform the brand’s strategy of where to build inventory and which marketplaces/countries are selling the best. When I think about this type of company the main value is in the integrations it sells. It’s similar to a railroad, where they lay the tracks between all the channels and countries and rent it out to customers (albeit at >70% gross margin and low marginal costs doing this via software).
In fact, Europe can be fertile soil for these kinds of companies given the different markets that make up European GDP, though this dynamic can also work in very specific categories like headless commerce. Here are some other categories I have thought about recently:
Tax compliance: Integration-as-a-Service dynamics are present in tax compliance software that help commerce businesses automate tax processes and calculations when selling across Europe (e.g. Fonoa or Taxdoo). No commerce business would seemingly want to build its own rules and knowledge of sales tax/VAT in selling across Europe from scratch when they could just get this from one vendor quickly. The cost of getting this wrong can also mean poor customer experience and additional cost, both of which are important to commerce companies given margins are low and so every € counts.
International payments: One company Rapyd I heard about 3 years ago was executing on this Integration-as-a-Service strategy, building integrations with local debit providers in India on behalf of its commerce customers. This enabled customers to accept local payment methods by integrating with Rapyd which improves conversion and drives revenue. This makes total sense.
Headless commerce: I think there’s an opportunity in the web front-end space for companies like Vue Storefront to build the ‘glue’ layer between all the different components (CMS, payments, ecommerce/fulfilment etc) of running a high performance online store. By integrating with all of the best-in-class solutions for these different components of a modern store, Vue can offer its customers interoperability/best-in-class solutions without the risk of vendor lock in which has been a longstanding issue in the ecommerce platform space (think Adobe Magento, Salesforce Demandware, SAP Hybris etc). This is another example where I believe Vue’s main proposition is these integrations, where Vue has to build these once (whilst also maintaining them) and can leverage these across lots of customers over time.
From an investment perspective, I think these first-mover railroad businesses will grow at higher rates over time as there is a headstart dynamic at play given it takes time and resources to build and maintain high-quality integrations. Customers should naturally want to work with the vendor that has been building integrations the longest and thus has the lowest risk of downtime/things not working which would risk revenue/customer experience. This makes these types of companies more defensible than software products without much underlying technology (think of some marketing technology companies that really should just be a feature in someone else’s system) . The Integration-as-a-Service companies with the largest headstart should be category leaders (assuming good execution and strong technology teams) as there would be little to no value of a competitor offering similar integrations. The key will also be to build more functionality (e,g. analytics or reporting) on top of the integrations to grow this defensibility over time.
