Problem One: Your ICP Is Defined by Your Data Provider's Filters

When you sit down to define your ICP, you default to what is available. Industry. Headcount. Revenue range. Geography. Tech stack. Funding stage.

These are the fields your database surfaces. So they become your ICP.

The problem is that your actual best customer is usually more specific than any combination of those fields captures. A company that is 51 to 200 employees in the SaaS space means something very different depending on whether they are a horizontal productivity tool or a vertical AI code review product for enterprise engineering teams. The filters treat them identically.

As your product evolves, your market shifts, and your CS team reshapes what a good customer actually looks like, the gap between your real ICP and your filter-defined ICP grows. The filters do not update. They stay set to whatever you configured when you last revisited them.

Problem Two: Your ICP Has No Dimension of Time

A company that fits your ICP today may not fit it next quarter. A company that does not fit it today may fit it in six weeks.

Companies enter and exit your ICP continuously. A startup that was too small six months ago just raised a Series A and hired their first RevOps lead. A mid-market account that was a strong fit just went through a restructure and eliminated the function your product serves.

Static ICP definitions cannot reflect any of this. They are a snapshot of who your best customer was at the moment you defined it. They have no mechanism for tracking how that changes over time or which accounts in your market are crossing into or out of fit right now.

ICP is not a fixed set of criteria. It is a function of time. ICP(t).

The TAQ Framework: A Live ICP Definition

At OpenFunnel we broke ICP into three components that reflect how fit actually works in practice.

TAQ: Traits, Activities, Qualifiers.

Each component moves at a different speed. Together they give you an ICP definition that is both precise and live.

T: Traits

Traits are the firmographic and descriptive characteristics of the company. This is the closest to a traditional ICP filter, but pushed to be more specific than a database field allows.

Not just "SaaS, 51 to 200 employees." But "AI Code Review SaaS, Enterprise only, 51 to 200 employees, US."

The description carries information that no filter captures. It tells you what the company does, who they sell to, and at what scale. Two companies can share identical firmographic filters and be completely different prospects depending on what they actually build and who they sell to.

Traits move slowly. A company's core description, market focus, and employee range do not change week to week. T is the stable foundation of your ICP definition.

A: Activities

Activities are the active pain points. The "why now."

This is what the company is doing right now that signals a buying window is open. Not what they did six months ago. Not what their job description says they care about. What is actually happening at the account today.

"Hired someone this morning to revamp their GEO strategy." "Actively scaling outbound to 2x cold-calling volume." "Posting three SDR roles in the last two weeks."

Activities are the time dimension of ICP. A moves every single day. A company can have the right Traits and Qualifiers and still be completely outside your buying window if there is no active pain point driving urgency right now. And the same company can move into your window overnight when the right activity appears.

This is the component that static ICP definitions miss entirely. There is no database filter for "what is this company actively doing right now."

Q: Qualifiers

Qualifiers are the must-haves. The conditions that need to be true for your product to actually land and deliver value.

"More than 5 SDRs." "Existing built-out RevOps function." "Series A in the last 3 months."

Qualifiers are different from Traits because they are binary gates, not descriptive criteria. A company either meets them or it does not. And if it does not, the deal either will not close or will not stick.

Sources for Q can be anything: job postings, CRM signals, funding databases, technographic tools. The qualifier defines what you need to verify. The data source is whatever gets you there most reliably.

Qualifiers move at medium speed. Slower than Activities, faster than Traits. A company crosses the 5 SDR threshold and becomes a fit. They churn Salesforce for HubSpot and fall out of fit. These changes are trackable if your data layer is watching for them.

Why This Matters for Your Data Layer

TAQ is only useful if your data layer can support it.

T and Q can be handled by a good database with regular enrichment. They change slowly enough that periodic updates are sufficient.

A cannot. Activities are live events. They exist in job postings, LinkedIn activity, hiring patterns, and market signals that change daily. A data layer that refreshes weekly or monthly cannot capture A accurately. By the time the data updates, the activity that opened the buying window may already be in the rearview mirror.

This is why ICP(t) requires continuous market monitoring, not a static filtered list. The A component of your ICP is the highest-signal and highest-decay component. It is also the one most existing tools are worst at tracking.

A live ICP definition without a live data layer is still a static list with a better name.

Frequently Asked Questions

What is an ICP in B2B sales?
ICP stands for Ideal Customer Profile. It defines the characteristics of the companies most likely to buy your product, get value from it, and stay. Traditional ICP definitions use firmographic filters: industry, headcount, revenue, geography, tech stack. A more precise definition includes a specific description of what the company does, the active conditions that indicate a buying window is open, and the hard qualifiers your product requires to land successfully.

Why do static ICP definitions break over time?
Static ICP definitions are built around available data filters at a single point in time. As your product evolves and your market shifts, the gap between your filter-defined ICP and your actual best customer grows. They also have no time dimension: they cannot reflect which companies are currently in a buying window, which have just crossed into fit, or which have exited fit since you last updated the definition.

What is the TAQ framework?
TAQ is a live ICP framework that breaks ICP into three components: Traits (firmographic and descriptive characteristics of the company), Activities (the active pain points and events indicating a buying window is open right now), and Qualifiers (the hard must-haves for your product to land). Each component moves at a different speed. Traits are stable. Activities change daily. Qualifiers change at medium pace.

What is the difference between Traits and Qualifiers in TAQ?
Traits describe what the company is: its market focus, product category, size, and geography. They are descriptive. Qualifiers are binary gates that must be true for the deal to work: a specific CRM, a minimum team size, a funding stage, an existing function. A company can have the right Traits but fail a Qualifier and still be the wrong prospect.

What are Activities in an ICP definition?
Activities are the real-time signals that indicate a company is in an active buying window right now. They are the "why now" component of ICP. Examples include a company actively scaling their outbound team, posting roles that indicate a new strategic priority, or bringing in a new executive to rebuild a specific function. Activities are the highest-decay component of ICP and the hardest to track with static data tools.

Your ICP is not a row of filters you set once. It is a description of your best customer at a specific moment in time, with a time dimension that changes as your market moves. The teams that treat it as live are prospecting against reality. The ones that treat it as static are prospecting against a definition that stopped being true the moment they wrote it down.