48,000 unqualified contacts. One $5M pipeline.
A national real estate group, recruiting from 48,000 contacts nobody had ever qualified. Two months later, it had a talent pipeline.
- Industry
- Real estate recruitment
- Platform
- HubSpot, one portal
- Build
- 2 months
- Project lead
Cassy Woodforth- Award
- 2025 Impact Awards
01 The briefTurn state based activity into one national system for finding and converting high GCI agents, without taking the relationship off the managers who own it.
A database nobody had ever qualified.
The Agency recruits working agents away from other agencies. The people it wants are already employed, earning well, and in no hurry to move.
Winning an agent takes 12 to 24 months of relationship, across five states with a manager owning each. HubSpot held roughly 48,000 contacts.
What did not exist was the structure underneath: no contact types, no territories, no income tiers, no consent position recorded against a contact. Nothing could be scored, segmented or sent.
It had stayed that way because all four are decisions somebody has to own, and nobody owns them by default. The agents worth recruiting are also the least likely to fill in a form.
Four things had to be true first
- Know who is in there. An agent, a landlord and a buyer were in one list.
- Earn the right to send. Compliant reach is a consent problem, not a settings problem.
- Agree what good looks like. A score five managers will act on is one they helped build.
- Report it once. One set of numbers everyone reads.
One recruitment system, built in order.
One national pipeline, first touch to signed. Five jobs in sequence, and every one used to be a person guessing.
Work out who is actually in there
Agent, landlord or buyer, which state, who they work for, what they earn. Until that existed nobody could answer who The Agency could hire.
Work out who is worth a phone call
Fit is what they earn and where they work, engagement is whether they open anything. Both, and they rise to the top.
Ask permission before sending anything
A short personal note from the manager who runs their state, asking to stay in touch. Nothing else works without it.
Keep in touch without sounding like a mailout
Eleven journeys, each under a real manager’s name with their own market in it. Not right now goes onto a slower track.
Give everyone the same set of numbers
Every conversation in HubSpot, feeding two dashboards and 21 reports, so the meeting is about what to do next.
The order is the whole trick. Skip a step and the one after it produces a number nobody believes.
Nobody sits in a fixed square. Engagement moves week to week, and a tier changes with a good year.
Fit is what they earn and where they work, and a manager cannot talk it up. Nobody is ranked against anybody else.
Workshops before we touched a single record.
Adoption was the risk, so the build ran in the open: mapped with them, agreed with them, turned on in pieces.
Look before proposing anything, In the portal
We got into their actual HubSpot account first, so the proposal came from what was in there.
Scope: start where there is income data to score on
Find out where the data comes from, On a board
Every system feeding the database, drawn somewhere the client could argue with it.
Every field traced to the team that owns it
Agree what the process should be, In the open
MQL to SQL to handover, drawn in full. The session asked the managers why it would not work.
The courtesy opt in, the first segments, the scoring model
Build it in front of them, On their portal
Released in pieces they had already seen. Training ran in their own account, on people they recognised.
The first thing each manager saw was their own work
Four rules the build ran on
A person does the inviting
The managers invited the agents they wanted. That is why they got behind it.
Let the scoring earn its trust
It arrived a bit at a time, so they saw it work before they relied on it.
Ship something they can see
Basic scoring and one plain email shipped while the hard sorting was unfinished. Working and visible beat perfect and late.
AI writes the first draft only
Breeze wrote the first version of every manager email. A person approved each one.
48,000 contacts went nowhere because four decisions had no owner.
Got a database that big and no idea who is worth a call?
Four layers of HubSpot, each one holding the next.
From deduplicated data through to board level reporting.
Data foundation
Nothing could be scored until a contact had a type, a territory and a tier.
- Deduplicated and verified before import
- Contact types, territories and GCI tiers
- 18 contact and 34 deal properties
- 59 typed lists
Sales Hub
Where a scored agent becomes a deal with an owner and a next step.
- Combination lead scoring, fit plus engagement
- GCI tiers carry the fit score
- One national pipeline, 5 manager playbooks
- Forms that pre populate the record
- Deal owners assigned at pre qualification
Marketing Hub
Reach earned one opt in at a time, in five local voices.
- Manager led permission request
- 11 manager branded journeys
- Smart content per territory
- A “Not Right Now” track
- Breeze on first drafts only
Reporting
One set of numbers, so nobody argues about whose are right.
- 2 dashboards, 21 reports
- HubSpot Campaigns, end to end
- GCI pipeline and deal speed
- List growth and activity per manager
- Data quality as a metric of its own
The part that actually took the time.
Recruitment data is what an agent earns and whether they are thinking of leaving. Getting it wrong can cost somebody their job.
Access and permissions, Who can see and send to whom
- Two audiences, one portal. A recruitment campaign cannot reach a landlord list.
- Territory level visibility. A state manager sees their own market.
- Send on behalf of is granted as its own permission.
- Do not contact lists are fenced out at the list level.
Consent, done the slow way, Why the first send took weeks of work
- The one click way to make a database look reachable is the wrong call every time.
- A personal note from the manager for each state, sent from their own address.
- Opt out is permanent and immediate, offered in the first message.
- Slower than a bulk flag, and the only version that survives scrutiny.
Data quality and deliverability, What happens to a record before it is allowed in
- Verified on the way in. Addresses checked and duplicates collapsed before import.
- Every field has an owner. Dead properties archived, each live one traced to the team that uses it.
- Location fields reconciled. Two that disagree misroute smart content and lead assignment silently.
Visibility and handover, What the client is left holding
- Tracking that actually tracks. Website engagement is half the lead score, so it is verified before the score means anything.
- Activity on the record. Calls, emails and meetings log against the contact.
- Data quality is reported beside pipeline. What nobody measures decays.
- Handover by design. Project HQ holds every recording and board, and it is theirs.
Four goals. Here is what came of them.
What The Agency asked for before the build started, and where it stood two months later.
The results in five figures
- 48,000 Contacts they started with
- 44 Qualified deals in two months
- 5 Agents signed
- 3.4× More people The Agency could email
- $5M Yearly commission in the pipeline
Goal 01
Simplify the recruitment process
We built one national pipeline, where a deal opens on a score or on a manager who has verified fit.
1
national pipeline, from first touch to signed, across all five states. Every deal in it has an owner and a next step, and the five manager playbooks run on it.
Goal 02
Attract and convert higher quality talent
We built a combination lead score, so high fit plus high engagement is a shortlist worth calling.
11
manager branded email journeys behind that score. Each one goes out under a real manager's name with their own market in it. In two months, 44 qualified deals came through and 5 agents signed.
Goal 03
Improve data quality and segmentation
We ran a deduplicated import, then set up contact types, territories and GCI tiers.
59
typed lists, on 18 contact and 34 deal properties. Nothing could be scored until a contact had a type, a territory and a tier. With those in place, a contact can be scored, segmented and sent to.
Goal 04
Better attribution and commercial reporting
We built two dashboards covering GCI, deal speed, list growth and activity per manager.
21
reports, one set of numbers for everyone. Every conversation is logged in HubSpot and feeds the two dashboards, so the meeting is about what to do next.
A standing cadence underneath all four. A report every Monday: what shipped, what ships next, what it changes.
Put together, 48,000 contacts nobody had ever qualified became one compliant national talent pipeline in two months. There are 44 qualified deals in it, 5 agents have signed, and those deals are tracking toward $5M in yearly commission.
The audience The Agency has permission to email grew 3.4×, earned one opt in at a time through a personal note from the manager who runs each state. Every conversation feeds the same dashboards, so everyone reads the same set of numbers.
“We partnered with Neighbourhood to untangle and rebuild our recruitment process in HubSpot, and the results have been game-changing. Cassy was absolutely amazing. She didn’t just help ‘set up’ HubSpot, but took the time to understand our business, our people, and the unique challenges of recruitment in real estate.”
Trent Penfold
Head of Brand, The Agency
Who ran it
Cassy WoodforthAI & Digital Strategist
Led the data map, the fit and engagement score, the manager led consent model and the board reporting.
HubSpot Certified Trainer, HubSpot Impact Award winner, Neighbourhood since 2024
Want the same run at your database?
The neighbours next door.
The closest builds to this one, ranked by industry, HubSpot hub and service.