When Should You Create Separate ICPs?

TL;DR. Create separate ideal customer profiles when customer groups require materially different buying journeys, product capabilities or routes to market. Keep a shared ICP when differences can be handled through segment tags, messaging or qualification rules within the same approach. Before splitting, gather evidence that the distinction changes how you attract, qualify, sell to or serve customers. Ranking segments decides where to focus; separate ICPs define different kinds of customers you are prepared to serve.

What makes a difference large enough to justify another ICP?

An ICP should help your team make decisions about which organizations fit your business and what serving them requires. If you need a starting definition, use How to Build an ICP.

For this decision, treat a difference as material when ignoring it would produce the wrong action: sending an account through an unsuitable purchase path, promising capabilities it needs but you cannot provide, or assigning a channel that cannot support its purchase.

Ask:

Would we make a meaningfully different decision if we knew which group this account belonged to?

Different industries, company sizes or locations are reasons to investigate. They are not automatic reasons to split. Conversely, organizations in the same industry may need separate profiles if their operating requirements demand different products and purchase paths.

Keep the distinction between organizations and people clear. Different users, champions and budget holders can belong within one ICP. B2B User Persona vs Buyer Persona explains that separate question.

Should you keep one ICP or create separate profiles?

Use this checklist to compare the groups. Record the actual decision each difference changes, rather than simply counting answers.

Decision area	Keep a shared ICP when…	Consider separate ICPs when…
Customer problem	Groups need the same outcome, with different examples or terminology	The problems require different success criteria and solutions
Buying journey	The same stages and proof work, with minor adjustments	Groups require different evaluation, approval or contracting paths
Product capabilities	A common product meets essential needs	Each group needs distinct capabilities to reach value
Route to market	The same channels and selling approach can serve both	One requires a substantially different way to reach and sell to accounts
Onboarding	The same process works with optional steps	Setup, implementation responsibilities or support commitments differ materially
Qualification	Shared fit rules work with segment tags	An attribute signals good fit in one group and poor fit in another

Do not turn this into an automatic scoring formula. One essential product requirement may justify a split even when other answers match. Several cosmetic differences may justify none.

Write the proposed boundary in one sentence:

“Accounts belong to this profile when they need ___, buy through ___, and require ___ to succeed.”

If the sentence only changes an industry label, try a shared ICP with segment notes first.

What evidence should you collect before splitting?

Build a side-by-side evidence sheet using customer interviews and records from actual opportunities. Include wins, losses, stalled evaluations and customers who struggled after purchase.

Collect evidence in these areas:

• Problem and trigger: What prompted the search? What outcome did the customer need?
• Buying journey: Who evaluated, approved and signed? What proof did each person require?
• Product requirements: Which capabilities were essential? Which were preferences?
• Route to market: How did the account find you, and what helped it progress?
• Delivery effort: What setup, training and support did the customer require?
• Results: Did the customer reach the agreed outcome, continue using the product and remain worth serving?

For each observation, record its source, the account context and whether it appeared elsewhere. Separate what customers said from what they did.

For example, “needs onboarding help” is too vague. Record whether the account needed a short explanation, data preparation or an implementation project. Those imply different decisions.

Use Win-Loss Analysis for B2B to investigate purchase decisions. Compare those findings with post-purchase experience before defining a profile around customers who buy readily but cannot succeed.

If evidence is limited, call the proposed profile a hypothesis. State what would support it and what would disprove it.

How is this different from ranking customer segments?

Segment ranking asks where to invest. ICP separation asks whether different customers need different approaches.

You can rank several segments within one ICP. For example, two industries might share the same problem, purchase path and product requirements, while one deserves more attention because you can reach it more easily.

You can also define separate ICPs while actively targeting only one. The second profile might describe a valid future opportunity whose product or delivery requirements you cannot currently support.

Keep two decisions distinct:

• Profile decision: Does this group need a separate definition and approach?
• Priority decision: Should we invest in serving it now?

Use Prioritize B2B Customer Segments for the second decision. Do not create another ICP merely to express that one segment ranks below another.

What would a justified split look like?

Consider this illustrative example.

A workflow software business serves independent teams and organizations coordinating work across several departments. Both want clearer task ownership.

Independent teams can evaluate the product themselves, invite colleagues and buy through a standard purchase flow. The larger organizations require central administration, access controls, an agreed rollout and approval from several stakeholders.

A split would be reasonable if evidence confirms that these requirements are essential and that the business intends to support both approaches.

Element	Independent-team profile	Organization-wide profile
Evaluation	Team tests its own workflow	Several stakeholders assess a shared rollout
Essential capabilities	Task ownership and collaboration	Central administration and access controls
Purchase path	Standard purchase flow	Assisted evaluation and negotiated agreement
Onboarding	Team sets up its workspace	Assigned rollout responsibilities and training

The boundary is not simply “small versus large”. It is whether the purchase and deployment require a different approach.

If larger accounts use the same product and purchase flow, a size tag may be enough. If the business cannot support organization-wide deployment, the immediate decision may be to disqualify those accounts rather than launch another ICP.

What does each additional profile cost to maintain?

Before approving another ICP, list the work needed to make it usable.

Depending on the distinction, that may include separate qualification rules, messaging, demonstrations, sales materials, onboarding instructions, reporting and product commitments.

Assign an owner to each item. Ask what existing work will receive less attention to make room.

A profile also needs upkeep when evidence changes. Review it when customers repeatedly cross its boundary, new capabilities remove an old distinction, or a route to market changes.

Keep a compact record:

• The boundary between this profile and the others.
• The evidence supporting that boundary.
• The decisions that change because of it.
• The person responsible for maintaining it.
• The conditions for merging, retiring or revising it.

Approve the split when its decision value justifies that work. A document that nobody uses does not establish a useful customer strategy.

What are the common mistakes?

Splitting by labels alone. Different industries or locations may need different examples without needing separate profiles. Test the buying and delivery implications.

Treating every buyer role as an ICP. A user and a procurement approver can be part of the same account. Map their roles within the buying journey.

Building around one unusual request. Investigate whether it reflects a repeatable customer need or a single account's circumstances.

Confusing custom work with a target market. A request becomes a profile hypothesis only when you deliberately consider serving similar accounts.

Using a split to avoid a priority decision. Defining more profiles does not create capacity to serve them. Choose what receives investment.

Keeping profiles after their purpose disappears. Merge them when shared capabilities and purchase paths make the distinction unnecessary.

What should you do before approving another ICP?

Complete this final check:

• We can describe the boundary using observable account characteristics.
• Evidence supports a different buying journey, capability requirement or route to market.
• We know which decisions change for each profile.
• We tested whether segment tags or optional steps would be sufficient.
• We have chosen whether to serve both profiles now.
• Owners can maintain the required materials and processes.
• We have defined what would make us merge or retire the profiles.

If the distinction remains uncertain, keep a shared ICP and record the suspected difference. Gather evidence through the next relevant customer conversations before committing to a separate approach.

What else should you know about multiple ICPs?

Related reading

• How to Build an ICP — define the customer characteristics that guide your decisions.
• Prioritize B2B Customer Segments — decide where to focus your resources.
• B2B User Persona vs Buyer Persona — distinguish account fit from individual roles.
• Win-Loss Analysis for B2B — examine evidence from real buying decisions.