The Marketing Data Platform

Introduction

HubSpot is a leading GTM platform for SMBs looking to consolidate their CRM, marketing, service, and operations tools in one place. HubSpot's marketing automation product was designed to work natively with their own CRM, making it a natural fit for teams operating entirely within the HubSpot ecosystem. However, teams that use Salesforce as their CRM alongside HubSpot Marketing Hub often find that the sync between the two platforms has limitations, largely because HubSpot's sync architecture was optimized for its own data model. Conversion has built an enterprise-grade Salesforce sync as part of a flexible data model, designed specifically to meet the needs of fast-growing enterprises that depend on Salesforce as their system of record.

Salesforce Campaign Syncing

Conversion supports the concept of a "Campaign," similar to Marketo Programs. Campaigns are an enterprise-grade feature that lets you organize assets, such as workflows, forms, and emails around a specific marketing initiative, and sync them directly with Salesforce campaigns. When you create a campaign and connect it to Salesforce, Conversion automatically creates the corresponding Salesforce campaign and maintains a continuous two-way sync of members, campaign statuses, and campaign member fields.

HubSpot also offers a campaigns feature, though it's designed primarily to group and organize marketing assets rather than to sync with Salesforce campaigns. HubSpot campaigns do not support bidirectional member or status sync with Salesforce.

Many enterprise marketing teams rely on Salesforce campaigns for revenue reporting and marketing attribution. Conversion's tight, bidirectional sync with Salesforce campaigns is built with this use case in mind.

De-Duplication

HubSpot's Salesforce sync de-duplicates records based on email address. In practice, this means scenarios where the same email is associated with both a lead and a contact in Salesforce can result in records being merged, blocked from syncing, or dropped.

Conversion takes a different approach by treating Salesforce as the system of record and identifying leads and contacts by their Salesforce Record ID. This maintains a true 1:1 relationship between every Salesforce record and its counterpart in Conversion, and gives users full control over when and how contacts are merged or de-duplicated. If a lead and a contact share the same email in Salesforce, Conversion will update both records independently rather than collapsing them into one.

Sync Rules and Inclusion Lists 

HubSpot uses inclusion lists to control which contacts are eligible to sync to Salesforce. Contacts that don't meet the defined criteria, or that fall out of an inclusion list over time, won't sync, which can create gaps between marketing engagement data and sales visibility.

Conversion offers more granular control over sync rules across the entire platform. Sync settings can be configured at the global level (for example, automatically creating a new contact in Salesforce whenever one is created in Conversion) or at a more specific level. Individual forms and CSV imports can each have their own sync configuration, and workflows can include dedicated "Sync to CRM" nodes that let you conditionally sync contacts based on enrichment logic. For instance, you could configure a workflow to exclude contacts who submitted a personal Gmail address from syncing to your CRM.

This flexibility means sync behavior can be tailored to your team's processes rather than managed through a single, centralized list.

Native Salesforce Actions in Workflows

Within Conversion workflows, you can perform Salesforce-specific actions to support your sales processes, including:

• Linking a contact to an existing Salesforce campaign
• Creating a Salesforce task
• Converting a lead to a contact

These actions allow marketing and revenue operations teams to trigger sales workflows directly from within Conversion, without needing to leave the platform or rely on manual steps in Salesforce.

Sync Audit Logs

Conversion includes a built-in audit log that continuously monitors the health of your Salesforce sync. You can view a complete history of all inbound (Salesforce → Conversion) and outbound (Conversion → Salesforce) sync events, including the status of each record and the reason for any failures. This visibility makes it straightforward to diagnose and resolve sync issues quickly, giving operations teams confidence that their data is staying in sync as expected.

Introduction

For maturing B2B enterprises, teams are increasingly feeling the importance of expanding revenue through the upsell/cross-sell motion. Almost 40% of ARR for late-stage B2B software companies comes from expansion revenue. Leadership may make aggressive changes to deploy a new product suite or suddenly realize that they haven’t prioritized lifecycle marketing properly. Marketing motions to support this can’t be deployed solely off of CRM data. You’ll need to activate your data warehouse to unlock critical insights and reach out to customers at the right time. 

Traditionally, teams have tried to use a reverse ETL, in-house data tools, or a CEP to power marketing automation through warehouse data. This leads to fragmentation, tool sprawl, missing leads, and general data chaos. New-age platforms like Conversion unify warehouse and revenue data to create a holistic view of the customer – making upsell/cross-sell motions effortless. Let’s do a deep dive on why product data is essential to your marketing automation platform if lifecycle motions are on your radar. 

What data lives inside your data warehouse

Data warehouses seem scary to the non-technical user. Typically siloed in engineering teams, it’s rare that business teams ever interface with the warehouse. In reality, they are a critical component of any fast-growing enterprise that has crucial insights into customer data. Warehouses store information such as: 

• An entire database of all users of your software
• Records of which products are used by which contacts and companies
• Product event logs that record millions of actions every day
  
  • Logins
  • Teammate Invites
  • Feature Usages
  • Page Views 
• Transformations of data to create simplified variables and data fields (Typically executed by a data engineering or analytics team)
  
  • Ex. Turning complex feature usage into an “Account Health Score"

This is just a peek into your warehouse. They are made up of thousands of tables, all of which relate to each other and store valuable information.

Guessing Games Are Killing Pipeline

Typical B2B SaaS orgs do a “vibes-based” upsell/cross-sell motion. AEs are instructed to “reach out when it feels right” or “reach out every month” which ends up being ill-timed or spammy. Marketing is instructed to throw new products in customers faces on a regular cadence. “They won’t churn! They’re locked into a year-long contract.” This is not an expansion strategy or lifecycle motion. It’s a glorified (and randomized) calendar. On top of this you’re slowly losing credibility with your customer base, who believe you don’t understand their business fundamentally. 

The guessing game approach is losing you real revenue every month. Pipeline becomes inflated with low-quality opportunities while high-probability expansion moments slip through the cracks. Meanwhile, leadership may not immediately feel the damage. Contracts are still active. Revenue hasn’t churned. But what’s happening underneath the surface is more dangerous: expansion potential is decaying. Accounts that could have grown naturally stall. Advocates become neutral. Neutral customers become price-sensitive at renewal. What looks like “stable ARR” is often suppressed expansion.

Expansion Is About Readiness, Not Reach

One of the most common mistakes teams make is treating upsell and cross-sell like demand generation. They expand audiences, increase message volume, and optimize for reach, assuming that more exposure will produce more expansion. In reality, expansion depends far less on volume and far more on readiness. No matter how many times someone gets an email for an expansion opportunity, they won’t budge until the business as a unit is ready to purchase. 

Readiness shows up in subtle ways. Buyers won’t explicitly knock on your door to tell you that they want to buy a new product or need more AI credits on their existing product. The signals for readiness live deep inside your data warehouse. Is someone > 90% of their limit on AI credits? Has an account crossed 10 total users in the past 2 weeks? These are potential signs that a contact or account is gearing up for a potential upsell or cross-sell. It’s now on sales and marketing alignment to act upon this signal and target the buyer through hyper-personalized automation. 

If a marketing automation platform cannot understand how customers are using the product, it has no reliable way to determine when expansion is appropriate. As a result, teams either move too early, risking irrelevance and losing trust, or too late, missing the moment altogether.

Why Static Fields Fall Short

You might be wondering – “I thought our data team added some product data to Salesforce”. In most cases, these are static fields that don’t reflect the complexity of the buyer account and their usage. It may have basic information such as “Uses XYZ products” or “Tier = Enterprise” but it certainly does not contain in-depth feature usage, events, complex objects, etc. 

Your warehouse is a living breathing universe of infinite data. Using it for a few custom fields is one thing but the real value comes from the ability to record events, attributes for those events, custom data objects, and attributes that can relate a contact or company to that object. Once you have a data setup that contains real-time information such as “Used X feature in the last 30d” or “Approaching seat limit”, you’ll be able to run powerful up-sell/cross-sell campaigns to hyper-complex audiences. 

Why Product Data Belongs Inside Marketing Automation

For upsell and cross-sell to work at scale, product data must be more than something teams analyze in dashboards. It must be directly actionable. When product data lives inside the marketing automation platform itself, expansion logic becomes automated rather than manual. Messaging stays tightly aligned with real usage. Whether it’s creating smart automations to send ads and notify sales the moment a key signal is achieved within an account or inserting custom warehouse variables into dynamic emails, the data warehouse enables GTM-wide alignment on lifecycle marketing. Sales, marketing, and customer success operate off the same signals instead of competing interpretations. Most importantly, revenue expansion becomes predictable and achievable rather than opportunistic and serendipitous.

Introduction

Your data warehouse is a goldmine of customer information. With information on which products customers are using, how they are using them, and key adoption signals, enterprises are able to run powerful upsell/cross-sell campaigns across sales, marketing, and customer success.

Yet most marketing automation platforms were never designed to operate on warehouse data. Legacy tools like Marketo and HubSpot were built on top of Salesforce, with rigid contact-centric data models. If you want your marketing automation to support custom warehouse objects, you’ll need a tool that supports a native integration with both Salesforce and your data warehouse. 

While HubSpot has recently introduced a native Snowflake sync, it primarily ingests basic fields rather than supporting a flexible, object and event driven data model. Conversion has built a flexible data model to support the ingestion of complex CRM, marketing, and warehouse objects. In this post, we’ll outline the key advantages of the Conversion + Snowflake sync and why it is more robust than HubSpot. 

What is a data warehouse?

Let’s start by going over how a data warehouse works. At their core, data warehouses are an extremely large collection of tables that relate information from one table to another. One table may store information about all the products your company offers and their associated properties (Name, SKU, Price, etc.). Another table may reference which companies are using which products by referencing the specific product IDs in the original table. This is known as a relational database. You may also have a table which acts as an event log, storing information about every time a user logs in or creates an item within your application. 

Warehouses are not just static spreadsheets – they are a grouping of complicated, realtime datasets. 

How The HubSpot + Snowflake Integration Works

The HubSpot + Snowflake integration treats warehouses as a spreadsheet, allowing you to configure basic field mappings. However, they do not cover the depth of custom object support that is required to utilize the true value of a warehouse. 

Once you connect your Snowflake instance to HubSpot, you are able to choose a table that you want to sync data from, which object within HubSpot you want to sync to (contact, company, etc.), and how you want to join that data via a unique identifier (email, ID, etc.). Finally, you can configure field mappings to send data from a Snowflake column into a property on your specified HubSpot object. The HubSpot + Snowflake sync mirrors the concept of importing a big CSV file and mapping the columns to different HubSpot fields. 

While the sync provides functionality to map basic properties (ex. Company Billing Tier), it doesn’t support the depth of an intricate data warehouse. 

How The Conversion + Snowflake Integration Works

The Conversion + Snowflake integration was built into the core of our marketing automation platform – not as a bolt-on that supports limited functionality. 

Let’s start with a few Conversion-specific definitions that will be critical to understanding the benefits of our Snowflake sync. Conversion supports an extremely important concept called custom objects. These are custom records that you define outside of the standard “Contacts, Leads, Accounts, Opportunities” objects. These can either be imported from SFDC or your data warehouse. Each custom object can have fields, which contain additional information about the object. 

These custom objects are related to contacts through relationships. Relationships link these custom objects to contacts, so their fields can be queried / triggered on in workflows and segmentations. Relationships can even hold data themselves, through “relationship attributes.”

Example: For an enterprise business that has multiple products across their entire suite, you can create a “Products” custom object. Properties for a “Products” custom object would include the Name, SKU, price, etc. Relationships on the other hand include how contacts and companies within Conversion relate to that specific object (ex. A contact’s last login date on that product or a company’s seat count for that product).  

Conversion also has Events which serve as records of actions taken by contacts, that can be triggered on. For example, you can trigger workflows and actions every time someone fills out a form or opens a marketing email. 

Once you’ve connected Snowflake to Conversion, you have a multitude of options on how you can sync your data. You can:

• Create / update contact records
• Create custom events
  
  • Add properties to custom events
• Create / update custom objects
  
  • Add properties to custom objects 
  • Add relationships to custom objects 

Within HubSpot, we were able to simply create contacts & update fields on any object. The Conversion data model supports complex objects which allow for deeper personalization. 

• Syncing in events and their associated properties from event-specific tables unlocks huge opportunities for automation. 
  
  • Ex. Trigger a marketing email every time a user creates an invoice where invoice value > $5,000
  • Ex. Trigger a marketing email when a user invites > 5 teammates to the platform and the teammate role was defined as “Admin”
• Syncing in table records as custom objects and then tying contacts & accounts to those objects as relationships. These custom objects can be used to create powerful segments and triggers with our Unified Statement Query Engine
  
  • Ex. A “Products” custom object that ties contacts & accounts to a specific product. Relationship attributes for contacts include role (User, Admin), last login date, etc. Relationship attributes for companies include Seat Count, Billing Tier, Renewal Date. 
    
    • Example audience trigger: Target all contacts who are an Admin using the “CRM” product where Company Seat Count > 5

Going beyond a 1:1 field mapping and supporting complex objects means that there needs to be a custom way to transform data within Conversion. Conversion has a built-in AI SQL editor which allows you to manipulate the data in your warehouse before you bring it into Conversion via contacts, custom objects, events, etc. 

Why is this important? Imagine you wanted to bring in a custom field into HubSpot but the table you chose in Snowflake didn’t have it. You would have to ping your not-so-friendly neighborhood data engineer (who has a backlog of tasks, none of which are related to MarOps). They would have to go into the database, manipulate the data to your liking, and ping you when complete. At a large enterprise, this process can take up to a week – just for 1 field. Now, imagine doing this every time you want to build an automation.

The Conversion AI SQL editor gives non-technical marketers the ability to write natural language queries to manipulate warehouse data and build custom fields, objects, and events. 

Imagine you had a table with millions of raw product events such as invited_teammate and project_created. Maybe you don’t want to trigger emails every time a teammate is invited but only when they’ve invited > 3 teammates. You can easily query in natural language “Create a separate event for when a company invites more than 3 teammates using the invited_teammate row”. The editor will write the SQL for you and create your event, like magic!

Summary

The HubSpot + Snowflake sync provides a rudimentary basis for mapping existing fields 1:1 from a specific Snowflake column to a field on a HubSpot object. This has extreme limitations when it comes to powering complex marketing automation campaigns centered around product activity, usage, and other warehouse objects. 

The Conversion sync allows you to manipulate your Snowflake data using AI to craft custom tables, fields, and columns within your Snowflake instance. This eliminates the data engineering bottleneck of continuous pinging other teams for support and new column or field creation.

Using your manipulated data, you can create complex data objects and relationships within Conversion which allow for powerful automation and personalization. You can create custom object records and tie contacts/companies to these objects as relationships, create custom events which hold custom properties, and create/update contacts.