Last updated: April 2026

Every business says it wants better customers.

More profitable accounts. Longer relationships. Higher lifetime value. Fewer difficult clients who consume resources and generate complaints.

The problem is not that companies lack ambition on customer selection. The problem is that the data they use to make those decisions is frequently wrong, incomplete, or inconsistent. And when your data is broken, your segmentation is broken, no matter how sophisticated your framework.

Robert Simons of Harvard Business School developed a four-step framework for customer selection that remains one of the clearest frameworks available for thinking through this problem. But Simons assumed your data was reliable. In practice, that assumption does a lot of heavy lifting.

Here is the framework, with the data quality dimension built in at every step.

Step 1: Define Your Priority Customers

The first step is deceptively simple: decide who your best customers actually are. Not who you wish they were. Not who you initially targeted when you launched. Who, based on actual evidence, generates the most value for your business with the least friction?

This requires you to look honestly at your customer data across several dimensions:

Revenue and margin by account, not just top-line revenue. A client who pays a large retainer but requires disproportionate support hours may be far less valuable than a smaller account that runs smoothly.
Retention and renewal patterns. Customers who stay are almost always more valuable than customers who churn, even if the churning customers initially looked attractive.
Referral behavior. Which customers send you new business? That referral value rarely appears in a CRM but it is real and significant.
Strategic fit. Which customers align with where you want to take your practice? An account that is profitable today but pulls you away from your core positioning has a hidden cost.

The data quality trap at Step 1 is assuming your CRM or accounting system reflects reality. In most organizations it does not. Revenue data is often incomplete, margin data is rarely tracked at the account level, and retention history is frequently buried in spreadsheets rather than systems. Before you can define your priority customers, you need to trust the data you are using to define them.

Step 2: Understand What Your Priority Customers Value

Once you know who your best customers are, the next question is what they actually value about working with you. This is where many businesses make a critical error: they assume they already know.

They are usually wrong.

What customers say they value in a sales conversation is frequently different from what actually drives their loyalty and referral behavior. A client may say they hired you for your technical expertise. What actually keeps them renewing is that you translate complex findings into language their leadership team can act on. Those are not the same thing, and conflating them leads to a service delivery model and a marketing message that misses the real value driver.

Getting this right requires primary research. Structured client interviews, not just satisfaction surveys, are the most reliable method. Ask your best clients directly: what would they lose if they stopped working with you? What do you do that their previous provider did not? What have you helped them accomplish that they could not have done without you?

The data quality trap at Step 2 is over-relying on survey data without validating it against behavioral evidence. If clients say speed of delivery is their top priority but your fastest turnaround accounts have the same retention rate as your slowest, speed may not be the real driver. Cross-reference what clients say with what they actually do.

Step 3: Assess Your Capabilities Against Customer Needs

With a clear picture of who your best customers are and what they genuinely value, Step 3 asks a harder question: are you actually well-positioned to serve them better than your competitors?

This is a capabilities audit. It requires honest assessment across three dimensions:

Distinctive competence. What do you do demonstrably better than alternatives? This is not a list of services. It is a specific answer to why a well-informed buyer would choose you over a capable competitor.
Operational capacity. Can you reliably deliver at the level your priority customers expect, at scale? A capability that depends on one person or one process that cannot be replicated is a fragile competitive advantage.
Credibility signals. Can you demonstrate your capability in ways that are visible and verifiable to a prospect who does not yet know you? Case studies, measurable outcomes, and third-party validation all matter here.

The data quality trap at Step 3 is relying on internal perception rather than external evidence. Your team may believe you are the strongest option in your market. But if your website, your case studies, and your proposal materials do not make that case clearly and credibly, your self-assessment is not doing you any good.

Step 4: Develop a Customer Selection Discipline

The first three steps are analytical. Step 4 is behavioral, and it is where most businesses fail.

Customer selection discipline means consistently applying what you learned in Steps 1 through 3 to actual business development decisions. It means having the confidence to decline work that does not fit your priority customer profile, even when the revenue is tempting. It means updating your ideal customer profile as your data improves and your market evolves. And it means making sure everyone in your organization who touches business development is working from the same definition of who you are trying to serve.

In practice this requires:

A documented ideal customer profile that is specific enough to be useful. “Mid-size B2B companies” is not a customer profile. “Professional services firms with 25 to 150 employees, an active digital marketing function, and a history of investing in analytics” is closer.
A regular review cadence. Your best customer profile should be revisited at least annually. Markets change, your capabilities evolve, and the customers who were ideal three years ago may not be the right target today.
Consistent application in business development. If your ideal customer profile says something clearly but you regularly make exceptions for revenue reasons, your profile is not actually guiding decisions. The exceptions are.

The data quality trap at Step 4 is treating customer selection as a one-time exercise. It is not. It is an ongoing discipline that depends on continuously improving data about who your best customers are, what they value, and how well you are positioned to serve them.

The Data Quality Foundation

Simons’ framework is sound. But every step depends on data you can trust. If your CRM is incomplete, your margin data is unreliable, your client feedback is anecdotal, and your capability evidence is thin, you will get imprecise answers at every step.

The investment in data quality is not a separate initiative from customer strategy. It is the foundation that makes customer strategy possible.

If you are not sure whether your data is reliable enough to support these decisions, that is a conversation worth having.

Pick the Right Customer