The Three Quotients of Agent Success

For a long time, we've known that success for humans isn't just about IQ, it’s IQ and EQ.

IQ (intelligence quotient) = how smart you are.
EQ (emotional quotient) = how well you relate to and work with others.

But as we head into the agentic world, I think we need to change that formula (atleast for agents):

Agent success = IQ × EQ × CQ

CQ = Context Quotient.

It’s how much the agent knows about YOU, your business, your goals, your constraints, your history, and your preferences.

But here's the thing: this is multiplicative, not additive. If an agent has zero context about you, it has zero value -- no matter how smart it is.

And that’s an important concept to grasp as we head into the agent era.

So today, I want to break down:

• What CQ actually means in practice

• Why context beats raw intelligence

• What this means for how we build and choose agents

—@dharmesh

What Actually Is CQ?

Context Quotient is how much relevant information an agent has access to at decision time.

Importantly, it’s not only "what data does the agent have" -- it’s "does this agent have the right context to make good decisions?"

Here are some examples of what an agent with high CQ should know:

For a sales agent:

• Your top rep's call recordings and what actually closes deals (not just EVERY transcript)

• Your pricing exceptions and why they were granted (the thought process behind decisions)

• Your lost deals and the real reasons (not just the CRM dropdown)

For a support agent:

• Your angriest detractor's NPS comments

• Your happiest promoter's renewal notes

• Your product's known issues and workarounds

For a marketing agent:

• Your brand voice and what messaging works

• Your audience segments and what they respond to

• Your campaign performance history

The point is, CQ isn't generic data you’re giving an agent. It's your data -- the accumulated wisdom of how your specific business works, and where it doesn’t.

The Thought Experiment

Here's a simple way to think about this:

You have two agents to choose from.

• Agent A: 200 IQ. Bonafide Genius. Knows nothing about you.

• Agent B: 150 IQ. Brilliant. Knows your business cold.

You pick Agent B, every single time.

Why?

Because raw intelligence without relevant context is just confident guessing. (Also known as "management consulting"). 😀

Agent A might give you a technically perfect answer, but there’s a good chance that it would be completely wrong for your situation, as it doesn't know your specific constraints, history, exceptions, or your actual goals.

Agent B, on the other hand, gives you an answer that actually works -- because it knows what "works" means in your world.

This is also why you can't just plug in the smartest AI model and just expect results. The model might be incredibly capable, but if it doesn't understand your business context, you get no value.

This is also why I keep emphasizing that this formula (Agent success = IQ × EQ × CQ) is multiplicative, not additive. If CQ goes zero, so does the result.

What This Means for Building Agents

The conversation right now is heavily focused on IQ. Everyone's watching which model scores highest on benchmarks, which one has the best reasoning, or which one can handle the most complex tasks for the longest runs.

…And all of that matters… We do need more capable models because better reasoning will enable better agents.

But I think the real differentiation in day-to-day agent tasks is going to come from this concept of CQ.

Because model capabilities are converging, all major AI companies are building incredibly capable systems, narrowing the IQ gap.

The CQ gap, meanwhile, is wide open.

Building high-CQ AI agents would mean:

1. Capturing the right context. Why was that exception cleared? What precedent does it set for the future?

2. Maintaining context over time. Agents need to learn from interactions and improve at understanding your specific needs.

3. Connecting context across systems. Your business is distributed across systems (CRMs, analytics tools, more). High-CQ agents need access to all of it.

This is what we're furiously working on at HubSpot. As frontier labs continue to build highly capable models (high IQ), we’re focusing on building agents that deeply understand customer context.

As founder/CTO of HubSpot, you can call me biased, but context is particularly critical in go-to-market, and unified platforms like ours that connect marketing, sales, and service have a structural advantage here.

Having that said, when you're evaluating agents -- whether you're building them or buying them for your business -- don't just ask "how capable is the underlying model?".

Always ask:

• How much context does it have access to?

• How does it learn about my specific business?

• Can it remember decisions and improve over time?

• Does it connect to the systems where my actual context lives?

Because in the agentic age, the agent that understands your business is more likely to win.

There’s also the other dimension explore in this -- around EQ -- but that's for another post.

For now, if you're building agents, how are you approaching the context quotient? If you're evaluating them, what signals are you looking for?

Let me know. We’re all still figuring this out, and I'd love to hear what's working (or not working) for you.

—Dharmesh (@dharmesh)