Every month I teach a Tech Term You Should Know (TTYSK) and a tech essay to level up your technical literacy and communicate well with dev teams. Ask me anything and I'll cover it in an upcoming issue.

This issue's TTYSK is "Data Isolation". Scroll to end to learn more 👇

Join me and 400+ product managers for a special event on Thursday, October 3rd.

I’m hosting a webinar on How to Master Technical Literacy: Evolving your Approach Thursday, October 3rd and you’re invited!

In this talk, you'll find out where your current technical skills are, learn what skills are essential for building technical literacy as a PM, and evolve your approach to technical literacy.

Interested? Register for the special event 👉️ here. Can’t wait to see you!

In 2012, when Adobe launched Creative Cloud, they focused on serving large enterprises like Amazon and Netflix—companies that needed custom solutions for managing creative assets. At the time, the best way to meet those needs was through a single-tenant architecture, where each client had its own dedicated instance of the software, with separate databases and resources.

This setup made sense for Adobe’s early enterprise customers because it gave them maximum control. Each company had its own “copy” of Creative Cloud, isolated from others, and they could customize the platform to meet their exact workflows. But as Adobe started growing and attracting small-to-medium businesses (SMBs) and individual users, the single-tenant model became a bit of a headache and a move to a multi-tenant architecture was necessary.

What Is Single-Tenant vs. Multi-Tenant Architecture?

In single-tenant architecture, each customer has their own instance of the software—essentially, they get their own "private server." It’s great for companies that need a lot of customization or have strict security requirements. But here’s the downside: every time you roll out an update or add new features, you have to apply those changes to each customer’s instance individually. This gets costly and complicated as more customers come onboard.

On the other hand, multi-tenant architecture means all your customers share the same instance of the software, with data kept separate. This makes everything easier to manage because updates and fixes happen across the board, and it’s a lot cheaper to scale.

Adobe’s Early Days: Single-Tenant Architecture

When Creative Cloud launched, Adobe used a single-tenant setup to accommodate the complex needs of their enterprise clients. Companies like Amazon and Netflix needed more control, customization, and compliance due to the sensitivity of the creative data they handled. Each customer got their own isolated instance, which made it easy to configure Creative Cloud to their specific needs without affecting anyone else.

But by 2015, Adobe had a much bigger problem on their hands: Creative Cloud was no longer just for massive enterprises. The platform had exploded in popularity with 12 million subscribers, including freelancers, SMBs, and creative teams around the world. These users didn’t need (or want) the heavy customization that enterprise clients demanded—they just wanted access to powerful tools at a reasonable cost.

Maintaining separate instances for everyone wasn’t cutting it anymore.

The Shift: Why Adobe Moved to Multi-Tenant

By 2015, the writing was on the wall: Adobe needed to switch to a multi-tenant architecture. They started the transition in 2016, and while the shift didn’t happen overnight, it was a huge step forward in scaling Creative Cloud. The process itself took about two years from start to finish, involving a lot of work from their engineering team to consolidate data, optimize the infrastructure, and ensure customers wouldn’t face downtime or disruptions.

The risks? Moving from single-tenant to multi-tenant is requires a significant amount of engineering effort and strategic planning. Adobe had to plan for:

• Data migration challenges: Moving enterprise clients who had been on dedicated instances for years to a shared environment without losing any of their data.

• Security: Ensuring that even though customers would now share the same infrastructure, their data remained safely partitioned and compliant with regulations like GDPR.

• Customer expectations: Keeping enterprise clients happy during the transition. These companies relied on Creative Cloud for mission-critical operations, so there was little room for error.

But the reward was worth it. By the time Adobe completed the move around 2018, they had a platform that could serve everyone—from giant corporations to individual creatives—using the same underlying infrastructure. From a long-term business and product perspective, they could push updates, roll out new features (like Adobe Stock integration and Adobe Sensei), and apply security patches to millions of users all at once.

How Long Did It Take to Make the Move?

For Adobe, the full transition from single-tenant to multi-tenant took around two years. It was a complex project that involved migrating huge amounts of data while ensuring zero disruption for their existing customers, and optimizing the shared infrastructure to handle millions of users. As you can imagine, this takes an incredible amount of careful strategic planning before technical implementation and close communication between PMs and engineers is vital.

Why this matters for Product Managers

You might not be managing a product as big as Creative Cloud, but the decision Adobe faced is one that you might run into as your product evolves. Starting with a single-tenant architecture is sometimes necessary when you’re catering to big, high-maintenance customers. They need custom features, tighter security, and more control over their data. But as your customer base grows and you start serving more users with similar needs, multi-tenant architecture becomes a smarter choice.

Here are a few things to keep in mind:

1. Cost Efficiency: Single-tenant models get expensive quickly. Every new customer means maintaining a separate environment. With multi-tenant, you’re only managing one instance of the software for all your users, saving you time and money.

2. Maintenance & Updates: In single-tenant setups, every time there’s an update or bug fix, you have to apply it to each instance. With multi-tenant, you push changes once and all customers benefit.

3. Security & Compliance: If your customers demand strict data isolation (think finance or healthcare), single-tenant can be useful. But multi-tenant has come a long way, with strong data partitioning and encryption that meets most compliance needs.

4. Scalability: Multi-tenant architecture is far more scalable compared to single-tenant architecture and was essential in order to scale Creative Cloud to millions of users without blowing up their costs. It’s easier to add more customers, and you can manage the platform without constantly juggling multiple instances.

If you’re thinking about making a similar move, expect a significant amount of effort and coordination across your product, engineering, and customer success teams. The timeline will vary based on the complexity of your system and how many customers you’re moving over, but it’s important to prepare for months, if not years, of work.

The Takeaway from Adobe’s Shift

Adobe’s decision to move from single-tenant to multi-tenant wasn’t just about saving money—it was a strategic move to ensure Creative Cloud could grow without sacrificing performance or user experience. As a product manager, the architecture you choose impacts your product’s ability to scale, your team’s ability to manage it efficiently, and your customers’ satisfaction.

If you’re finding that maintaining multiple isolated environments is becoming too costly or complex, it might be time to consider a multi-tenant approach. But if your product serves clients with specific security or compliance needs, single-tenant may still be the right choice.

The key is knowing when it’s time to switch gears, just like Adobe did.

💡 Tech Term You Should Know (TTYSK)

Data Isolation

Data isolation refers to the practice of keeping one customer's data completely separate from another's in a shared environment. In a multi-tenant architecture, where multiple customers use the same instance of software, data isolation makes sure that even though resources are shared, each customer's data remains private and secure.

Here’s why data isolation matters:

1. Security: One of the biggest benefits of data isolation is preventing data leaks or breaches. Even though multiple customers share the same infrastructure, proper isolation ensures no one can access another's data, keeping sensitive information secure.

2. Compliance: In industries like healthcare, finance, or government, data isolation is essential to meet strict regulations like HIPAA or GDPR. These regulations require data to be stored and handled in ways that prevent unauthorized access, even in shared environments.

3. Customization: Data isolation also allows customers to have customized configurations and preferences in shared systems without affecting others. For example, while many companies might use the same SaaS tool, they can still tailor the way their data is managed without impacting the broader user base.

In a multi-tenant setup, data isolation is often achieved through methods like:

• Database partitioning, where each customer’s data is stored in separate, logically isolated sections of the same database.

• Encryption to ensure that even if data were to be accessed by someone else, it would be unreadable.

• Access control mechanisms that restrict who can see or modify data.

When discussing security or compliance requirements with engineers or customers, it’s helpful to understand how data isolation works and what it protects. It plays a major role in the scalability of multi-tenant systems by ensuring that even though customers share infrastructure, their data remains private and secure.

Missed the mid-month PM & Tech Jobs Newsletter?

Looking for a new job? Our PM & Tech Jobs newsletter is issued monthly with product role job listings from senior to entry-level roles.