HubSpot For Asset Management: The 2026 Guide To Custom Objects, Compliance, And Real Platform Limits

The Quick Verdict: When HubSpot Fits An Asset Manager (And When It Doesn't)

The rest of this guide assumes you are somewhere between those two poles, which is where most firms genuinely sit. What follows is the detail that decides it.

What "HubSpot For Asset Management" Actually Means

HubSpot ships with four standard CRM objects — Contacts, Companies, Deals and Tickets. None of them is an investment concept. There is no portfolio, no holding, no subscription, no capital call, and no NAV. Everything that makes a CRM recognisable to an asset manager has to be added on top, and there are exactly four mechanisms for doing that.

The critical framing, and the one most content on this topic gets wrong: these mechanisms let HubSpot represent investment data. They do not make HubSpot a system that owns it. A Holding record in HubSpot is a reconciled copy, refreshed on a schedule, that exists so a relationship manager can have an informed conversation and a marketer can segment intelligently. It is not the position your accounting team reconciles against, and designing as if it were is the fastest way to a failed implementation.

The Tier Question: Why Custom Objects Change The Entire Business Case

Custom objects require HubSpot Enterprise. They are not available on Starter or Professional. This one fact reshapes the budget conversation more than any other, and it is conspicuously absent from most articles recommending HubSpot to asset managers.

The consequence is straightforward. Without Enterprise, a portfolio has to be crammed into a Deal record, a holding into a contact property, and a compliance review into a task. Firms do try this. It works until the first time someone needs to report holdings across clients, at which point the model collapses and the project restarts — on Enterprise, having spent a year proving the point the tier documentation already made.

So the honest sequence is: confirm the tier, then design the data model, then build. Not the reverse. If Enterprise is out of budget, that is a legitimate and useful finding — it means the decision is between a lighter HubSpot deployment that handles marketing and relationship management only, with portfolio data left in the systems that already hold it, or a different platform altogether.

Designing The Data Model Inside Ten Custom Objects

An Enterprise account can define up to 10 custom objects. That is the real design constraint, and it is why an asset management data model has to be planned rather than accumulated. A sensible core model spends five of the ten:

Five objects spent, five left for whatever your firm actually does differently — a Fund object for managers running pooled vehicles, a Capital Commitment or Subscription object for private-markets firms, a Mandate or Consultant object for institutional distribution. That headroom is the reason to resist modelling things as custom objects when a property or an association label would do.

Associations Are The Part People Underestimate

Associations are what turn five object types into a usable investment model: Holding to Investment Account, Account to Portfolio, Portfolio to Client, Client to relationship manager. HubSpot supports association labels, so a single Contact can be attached to a Portfolio as primary owner, joint owner, beneficiary or authorised representative — distinctions that matter enormously in financial services and that most generic CRM implementations flatten into a single undifferentiated link.

There are ceilings here too. HubSpot applies a default limit on how many records of a given type can be associated to one record, and since mid-2024 you can configure custom association limits through the Associations API and in Data Management → Data Model, up to a substantially higher ceiling. For most firms the defaults are generous. For a firm attaching every individual holding directly to a client record, they are not — which is exactly why Holdings should hang off Investment Accounts rather than off Clients.

HubSpot vs Financial Services Cloud vs Purpose-Built: Side By Side

Three genuinely different answers compete for an asset manager's CRM budget, and they are not interchangeable. The comparison below is about architecture and fit, not about which vendor is better.

Platform By Platform

Two things fall out of that table. First, none of the three is your book of record — that stays with your custodian, fund administrator or portfolio management system in every scenario. Second, the choice is really a question about where your constraint lives. If you cannot see your distribution pipeline, HubSpot's argument is strong. If you cannot represent your client relationships, Financial Services Cloud's argument is strong. If you cannot track allocators and consultants, a purpose-built platform's argument is strong.

Where HubSpot Falls Short For Asset Managers

These are the gaps that surface during implementation rather than during the sales cycle. None of them is disqualifying on its own; all of them need a deliberate answer before you commit.

The Compliance Reality: What HubSpot Does And Doesn't Satisfy

Direct answer: HubSpot does not make your firm compliant, and no CRM does. It automates the workflow around compliance — reviews, reminders, evidence collection, approval routing — while the retention obligation itself is met somewhere else. Treating the CRM as the compliance system is the single most consequential mistake in this category, and it is one the marketing literature around CRM in financial services actively encourages.

The obligations differ by registration, and conflating them is common:

• Registered investment advisers keep books and records under Investment Advisers Act Rule 204-2 — generally a five-year retention period, with the first two years readily accessible.
• Broker-dealers keep records under Exchange Act Rule 17a-4, which the SEC amended in October 2022 to permit an audit-trail alternative to the traditional write-once-read-many format and to address the use of third-party recordkeeping services.
• Dual-registered and hybrid firms have to satisfy both frameworks simultaneously, which in practice means designing to the stricter of the two.
• Advertising and testimonials fall under the SEC Marketing Rule, Rule 206(4)-1, fully effective since November 2022 — directly relevant because investor marketing is precisely what firms buy HubSpot to do.

The practical architecture that follows from this is consistent across the regulated firms we work with: HubSpot generates and logs the communication, and a dedicated archiving platform — Smarsh, Global Relay or Proofpoint are the names that come up most — provides the compliant, tamper-evident retention. HubSpot supplies the workflow and the audit trail of who did what and when; the archive supplies the durable, producible record.

What The Marketing Rule Means For A HubSpot Build

Because Rule 206(4)-1 governs advertising, testimonials and endorsements, an RIA running campaigns out of HubSpot needs review and approval built into the sending process rather than bolted on afterwards. Practically, that means marketing emails and landing pages route through a compliance approval step before send, every approved version is retained with its disclosures intact, and any testimonial or endorsement carries the required disclosure about client status and compensation. This is entirely buildable in HubSpot — it just has to be built, and it has to be built before the first campaign goes out rather than after an examination asks for it.

• Map which registration you hold, and therefore which retention rule governs you, before configuring anything.
• Select the archiving platform first — it constrains how communications must flow through HubSpot.
• Build compliance approval into the marketing send process, not around it.
• Model the evidence: which HubSpot record proves a review happened, who performed it, and when.
• Have your compliance officer validate the design, in a sandbox, before production data lands in it.

This section describes how firms commonly architect their systems and is not legal or compliance advice. Retention periods, supervisory obligations and marketing review requirements vary by registration, jurisdiction and firm; confirm your specific obligations with your compliance counsel and against the current rule text.

The Integration Layer: Where Asset Management Builds Actually Break

Getting custodian and portfolio management data into HubSpot is not conceptually hard. Getting it in every night, reliably, within the platform's rate limits is where these projects succeed or fail. The published ceilings matter more than any feature:

The arithmetic is worth doing explicitly, because it is the argument that wins the design debate. Fifty thousand holdings updated one record at a time is fifty thousand calls, plus a lookup call each if you search for them first — call it a hundred thousand, against a burst ceiling that allows roughly nineteen a second. The same fifty thousand holdings through batch endpoints is five hundred calls. Same data, same outcome, two orders of magnitude apart in platform cost.

Three Patterns That Keep Syncs Healthy

• Batch everything, always. Treat single-record write endpoints as a debugging tool, not a production pattern.
• Use a unique external identifier, not search. Store the custodian's own account and position identifiers as unique properties so records can be upserted directly instead of searched for. This removes the strictest rate limit from your critical path entirely.
• Sync deltas, not the world. Most nights, most positions have not changed. Pull only what moved since the last successful run, and keep a full reconciliation as a weekly or monthly job rather than a nightly one.

Webhooks handle the other direction — events in HubSpot that downstream systems need to know about, such as a completed compliance review or a change in client status. They are not a substitute for the scheduled sync; they complement it.

Automation And Calculations, Written Correctly

Standard HubSpot workflows cover the majority of asset management automation: scheduling portfolio reviews from a last-review date, triggering KYC and AML renewals ahead of expiry, creating follow-up tasks after meetings, escalating a compliance review that has gone past due. That is declarative work and it belongs in workflows, where your operations team can maintain it.

Calculations are different. Rolling holdings up into a portfolio value, deriving a risk score, or computing allocation percentages needs a custom code action — and custom code actions run inside real limits:

• Written in Node.js or Python, executed on serverless infrastructure.
• A hard execution ceiling of roughly 20 seconds, with memory capped around 128 MB.
• Up to 50 HubSpot properties available as inputs to a single action.
• String outputs capped near 65,000 characters.
• A restricted library set — the HubSpot API client and a HTTP library, not arbitrary packages.
• Automatic retries on rate-limit and server errors, backing off over an extended window.

A Correction Worth Making

A portfolio-value snippet circulates widely in articles on this topic, and it does not run. It looks like this:

// `holdings` is never defined; nothing provides it.
const totalValue = holdings.reduce((sum, h) => sum + (h.units * h.nav), 0);
output.set("portfolio_value", totalValue);

Three things are wrong with it. A custom code action receives the properties you explicitly map as inputs — it is not handed an array of associated holdings, so holdings does not exist and has to be fetched. HubSpot property values arrive as strings, so multiplying them without conversion produces string concatenation or NaN rather than a number. And output.set() is not the output mechanism; results are returned through the callback. The working shape looks like this:

const hubspot = require('@hubspot/api-client');

exports.main = async (event, callback) => {
  const client = new hubspot.Client({
    accessToken: process.env.PRIVATE_APP_TOKEN
  });

  // The action is handed the record it fired on, not its children.
  const portfolioId = event.object.objectId;

  // Associated holdings must be read, then their properties fetched.
  const links = await client.crm.associations.v4.basicApi.getPage(
    'portfolio', portfolioId, 'holding'
  );

  let total = 0;
  for (const link of links.results) {
    const holding = await client.crm.objects.basicApi.getById(
      'holding', link.toObjectId, ['units', 'nav']
    );

    // Properties come back as strings — coerce before arithmetic.
    const units = Number.parseFloat(holding.properties.units);
    const nav   = Number.parseFloat(holding.properties.nav);
    if (Number.isFinite(units) && Number.isFinite(nav)) total += units * nav;
  }

  // Results are returned through the callback, not a global setter.
  callback({ outputFields: { portfolio_value: total.toFixed(2) } });
};

Note what that loop implies. One request per holding, inside a twenty-second budget, means this pattern is fine for a portfolio holding tens of positions and unsuitable for one holding thousands. Past that scale the calculation belongs in your integration pipeline, which writes the finished portfolio value into HubSpot as a property rather than asking HubSpot to derive it. Knowing where that line sits for your firm is a design decision worth making early — it is far cheaper than discovering it when the action starts timing out in production.

The Reference Architecture

Every successful asset management build we have seen resolves into the same four layers. The value of stating them explicitly is that it makes the boundaries arguable — and the boundaries are where these projects go wrong.

Alongside these four sits the archiving platform, which is deliberately not a layer in the stack — it observes communications rather than participating in the data flow. Drawing it as a layer is a common diagramming error that leads firms to assume HubSpot can be configured into compliance.

The 2026 AI Layer, And Where It Is Actually Useful

HubSpot's AI has been reorganised over the past year. Agents that shipped under the Breeze name now live under Agent Hub, with the builder tooling renamed Agent Builder, and Breeze has stepped back as the umbrella brand. Generally available agents cover customer support, prospecting and CRM data quality, with research and customer-health agents in beta. Standard data enrichment, previously sold as Breeze Intelligence, is now included with core seats rather than consuming credits for basic firmographic fields.

For an asset manager, the honest assessment is that the data quality agent is the one that pays for itself. CRM data decay is the chronic problem in this sector — stale contact details, duplicate entities, relationship managers who left two firms ago — and it degrades every downstream number, including the ones leadership uses to make decisions.

The prospecting and customer agents warrant more caution in a regulated firm, for a reason that has nothing to do with the technology's quality: anything that generates outbound communication to investors is advertising, and advertising is governed. An AI agent drafting investor outreach sits squarely inside the Marketing Rule's scope. That does not rule it out — it means the same review and approval path that governs human-written campaigns has to govern agent-generated ones, with a person approving before anything reaches an investor.

Decision Framework: Five Questions, In Order

Work these in sequence. Each one can end the evaluation, which is the point — the cheapest decision is the one you reach before the build starts.

Implementation Realities

A focused build — investor marketing, relationship-manager activity, pipeline reporting, no custom objects — typically runs 6 to 10 weeks. A full build with a portfolio and holdings data model, a custodian or portfolio management system integration, and compliance workflows generally runs 3 to 6 months.

What drives the difference is almost never CRM configuration. It is data migration, reconciliation rules and archiving integration. Three specifics account for most of the overrun we see:

The reconciliation question deserves particular attention because it is the one firms consistently defer. When the nightly sync reports a position that contradicts what a relationship manager is looking at on screen, what happens? Does the record update silently, flag for review, or block? There is no universally right answer, but there is a universally wrong one, which is not deciding until it happens in production in front of a client.

A Pre-Build Checklist

• Licence tier confirmed in writing, against the data model you intend to build.
• Object model drawn and reviewed — including which of the ten definitions stay unspent.
• Unique external identifiers agreed with the custodian or portfolio system owner.
• Reconciliation behaviour decided for every field that syncs.
• Archiving platform selected and its HubSpot integration path verified.
• Compliance sign-off on the design, in a sandbox, before production data lands.
• Named owner for the data model after go-live — models decay without one.

The Bottom Line

HubSpot is a genuinely capable platform for asset management firms whose growth is constrained by distribution — and a poor choice for firms who need a financial data model, institutional IR tooling, or a compliant archive to arrive pre-built. Both statements are true at once, and the value of an evaluation is working out which describes you.

1. 1

   Confirm Enterprise before designing anything. Custom objects are Enterprise-only, and every asset management data model depends on them.

2. 2

   Design inside ten custom objects. Spend them deliberately and keep headroom for what your firm does differently.

3. 3

   Keep the book of record where it belongs. HubSpot holds a reconciled copy for relationship purposes, never the authoritative position.

4. 4

   Pair HubSpot with a real archive. The CRM automates compliance workflow; it does not satisfy the retention obligation.

5. 5

   Design the integration around batch endpoints and external identifiers from day one. Rate limits, not features, decide whether the sync survives.

6. 6

   Adopt AI for data quality first, investor-facing generation second — and only once an approval path governs it.

If your firm's answer is that HubSpot fits, the work is architectural rather than administrative: the object model, the sync design, and the compliance boundary decide whether the implementation holds up two years from now. If the answer is that it doesn't fit, finding that out during an evaluation costs a few weeks. Finding it out eighteen months into a build costs considerably more.

Frequently Asked Questions