For a long time, Braze represented the pinnacle of CRM.
Segmentation, journeys, conditions, A/B tests. If you mastered the UI, you could run “data-driven marketing” at scale.
But the moment AI stopped being a feature and became a decision-maker, that entire model started to crack.
Truth be told, as recently as a year or two ago, I was the one leading and directing projects to adopt Braze or migrate to similar CRM tools, serving as a technical or business leader across multiple companies. However, the paradigm has shifted. We no longer need expensive UIs to draw boxes; we need reasoning engines that understand the soul of our data.
Every classic CRM platform is built on the same premise:
Humans define the rules.
If the user does X → send message Y
If they don’t react → wait Z days → try again
Segment first, then communicate
This worked when human thinking speed was the bottleneck.
AI flips the premise entirely:
Rules don’t need to be designed. They can be inferred.
With AI:
Segments don’t need to be pre-defined
Conditions don’t need to be exhaustively listed
Hypotheses don’t need to be manually tested first
Instead, the system asks — continuously:
“Given everything we know right now, what is the most plausible message for this user at this moment?”
A workflow UI cannot answer that question. A reasoning system can. (That's why Google's gemini can beats ... )
Braze is excellent at orchestration:
When to send
Through which channel
To which pre-defined audience
But in an AI-native world, orchestration is a solved problem — a lower layer.
The real leverage has moved upstream:
How is the message created?
What narrative frame should be used?
Which data points actually matter this time?
With AI:
Content is not static — it’s generated
Campaigns are not predefined — they’re assembled
Analytics are not reports — they’re inputs to the next decision
You don’t “add AI” to this model. You replace the model.
Here’s the uncomfortable truth for CRM SaaS vendors:
They don’t truly understand your data.
They see events. They don’t see meaning.
They lack:
Domain-specific context
Product intuition
The stories behind the numbers
When you run AI internally:
You feed it raw domain data
You encode your definition of success
You let it reason directly on SQL outputs
At that point, CRM stops being a tool. It becomes an extension of your product’s cognition.
Asking “Can Braze do this?” starts to sound like “Can Excel do this?”
Technically, maybe. Practically, not at the required speed or depth.
Once AI is in the loop, the very concept of CRM changes.
No more:
Campaign calendars
Journey diagrams
Static segments
Instead:
Data pipelines
Objective functions
A system that keeps asking better questions
“Is this the right message?” “Is this the right tone?” “Is this even the right moment to speak?”
Braze can only ask these questions inside pre-drawn boxes.
AI doesn’t need the boxes.
Braze is a near-perfect answer to an old question:
How do humans manage messaging at scale?
AI introduces a new question:
How does the system decide what to say, by itself?
In teams that fully adopt AI, CRM stops being:
A subscription
A dashboard
A workflow builder
It becomes:
An internal system
A thin layer of code
Sitting on top of data + models
That’s why the real competitors to CRM tools today aren’t other CRM tools.
They’re:
Data pipelines
LLMs
And teams willing to let machines think
In that landscape, many CRM platforms will start to feel like expensive, slow UIs for decisions AI can already make better on its own.