Common ICP Mistakes: Too Broad, Aspirational, Outdated
Most ideal customer profiles fail in one of three ways: they are defined too broadly, they are built aspirationally instead of from evidence, or they are allowed to go stale. Each mistake is quiet — the ICP still exists, still looks reasonable on a slide — but each steadily points the entire program at the wrong accounts. Knowing the failure modes is the fastest way to build a profile that holds up.
These traps come up so consistently that the standard remedy is always the same: ground the profile in data and keep it current. This article catalogs the mistakes and the fix for each. For the underlying method, see how to build a B2B ICP and how to validate it using closed-won data.
Mistake 1: Too broad
The most common error is an ICP that excludes almost no one. "B2B companies in North America that need better marketing" is not a profile — it is the market. An over-broad ICP defeats the entire purpose of account-based marketing, which is concentration: pointing finite effort at the accounts most likely to pay off. When the list is everyone, reps fall back on personal preference and the program degrades into ordinary lead generation with an ABM label.
The fix: narrow until the criteria genuinely disqualify a large share of companies. Add specificity on vertical, size band, and at least one technographic or trigger criterion. A focused list you can actually engage beats a sprawling one you only nominally target — the same discipline that governs target-list size in the ABM strategy framework.
Mistake 2: Aspirational, not evidence-based
The second trap is defining the ICP around the customers you wish you had — marquee logos, glamorous verticals, enterprise deals — rather than the ones your data shows you win and keep. Aspiration feels motivating and is almost always wrong: the dream accounts often have longer cycles, lower win rates, and worse retention than the unglamorous segment quietly driving your revenue.
The fix: build from closed-won analysis. Pull the accounts you have actually won and retained profitably and abstract their shared traits, as described in validating your ICP using closed-won data. If the evidence contradicts the aspiration, trust the evidence — or at minimum run the aspirational segment as a separate, explicitly experimental tier rather than letting it dominate the core profile.
Mistake 3: Outdated and stale
An ICP set once and never revisited slowly diverges from reality. Products evolve, markets shift, new competitors reshape a segment, and the profile that was sharp 18 months ago now quietly mis-targets. Stale technographic criteria are especially insidious — "targets companies running [legacy tool]" can become actively misleading as the market migrates.
The fix: treat the ICP as a living model on a quarterly review cadence, with every new closed-won and churned account fed back in. Owned jointly by marketing, sales, and RevOps, the profile should get sharper over time rather than ossifying.
Mistake 4: Confusing the ICP with a persona
A subtler error: writing a profile that describes a person (a job title and a set of pains) when it should describe a company. This corrupts account selection because you end up filtering on the wrong unit. Keep the two distinct — the difference, and why you need both, is covered in ICP vs. buyer persona.
Mistake 5: Ignoring intent and timing
Finally, many ICPs capture only fit (who the company is) and omit behavioral signals (whether it is in-market now). The result is a list of well-matched accounts with no sense of priority. Add at least one intent or trigger source so the profile distinguishes good-fit-now from good-fit-someday — the three-layer approach in firmographic, technographic, and behavioral criteria explained.
Frequently asked questions
How specific should a B2B ICP be?
Specific enough that a meaningful share of companies are excluded. If your ICP would let in most of your total addressable market, it is too broad to guide account selection. A useful test: could two reps independently apply it to the same list and largely agree on which accounts qualify?
How often should you review your ICP?
On a quarterly cadence at minimum, and immediately after any major product, pricing, or market shift. Feed every new closed-won and churned account back into the analysis so the profile sharpens rather than drifts.
Can you have more than one ICP?
Yes, if you genuinely sell distinct products to distinct markets — but each should be a separate, validated profile, not a single vague profile stretched to cover everything. Most teams need fewer ICPs than they think.
What is the difference between an ICP being too broad and being aspirational?
Too broad means the criteria exclude too few companies. Aspirational means the criteria describe the customers you wish you had (big logos, dream verticals) rather than the ones your data shows you actually win and retain. A profile can be both at once.
Avoid these five and the profile that results is narrow, evidence-based, current, company-level, and timing-aware — exactly the foundation the rest of the ABM strategy framework is built on.