"Should we add AI?" – here is how to decide
I see SaaS founders either ignoring AI entirely or rebuilding their product from scratch out of fear. Neither is the right move.
Every week another think piece declares that SaaS is dead and AI agents will eat everything. Your investors are asking about your AI roadmap. Your competitors just shipped an “AI-powered” feature. And somewhere in the back of your mind, a quiet panic is building.
Here’s the uncomfortable truth: some of that fear is warranted. And some of it is pure FOMO dressed up as strategy.
The mistake I see founders make is treating “add AI” as a binary decision. They either dismiss it entirely (”our customers don’t need that”) or they pivot the whole product because they read a Sequoia deck. What they rarely do is actually think through the decision with a clear framework.
The real question isn’t “should I add AI?” It’s “where does intelligence create value in my product?”
There’s a useful way to think about this from business strategy research on switching costs and workflow lock-in. The original insight is simple: products survive disruption when they’re deeply embedded in how people work. I’ve adapted this slightly for the current AI moment, because the threat isn’t just a competitor. It’s a new category of tool that can replace workflow layers entirely.
The framework has three lenses. It’s not perfect. But it’s good enough to cut through the noise.
The Three-Lens Framework
Lens 1: Is your core value in the data, the workflow, or the decision?
Data products (CRMs, analytics tools, data warehouses) are relatively safe for now. The AI layer sits on top, it doesn’t replace them. Workflow products (project management, scheduling, document editing) are more exposed. AI agents can increasingly run workflows end-to-end. Decision products (recommendation engines, pricing tools, forecasting software) are the most exposed. This is exactly where AI is strongest.
Ask yourself honestly: what are customers actually paying for?
Lens 2: How hard is it for an AI-native competitor to replicate your moat?
Think about what actually keeps your customers from leaving. If it’s integrations and data history, you have a 3-5 year buffer. If it’s UI familiarity alone, you have maybe 18 months. If it’s “we’ve always used this tool,” you’re already losing.
An honest way to test this: could a well-funded team build your core feature set inside a general-purpose AI tool in under six months? If the answer is yes, that’s your urgency signal.
Lens 3: Does AI make your existing product meaningfully better, or just different?
This is where FOMO creates the most damage. I see founders add AI features because they feel pressure to show up at the demo with something shiny. But slapping a chatbot on a project management tool doesn’t change why customers buy. It just adds complexity.
The better question: is there a specific, frustrating part of your product that takes users significant time or judgment? That’s your AI opportunity. Not the whole product. Just that friction point.
The SaaS products that will thrive without AI features
Not every tool needs to become an AI product. Some categories are structurally resistant to AI disruption, at least for the next few years.
Compliance and audit tools. Regulated industries need traceable, human-accountable decisions. AI creates liability, not value. Vertical SaaS with deep industry workflows. Tools built for a specific niche (dental practice management, municipal permitting, independent school administration) have domain complexity that general AI models handle poorly. Infrastructure and developer tooling. When the product is about reliability and control, AI assistance is a nice-to-have, not a survival requirement.
If you’re in one of these categories, the right move is probably to watch closely and move deliberately. Not to sprint toward an AI roadmap.
The prompt: paste this into any AI tool
Instead of working through this alone, use this prompt directly. Paste it into ChatGPT, Claude, or whichever AI tool you already use. It will ask you a few focused questions and give you a grounded recommendation specific to your situation.
I'm a SaaS founder trying to decide whether I need to add AI features
to my product to stay competitive. I want you to help me think through
this clearly, without hype or FOMO.
Please ask me the following questions one at a time, wait for my answer,
then move to the next:
1. What does your SaaS product do, and who is your primary customer?
2. What is the most time-consuming or frustrating task your users do
inside your product?
3. What keeps your current customers from switching to a competitor
today?
4. Have any of your customers asked for AI features? If yes, what
specifically did they ask for?
5. Who is your most dangerous potential competitor right now, and are
they AI-native?
After I've answered all five questions, give me:
- An honest assessment of how exposed my product is to AI disruption
in the next 1, 2, and 3 years
- Whether I should build AI features now, wait, or focus elsewhere
- The single highest-value AI opportunity in my product if one exists
- What I should NOT do based on my answers
Be direct. Don't hedge everything. I'd rather have a clear opinion
I can push back on than a list of things to consider.The output won’t be a perfect roadmap. But it will tell you quickly whether you’re building from strategy or from fear.
When to actually move
If the assessment flags real exposure in the next 12 months, start a focused experiment on your highest-friction workflow. Three months, one problem, real customer feedback. Not a full rebuild.
If your moat holds up under scrutiny, the best thing you can do right now is focus. Improve reliability. Deepen integrations. Build what AI-native tools can’t easily copy: trust, compliance history, workflow depth.
The founders who’ll regret the next few years aren’t the ones who moved too slowly on AI. They’re the ones who let FOMO make the decision for them.