The App Era Is Over – Integration Is the New Frontier

01 /The Concept

StackOne buried a line near the bottom of its website that deserves top billing: its goal is to help developers of cloud-based B2B products say "yes" to their customers more often.

Not to every question. But at minimum to questions like: "Does your product integrate with the tool we're already using?"

To make that possible, StackOne built a platform that lets B2B SaaS developers add hundreds of integrations with third-party enterprise services – enabling seamless data flow in both directions.

The platform exposes a unified API. That means a developer can make a single call – say, "fetch customer data from the CRM" – without caring which CRM the client actually uses. StackOne handles the translation, routing the request to the right system in the right format.

A sign of the times: StackOne built the same kind of unified API layer for AI agents, which are increasingly being deployed inside companies to handle a wide variety of tasks.

The platform currently offers ready-made integration modules for 200 popular enterprise platforms, covering more than 3,000 distinct data exchange and task-execution actions.

StackOne also ships a tool that lets B2B product developers quickly build and publish new integrations for niche products not covered by the standard catalog – for cases where specific clients need specific connections.

Founded in the UK in 2023, StackOne already has paying customers. It raised its first €3.3M in the spring of last year. Its latest round – $20M – was led by Google Ventures.

02 /Why It Matters

Why would Google put $20M into a platform for what sounds like a mundane infrastructure problem?

Start with scale. Modern companies run on an extraordinary number of cloud services. Companies with up to 200 employees use an average of 42 different platforms; companies between 200 and 750 employees use 103; companies above 1,500 employees use around 150.

All those services need to talk to each other inside unified business processes. As a result, nearly 40% of enterprise buyers consider fast integration with existing infrastructure the most important criterion when evaluating a new product.

The problem: building even a single integration module takes a developer at least three weeks of work. A product that needs integrations with a hundred services? That's 300 weeks of integration work alone – a completely unmanageable burden for most teams.

And building an integration is only half the problem. It then needs to be maintained as the external service updates its APIs and features. That ongoing maintenance burden is enough that 50% of existing integrations eventually get dropped from products – because keeping them current is more hassle than it's worth.

The result: 83% of B2B SaaS developers list integrations as a top priority, yet can't handle it alone. Platforms like StackOne – which let developers embed hundreds of integrations essentially by waving a wand – have become genuinely valuable.

The payoff is measurable: nearly 60% of developers say having the right integrations directly increases their close rate on new deals. And customers who have integrated a product into their infrastructure churn at roughly 60% lower rates than customers using the same product standalone. Once something is embedded in your processes, the activation energy required to rip it out – even if you're not in love with it – is usually more than it's worth.

There's also a market expansion angle: integrations with niche tools often let developers break into verticals where those tools are dominant. About 52% of B2B product teams report successfully entering new market segments this way.

The integration platform market reflects all of this. It hit $14.5B in 2024 and is projected to reach nearly $75B by 2033, driven by the continued growth in the number of services companies use – which makes integration more necessary and more valuable over time.

StackOne is a relatively new entrant in this space, but it's not alone. Several predecessors have already raised significant capital:

Merge ([covered here](/review/integracija-cherez-odno-mesto)) raised $74.5M for a similar platform – first reviewed when it had just raised $4.5M back in 2021.

Prismatic ([covered here](/review/samoe-to-dlja-programmistov)) raised $34.3M.

Finch ([covered here](/review/zolotaja-integracija)) focused exclusively on HR-system integrations and still raised $62.1M.

Paragon ([covered here](/review/s-kogo-brat-dengi)) raised $21.2M on a platform that promises to build integrations 10x faster than doing it from scratch.

03 /Opportunities

The macro trend is the internet becoming a single unified information fabric, where a proliferating ecosystem of services interoperates seamlessly. The corresponding opportunity is building the infrastructure that makes that interoperability real.

The simplest version: connecting services through existing APIs. Even that is nearly impossible to do comprehensively, which is why a cluster of specialized platforms – including those mentioned above – has emerged just in the relatively narrow niche of enterprise service integration.

But the integration problem is actually deeper, because vast numbers of services don't have public APIs at all – yet still need to be integrated to power useful products. This gave rise to a separate category of platforms that provide unified API access by scraping external sites and serving the data through their own API layer. Deck ([covered here](/review/hochesh-stat-monopolistom)) built exactly this – covering 100,000 services and raising $16.5M in the process.

A distinct and emerging problem: building integration platforms *for AI agents* – enabling them to interact with other agents or with traditional web services. StackOne is partly moving in this direction, but some startups have focused on it exclusively. Anon ([covered here](/review/bez-jetogo-oni-nichego-poleznogo-dlja-tebja-ne-sdelajut)) built a platform that lets AI agents authenticate and perform actions inside third-party services on behalf of the user who authorized them – even simple things like sending a message on Instagram. Anon raised $6.5M.

And Crosshatch ([covered here](/review/shapochki-iz-folgi-vyhodjat-iz-mody-teper-nuzhno-vot-takoe)) found a genuinely creative application of the same infrastructure: if a developer connects to its platform, users can grant the app access to their personal data from third-party services. A travel app, for instance, could pull what destinations the user is currently browsing on Instagram, retrieve past trip history from Google Maps, and immediately surface the most relevant ideas for a next trip – with no filters or search required.

All of this connects to a broader shift that's already underway: people are tired of juggling dozens of separate apps and services. That frustration is fueling the super-app trend – applications that consolidate multiple functions from completely different domains into a single experience, essentially integrating many services under one roof. The super-app market is projected to reach $426B by 2030.

Integrating everything with everything, through every possible mechanism, for every possible purpose – this is clearly a space that's accelerating. The question is which piece of it you want your startup to own. The highest-leverage positions are protocol-level infrastructure and the first unified API layer in any category that's still fragmented – those compound over time the way Plaid did.