How NCBA CLUSA Leveraged HubSpot’s AI Data Agent With RevPartners

The Challenges

NCBA CLUSA wanted to get more out of HubSpot and make some of its most time-consuming data and research work easier to manage. RevPartners helped to identify a few areas where the existing process could be made more efficient and scalable. 

1. Company data needed more consistency 

Some company records were missing important information like names and domains, while other data had been entered in different ways. For example, the same U.S. state could appear under several different spellings.

At scale, those inconsistencies made it harder to use the data reliably for segmentation, reporting, and targeting.

2. Identifying potential members required manual research 

NCBA CLUSA has a unique qualification challenge. Two organizations can operate in the same industry, but only one may actually be member-owned and qualify as a potential member.

Standard HubSpot fields like industry and company type don't capture that distinction. The team had to research organizations individually and use that information to determine whether they were truly cooperatives.

3. Important HubSpot work required valuable team time 

Without a dedicated data or RevOps function, NCBA CLUSA's membership operations lead managed HubSpot alongside the team's core membership work.

That included system administration, keeping data accurate, and researching potential members. As the organization looked to grow, there was an opportunity to automate more of that work and give the team more time for direct member engagement.

4. Preparing for meetings required additional research 

Before each membership conversation, the team needed context on the organization, including what it did, its sector, whether it was truly a cooperative, and a relevant way to approach the conversation.

There wasn't a system in place to gather that information automatically, so the research had to be done for each meeting. As the pipeline grew, the amount of preparation grew with it.

The Solutions

RevPartners focused on helping NCBA CLUSA get more from the tools already available in HubSpot. The system was designed to use free HubSpot enrichment and automation first, then turn to AI only when it was needed. This helped NCBA CLUSA make the most of its monthly AI credits while still automating work at scale.

Together, the teams put five key solutions in place:

1. Automatically filling in missing company domains 

RevPartners built a workflow that uses the email domains of associated contacts to fill in missing company domain names.

This step doesn't use any AI credits, and gave HubSpot the domain information needed for AI-powered research later in the process.

2. Filling in missing company information with AI  

When HubSpot has a company's domain but other information is missing, the workflow first tries HubSpot's free enrichment.

If that doesn't fill the gaps, Data Agent researches the company and can add its name, industry, year founded, address, country, and state. All of that information can be gathered in a single AI run of about 10 credits per record.

3. Cleaning up state data without wasting AI credits  

RevPartners created a two-step process to standardize state information across more than 7,000 eligible company records.

First, a free workflow looks at the existing state or region and matches recognizable variations to the correct U.S. state. Only when it can't confidently make a match does Data Agent use the company's name and domain to research the likely state.

This meant the vast majority of eligible records could be standardized without using any AI credits, saving those credits for the records that actually needed additional research.

4. Using AI to identify and classify cooperatives 

RevPartners built an AI-powered workflow to help solve one of NCBA CLUSA's most important qualification challenges.

When a company has a known domain but no Co-op Sector, Data Agent researches the company's public presence using its name and domain, along with its homepage, About page, and LinkedIn. It then classifies the organization using NCBA CLUSA's existing list of approved co-op sectors.

Because the AI can only choose from those predefined options, the resulting data can be used directly for segmentation and reporting without additional cleanup.

5. Automating research before meetings  

RevPartners also built an AI-powered smart property that prepares a short company briefing ahead of a scheduled meeting.

The briefing covers what the organization does, whether it's truly a cooperative, when it was founded, and a relevant starting point for the conversation.

Rather than running this research across the entire database, it only runs for companies with an open deal and an associated upcoming meeting. That keeps AI usage focused on the records where the information is most useful.

The Impacts

The new workflows helped NCBA CLUSA reduce manual work, improve its data, and use AI more intentionally. The team was also able to do more with the HubSpot tools and AI credits already available to them. 

1. 60–70% of missing company names resolved automatically

Automated enrichment corrected approximately 60–70% of affected company records without requiring manual intervention. 

2. Approximately 7,000 company records processed

The state normalization workflows processed approximately 7,000 company records, with the substantial majority resolved without using any AI credits.

By starting with free HubSpot enrichment and only using AI when necessary, NCBA CLUSA was able to address most of this data-quality challenge without creating ongoing AI spend.

3. More companies classified by cooperative sector

Once the team confirmed there was enough AI credit capacity, cooperative sector classification expanded from an initial two-month historical window to several additional months of backfilled data.

This directly helped close a key gap where standard HubSpot data couldn't answer NCBA CLUSA's core qualification question.

4. Approximately 2,500 companies with no contacts identified

The work also uncovered approximately 2,500 company records with no associated contacts.

Instead of that gap remaining hidden in the database, NCBA CLUSA now had a prioritized, actionable list of records for cleanup.

5. Pre-call research automated

Pre-call preparation was automated specifically for companies with an active deal and a scheduled meeting.

This gave the team useful context ahead of those conversations while keeping AI usage focused on sales-qualified activity rather than running across the entire database.