How to Set Up Customer Enrichment and Segmentation in HubSpot

Most companies pour their CRM effort into winning new customers and then run the ones they already have on autopilot — answering tickets as they come, and finding out an account is unhappy only when the cancellation email lands. Meanwhile, perfectly good customers who'd happily buy more go unnoticed because nobody has time to spot them.

The fix isn't more manual account reviews. It's teaching HubSpot to watch your customer base for you: enriching records automatically, scoring every account on how well it fits, how engaged it is, how healthy the relationship is, and how much it's adopted — then sorting customers into tiers and segments your team can actually act on.

Here's what customer enrichment and segmentation involves in HubSpot, why it's one of the highest-leverage things you can do with an existing customer base, and where we've learned to focus.

What "enrichment and segmentation" actually means

Think of it as three connected layers.

Enrichment keeps your customer records complete and current — filling in firmographic details and surfacing signals from how customers are interacting with you, without anyone typing them in. HubSpot's AI can also research accounts, summarize long histories, and flag intent signals buried in calls and emails, turning your CRM from a passive database into something that surfaces what matters.

Scoring turns that data into decisions. Instead of eyeballing accounts, you build scores that rank every customer on a few dimensions that matter — and set thresholds so "at risk" or "ready to grow" is an automatic status, not a gut feeling.

Segmentation makes it operational. Customers sort themselves into tiers and segments that update automatically, so your team, your automation, and your leadership dashboard are always working from the current, correct view.

The four scores worth building

A strong customer intelligence model rests on four scores, and each answers a different question.

Fit — how closely an account matches your ideal customer, based on things like size, industry, geography, and what they hold. It answers: is this the kind of customer we want more of?

Engagement — how actively a customer interacts with you, drawn from behavior like email activity, site visits, portal logins, and survey responses. It answers: are they leaning in or drifting away?

Health — the state of the relationship, built from customer sentiment (surveys) and a simple red/amber/green status your team keeps current. It answers: are they happy, or are we about to lose them?

Product adoption — how much of what you offer a customer actually uses. It answers: is there room to grow this account?

Read together, these four scores tell you exactly who to protect and who to grow.

Why this pays off

When your customer data enriches and scores itself, the wins show up where they matter most — retention and expansion:

At-risk accounts get flagged before they churn, so your team can intervene while it still counts
Warm upsell candidates surface on their own: good-fit, happy customers who simply haven't adopted much yet
Your team runs the whole customer base from one dashboard instead of scattered spreadsheets and gut feel
Customers stay sorted into current, correct tiers and segments without anyone maintaining a list
Your CRM does the watching, so people only spend time on the accounts that actually need it

Retention and expansion are where profit compounds in a B2B business. This is how you manage them as a system instead of a scramble.

Where teams go wrong

This is where strategy earns its keep. A few patterns we see constantly:

Building scores on data that isn't there. A score is only as good as the data feeding it. Build a "product adoption" score before the ownership and revenue data is actually in the CRM, and it quietly reads zero for everyone — and the team stops trusting the whole system. Data has to come first.

Treating health as a toggle instead of a habit. A health score that relies on a red/amber/green status only works if the team actually keeps that status current. Without a simple rhythm — set it at onboarding, revisit it when closing a ticket, review it yearly — the score drifts out of date and starts lying.

Alert overload. It's tempting to build a dozen triggers. But a system with two alerts people act on beats one with twenty they've learned to ignore. Start with one upsell trigger and one churn-prevention trigger, earn trust, then expand.

Ignoring the cost of AI. Enrichment and AI research run on credits. Flip everything to "all accounts, all the time" without a plan and the bill surprises you. Scoping which records enrich and research is part of doing this well, not an afterthought.

How we approach it

We treat this as building a retention-and-growth system, not filling in a settings page. Before touching the platform, we get clear on what a great customer looks like to you, which signals matter, and where the underlying data lives — because that's what determines whether the scores are real or aspirational.

Then we build in dependency order: enrichment first, so the data exists; the four scores next, tuned to your definition of a good customer; the health inputs and review rhythm; then the two triggers that actually alert your team. Finally, we sort everything into tiers and segments and stand up a single dashboard your CSMs and leadership can run the business from. Along the way we keep the AI usage scoped, so the intelligence scales without the cost running away.

The goal isn't a screen full of scores. It's a customer base that tells you who to save and who to grow — before you'd have noticed on your own.

Common questions

Do we need a certain HubSpot tier for this? Yes — scoring, surveys, and the supporting properties require Professional or higher, and some of the AI-built and multi-model pieces are Enterprise features. The enrichment and AI research also use HubSpot credits. We confirm all of this up front so there are no surprises.

Isn't scoring just for leads? That's the common assumption, and it's a costly one. HubSpot's scoring works just as well on your existing customers as it does on leads. Scoring the customers you already have — for fit, health, and growth potential — is one of the most underused, high-value things you can do with your CRM.

What if our product and revenue data lives in another system? Common, especially in payroll and HCM. We look at how to bring that data into HubSpot so the adoption and value scores are real. If it can't be connected yet, we're upfront about it rather than shipping a score that reads zero.

If you want to learn more about how to get customer enrichment and segmentation right in HubSpot, contact The Gist.