Creation as a Service Will Eat SaaS (And Slack Shows Us How)
For the last decade, Software as a Service (SaaS) has been the default answer to almost every business problem. Need CRM? Buy SaaS. Need collaboration? Buy SaaS. Need analytics? Buy SaaS.
It worked — until it didn’t.
The uncomfortable truth is that most SaaS products are built for the lowest common denominator. They solve generic problems well, but they are terrible at adapting to how your organization actually works. You end up bending your processes to fit the software, instead of the software working for you.
That is why I believe Creation as a Service will eventually beat Software as a Service.
Not because SaaS is useless — but because static software cannot keep up with dynamic organizations.
The Real Problem With SaaS: It Freezes Your Process in Time
SaaS products encode decisions:
- how workflows should look
- how data should flow
- what is “normal usage”
This is fine when your business is simple. It breaks when:
- your workflows evolve fast
- your internal logic is unique
- your organization differentiates on process, not just output
Most SaaS vendors sell you their idea of how your company should work. Over time, you either:
- accept the friction, or
- build painful workarounds, or
- leave.
In the age of AI, this friction becomes unacceptable.
If software can be created and modified cheaply, why should my business be constrained by someone else’s assumptions?
Creation as a Service: Platforms Over Products
What I want is not more features.
I want infrastructure, primitives, and platforms.
Creation as a Service means:
- the product gives you building blocks, not finished opinions
- you define workflows
- you encode your own logic
- you extend the system instead of fighting it
The value is not in the UI screens.
The value is in:
- extensibility
- composability
- ownership of behavior
This is the difference between:
- using software
- and owning how software behaves inside your company
AI is the accelerant here. It collapses the cost of customization. Suddenly, building internal workflows and agents is cheaper than forcing your team to adapt to rigid SaaS UX.
Why Slack Is the Perfect In-Between
I have zero interest in building my own Slack.
And anyone who thinks rebuilding Slack is “easy” has never shipped infrastructure-grade software.
Slack works because it sits perfectly between:
- finished product
- and programmable platform
You get:
- identity
- real-time messaging
- reliability
- permissions
- scale
And on top of that:
- bots
- events
- APIs
- workflows
- agents
Slack becomes a control plane for how work flows inside a company.
You do not rebuild communication.
You extend it.
That is the sweet spot:
Buy commodity complexity. Build competitive advantage.
Slack-Based Agents Are the Real Operating System
When your company runs on Slack-based “@agents”, something fundamental changes.
Slack stops being “chat”.
It becomes:
- a human + AI interface layer
- a workflow trigger surface
- a decision hub
- a control plane for execution
At that point:
- tools are no longer destinations
- they are capabilities invoked from where work already happens
This is not “AI features in SaaS”.
This is software becoming programmable labor.
That distinction matters.
Why Most SaaS Is in Trouble
Many SaaS companies:
- lock workflows into UI flows
- monetize per seat
- resist deep extensibility (because it breaks their product vision)
AI changes the economics:
- building custom logic is now cheap
- internal platforms become viable
- rigid SaaS becomes a bottleneck
The future is not:
SaaS + a few AI buttons
The future is:
Infrastructure + primitives + workflows + agents
If your product cannot be extended deeply, it will eventually be replaced by something that can.
Salesforce + Slack: A Once-in-a-Generation Opportunity
Salesforce has always struggled with:
- UX complexity
- configuration fatigue
- heavy interfaces
Slack solves the human interface problem.
If Salesforce is smart, Slack becomes:
- the front door
- the daily interface
- the agent control plane
And Salesforce becomes:
- the system of record
- the policy layer
- the data engine
That is not a feature upgrade.
That is a platform re-architecture.
If Salesforce fails to do this, someone else will turn Slack into the control plane for business systems — and Salesforce becomes a backend provider in someone else’s ecosystem.
What People Often Miss (Important Considerations)
1. Governance Becomes Product, Not Compliance
When everything is programmable:
- permissions
- audit logs
- blast radius
- approval flows
must be first-class primitives.
Without governance, Creation as a Service becomes chaos.
2. UX for Builders Is the New Moat
Platforms fail when:
- extension is painful
- docs are unclear
- developer experience is hostile
If building agents and workflows feels harder than tolerating bad SaaS UX, people will go back to SaaS.
3. Cost Control Becomes Strategic
AI-powered workflows introduce:
- hidden compute cost
- token spend
- runaway automation
Your platform must expose:
- cost observability
- quotas
- budget controls
Otherwise, finance will kill your platform adoption.
4. Security Boundaries Must Be Designed, Not Bolted On
Agents can:
- read data
- trigger actions
- call external systems
Without strong boundaries, you create:
- massive breach surfaces
- compliance nightmares
Security is not an afterthought in a programmable company.
It is part of the product.
5. Platform Lock-In Replaces Vendor Lock-In
The risk shifts:
- from being locked into SaaS
- to being locked into platform architecture
If your internal workflows depend on one platform’s primitives, switching becomes hard.
The trade-off is worth it — but it must be acknowledged.
The New Winning Model
The future winners will not sell workflows.
They will sell the ability to create workflows.
They will not sell features.
They will sell capabilities.
They will not ship opinions.
They will ship infrastructure with product-grade UX.
That is the real evolution:
From Software as a Service
to Creation as a Service
SaaS is not dead.
But rigid SaaS is already obsolete.
The future belongs to platforms that let software finally work for people — instead of people working around software.
Comments ()