Oct 14, 2026, 11:44:59 AM

Fundraising on firmer ground: rebuilding a VCT's HubSpot for a harder year

Discover how a UK VCT transformed its CRM processes to enhance fundraising efficiency, improve data quality, and adapt to a challenging market landscape.

By Marie Roberts
rebuilding a VCT's hubspot for a harder year

When a CRM holds three versions of the same investor application, every fundraising report arrives with a caveat. Someone has to explain which numbers to ignore, which adviser is actually new, and why last year's pipeline still looks busy.

A UK investment manager that raises from private investors, largely through financial advisers, wanted those caveats gone before its next fundraising round. Centralise rebuilt the HubSpot foundations underneath it: separate pipelines for new business and adviser relationships, a lead-scoring model, an application history cleaned and matched on a unique identifier, buyer-intent tracking, and dashboards the client's own team now runs and refines.

A market that just got harder to raise in

Raising from private investors in the UK has become harder this year. Venture capital trusts, the most closely tracked part of the tax-advantaged market, show it clearly. They raised £918m in the 2025/26 tax year, 3% up on the year before and the third highest total on record (AIC). A good part of that was investors getting in before upfront income tax relief fell from 30% to 20% on 6 April 2026.

The current season looks different. In a RAM Capital survey run in May and June 2026, 97% of advisers expected VCT demand to fall in 2026/27, and 55% expected a significant decline (IFA Magazine). Even last season, VCTs collectively raised 63% of their target amounts, down from 72% the year before (MICAP via IFA Magazine).

For managers that raise through advisers, a softer market rewards the ones that know which advisers are active, which applications are genuine, and where new interest is building. All three of those live in the CRM, and all three depend on the data being right.

Where the CRM was falling short

The client's HubSpot portal had grown alongside its fundraising and marketing activity, and it was being asked to do far more than store contacts. The team needed one view across adviser relationships old and new, leads and applications, events, website intent and pipeline reporting.

Two problems sat underneath everything else. The portal held duplicated application records and historical deals that no longer matched how the team worked. It also had no clean way to separate application years, adviser types and pipeline stages, so a report on this year's fundraising could quietly include last year's.

With a new launch coming, the team wanted confidence that its processes, integrations and reporting would hold up under more volume.

Two kinds of relationship, two pipelines

A new adviser the client has never worked with and an adviser who has placed clients for years need different handling, so the first job was to stop treating them as the same record type. New business opportunities and ongoing adviser relationships now run in separate lead pipelines.

New leads move through a short set of stages from first awareness to investment, with a clear exit for those that stop progressing. Stage changes happen automatically when something meaningful occurs, such as email engagement or website activity.

Leads that stop progressing now carry a reason and supporting notes. Over a season, that turns into a record of why opportunities stall, which is far more useful in a soft market than a list of leads that simply went quiet.

A scoring model built to be adjusted

A lead-scoring framework gives the team a ranked view of who is most engaged and most relevant. It combines engagement signals, such as website behaviour, event and email activity and direct conversations, with fit criteria like seniority and audience match.

Scores update retrospectively when the criteria change, so the team can reweight the model once live data shows which signals actually predict an application.

Cleaning the history without losing the context

The duplicated application records were the part of the project with the most direct effect on reporting. Duplicates inflate pipeline value, split an adviser's activity across several records, and make year-on-year comparisons meaningless.

Gartner puts the average cost of poor data quality at $12.9m a year per organisation (Gartner). That figure comes from large enterprises and overstates the bill for a firm this size. The pattern behind it holds at any scale, though. The long-standing 1-10-100 rule says a record costs $1 to verify at entry, $10 to clean later and $100 if it is left alone (ORM Tech).

The cleanup followed an agreed sequence:

  1. Remove outdated and duplicated deals for the affected fundraising years.
  2. Keep contact-level notes wherever they weren't stored on the deals being deleted.
  3. Reformat the source data into separate uploads for each application year.
  4. Map every column to the right HubSpot field: deal stage, application year, amount, adviser, company and investor.
  5. Re-upload the clean records in that structure.

The step that stops the problem coming back is the unique identifier. Every regulated adviser firm already has a reference number, and that number is now mapped as a company property, so incoming applications match to the existing company record instead of creating a new one. The team worked through the upload with Centralise in a live session: filtering the source data, checking errors and aligning fields before the clean records went in.

Seeing interest before the form is filled in

More and more investor research now happens inside AI tools, long before anyone fills in a form. FCA research published in August 2026 found that four in five less experienced investors aged 18 to 40 have used AI for help with investing, and 56% trust AI tools, more than TV and radio (47%) or the press (46%) (FCA). The same study found 86% understood the need to check the sources AI cites.

That puts two jobs on the client's marketing: spotting research activity it can act on, and being one of the sources AI tools cite.

Buyer intent

HubSpot's buyer-intent tools were configured around a defined target market. The setup tracks website visitor intent and research into topics the client's investors care about, creates leads automatically from relevant signals, and sends the team regular summaries of recent activity.

Company matching has limits. Large organisations with subsidiaries, several trading names or multiple domains don't always resolve to one record. The agreed operating model is to set the criteria broadly, then review and merge manually where HubSpot can't confidently tell that two records are the same organisation.

SEO and AI search visibility

Training covered visibility in both traditional search and AI search: SEO recommendations, page-quality checks, AI search prompts, competitor visibility and citation opportunities. The content recommendations focused on clear, practical explainers that search engines and AI tools can reference when someone researches the product category.

Campaigns, illustrations and reporting

A flagship investor event became a single campaign in HubSpot, pulling together its forms, emails, landing pages and performance data. From one place the team can see influenced contacts, engagement and likely return. The work also set out how to bring in attendees from external event platforms and record who registered, attended or cancelled.

Adviser submissions were checked end to end, so each one creates or updates a deal in the right pipeline with its supporting detail attached. The team also learned to read weighted pipeline value properly, so expected value reflects each deal's probability rather than its headline amount.

A set of dashboards now covers marketing, sales and fundraising together, including lead-score distribution, email and meeting activity, buyer intent, page views, campaign performance, deal stages, loss reasons, win rate, funnel progression and fundraising performance. They were built as a starting point, to be tightened once a full season of live data runs through them.

Where it leaves the client

The client went into its next round with a defined operating model for its CRM: how data enters, how leads progress, how applications attach to companies and advisers, and which reports the team trusts. New business and adviser relationships are visible separately. Application history is clean, and the unique identifier keeps it that way. The team was trained to run and adjust the system itself.

The system gets more valuable as live fundraising data runs through it. The next steps are practical: use the processes day to day, collect the questions that come up, and review performance together. Further cleanup, event integrations and a reweighted scoring model are already on the list.

Sources