Keeping a CRM database properly segmented can be difficult, especially when the process relies on people entering and maintaining the data manually. Definitions change. Teams change. People interpret fields differently. New records come in without the right information, and values that made perfect sense when they were first created become harder to understand a year later.
Eventually, you end up with a situation where nobody is sure what the categories mean, whether they are still accurate, or whether new records are being classified consistently. That matters because segmentation sits behind so much of what happens in HubSpot. It affects who you market to, how leads are prioritised, how sales teams organise their accounts and what appears in reports.
Before, segmentation beyond standard fields such as industry, revenue, employee count and country often relied on external data enrichment tools. Some of these tools can have a steep learning curve to set up properly, and using them to segment a large database can quickly become expensive.
Now, HubSpot Smart Properties can fill in the missing gaps natively within HubSpot.
Smart Properties are custom CRM properties that use HubSpot’s Data Agent to research or interpret information and populate the property automatically. You define what you want to know, how the answer should be structured and where the information should come from.
Depending on the use case, Smart Properties can use web research, the company’s website, existing CRM properties, or activities such as calls, emails, meeting transcripts and notes. This makes them particularly useful when the information you need for segmentation is more specific than what standard enrichment fields can provide.
Smart Properties can access information available on the internet as well as data in the CRM.
Web research: data from searching the internet. This data is pulled using the Company Domain Name property value for companies and First name, Last name, Email, and associated company domain name for contacts. Contacts without an associated company pull from the name and the email.
Company website: data from the company's primary domain (e.g., hubspot.com). This data is pulled used the company domain name property value for the company record.
Property data: data from another property on the same record. Only one property can be selected. If you select a file property, it will use the first uploaded PDF file.
Activities and transcripts: data from logged calls, emails, meeting transcripts, and notes associated with the record.
Smart Properties are useful anywhere you need information that is more specific, contextual or interpretive than a standard CRM field.
Smart Properties are most useful when you need information that is specific to your business and difficult to capture reliably through standard CRM fields alone. Because you can define both the question and the data source, they can be used for much more than simple enrichment, from researching account fit to turning conversations into structured CRM data.
HubSpot itself highlights use cases such as ICP enrichment, lead qualification, strategic-priority research and call transcript analysis. Here is my simple getting started list.
ICP and target account segmentation: Classify companies against your own criteria, such as industry fit, market fit, business model or strategic relevance.
Persona classification: Turn inconsistent job titles into clearer categories such as Marketing Leader, Sales Leader, Founder or Operations.
Company type and business model: Identify whether a company is, for example, SaaS, ecommerce, marketplace, agency, consultancy or another type relevant to your business.
Technology and platform usage: Research whether a company appears to use a certain technology, platform or competitor product.
Research based on CRM activity: Use information from calls, emails, notes or transcripts to surface themes, needs or other useful context from existing interactions.
Data clean-up and standardisation: Use AI to classify inconsistent values into a controlled set of options that are easier to use in reporting and automation.
HubSpot provides several example prompts that show the types of research Smart Properties can handle in practice, with some additional use cases.
ICP segmentation is one of the strongest use cases for Smart Properties because the criteria that define a good-fit company are rarely captured neatly in standard CRM fields.
Industry, company size and location can be useful starting points, but most businesses also care about things that are more specific to how they sell, such as business model, market focus, product fit or other strategic relevance.
Smart Properties let you turn those criteria into structured CRM data. Instead of relying on manual research or broad enrichment categories, you can ask HubSpot to research each company against your own definition of fit and store the result in properties that can then be used in lists, scoring, reporting, workflows and sales prioritisation.
Is this someone you can sell to?
The broader goal of CRM segmentation is to agree on the company and contact profiles that actually matter to the business, then use those definitions to shape what happens in the CRM. Companies and contacts that match your target profiles should follow a deliberate, prioritised lifecycle journey, with the right nurturing, scoring, routing and sales activity around them. Records that are clearly outside your target market, for example companies you cannot realistically sell to, should follow a separate path so they do not distort reporting, scoring or marketing activity.
Start by defining what actually makes a company a good fit. For some businesses, that might be geography, industry or company size. For others, it could be the technology they use, their business model or another firmographic characteristic.
The important thing is to identify the factors that genuinely affect whether a company is worth targeting.
Next, decide how you want the information to be grouped and labelled. The categories should use language that makes sense internally and that everyone can understand consistently.
That might be grouping companies to Small Business, Mid-Market and Enterprise in broader industry categories, for example, such as Technology & SaaS, Financial Services, Professional Services, Retail & Ecommerce, Healthcare, Education.
The goal is to create a simple, controlled classification system that makes sense in this particular case.
Once the company fit is clear, move to the people within those accounts. Here, useful classifications might include persona or functional area, seniority, and for example, buying role.
For instance, you might group contacts as Marketing, Sales, Operations, HR or Finance, then add a simple buying-role classification such as Decision Maker, Influencer, Champion, User or Other. You could also use seniority levels like C-suite, VP/Director, Manager and Individual Contributor.
The aim is to make it easy to understand not just whether the company is a good fit, but whether the person is relevant to the buying process.
Now that you’ve defined the ICP universe that matters to your business, sales and marketing processes, it’s time to turn those definitions into CRM properties.
There are a few ways to approach this, but if you want to keep control over both the property values and HubSpot credit usage, a more conservative setup can work well.
Rather than letting AI write directly into your core segmentation properties, use Smart Properties as the research layer. Review or process those outputs first, then use workflows or controlled logic to update the operational properties.
In practice, this means creating both the final segmentation property and a Smart Property that feeds into it through workflows.
Why use this approach? Mainly because it gives you more control.
You can decide which parts of your ICP segmentation universe should be populated first, rather than running AI research across everything at once. You can also manage HubSpot credit usage more carefully and avoid spending credits on information you may already have from another source.
It also keeps the final segmentation property flexible. The value might come from a Smart Property, but it could also be filled manually, imported from another data source, or captured directly at the point of entry, for example through a form.
The Smart Property then becomes one of several possible inputs into the final field, instead of the only way that field can be populated.
The first goal is usually to deal with the database you already have. Existing contacts and companies may be missing key ICP information, use inconsistent values, or have been created before your current segmentation model existed.
Instead of trying to populate every property at once, you can prioritise the most important parts of your ICP universe, run the research in stages and use workflows to move approved values into your final segmentation properties.
Once the existing database is under control, the focus shifts to keeping it that way.
New records will continue to enter HubSpot through forms, imports, integrations and manual entry. Some of the information you need may already be available at the point of entry, while other fields may need enrichment or AI research.
The goal is to create a clear order of operations. Use information you already have first, then use Smart Properties only where something important is still missing. Workflows can then populate the final segmentation properties and move the record into the right lists, lifecycle processes or prioritisation flows automatically.
At the end, segmentation becomes part of the ongoing CRM process rather than something you need to clean up again every few months.
Once the logic is working, you can automate the process with workflows.
Use a workflow to trigger the Smart Property fill only for records that actually need it. If HubSpot credits are exhausted, the workflow pause and continue once credits become available again, so the process does not have to be managed manually.
After the Smart Property has been populated, use workflow logic to copy or map the result into the final segmentation property. This lets you keep the AI-generated research separate from the operational field your lists, reporting and lifecycle processes rely on.
We can help you define the right ICP structure, build the properties and Smart Property logic, and connect everything into workflows that keep your database segmented over time. Get in touch to talk through your ideas.