Recruitment/Data
National recruitment group
Challenge. Three spellings of the same client company across 60,000 records, deals owned by people who had left, and a board pack the leadership team openly did not believe.
Approach. Full data audit, a rebuilt property and pipeline architecture, deduplication in staged passes with human review, then reporting moved natively into HubSpot.
- Duplicate rate
- 18% to 3%
- Report build time
- 2 days to 20 min
- Records reviewed
- 61,400
The first honest pipeline number we have had in four years. Nobody argues about the report any more.
SaaS/Automation
B2B SaaS platform
Challenge. Inbound leads were sitting unassigned for days, lifecycle stages meant something different in every team, and routing lived in one person's head.
Approach. Lifecycle model defined and documented, scoring rebuilt against real closed won data, round robin routing with escalation, and Slack alerts on high intent signals.
- Lead response time
- 31h to 24m
- Unassigned leads
- 22% to 0%
- Workflows consolidated
- 84 to 19
It is the first time the process on the whiteboard matches the process in the system.
Manufacturing/Marketing
Precision manufacturer
Challenge. A rebuilt website had launched with no tracking plan, so spend across paid and trade channels could not be tied to any enquiry.
Approach. Migration to HubSpot CMS with a proper form and campaign taxonomy, source attribution wired end to end, and technical SEO cleanup on the product catalogue.
- Attributed enquiries
- 12% to 91%
- Organic sessions
- Up 2.4x
- Cost per enquiry
- Down 38%
We finally know which trade shows are worth attending and which are not.
Financial Services/AI
Financial services broker
Challenge. Advisers spent hours writing pre-meeting summaries by hand, using notes spread across email, spreadsheets and the CRM.
Approach. Data cleaned and consolidated into HubSpot first, then an AI assistant scoped to draft pre-call briefs from CRM history only, with a human check before sending.
- Prep time per meeting
- 45m to 6m
- Notes captured in CRM
- 38% to 96%
- Advisers onboarded
- 27
They refused to build the AI part until the data was right. That turned out to be the whole job.
