Segment vs mParticle

Customer Data Platform Comparison

Both products sit between your sources and your destinations. A tracking call from the web app, an event from a backend job and a row from the warehouse all land in the platform, get tied to a person, and are forwarded to analytics, email, ads and a CRM. The differences that matter to a B2B team are how the platform counts what you send, how it treats accounts as well as individuals, and what a privacy or legal reviewer sees when they ask where consent lives.

Segment: the documented strengths

• Accounts as well as users: The identity resolution overview lists support for user-to-account relationships and says it generates a graph of relationships between users and accounts for B2B companies. Sending the Group call lets you see behavior at the account level.
• Identity controls: Custom external IDs, customizable ID rules that enforce uniqueness on chosen identifiers, and merge protection that deals with non-unique anonymous IDs. Identifiers are treated as case insensitive.
• A warehouse is optional: The Segment pricing page answers the question directly: it works with sources you already have, and when you do connect a warehouse you can enrich profiles from it and sync resolved profiles back.
• Reverse ETL: The Reverse ETL documentation says it extracts data from the warehouse with a query you provide and syncs it to third-party destinations, covering event and object data such as subscriptions and product tables.
• Workspace terms: Seats are unlimited on both plans, payment is by invoice, HIPAA eligibility is listed as available, and Protocols for data governance is an add-on.

mParticle: the documented strengths

The mParticle platform page calls it a hybrid CDP that activates data in real time or from your warehouse, and lists IDSync, a Profile API, a Data Catalog, real-time audiences and 300+ integrations. Three things stand out in its documentation. Identity resolution is configurable per business case, with strategies for funnels, links, isolation and no-login products. Consent is stored as state on each profile. And Warehouse Sync, which the guide calls its reverse-ETL solution, pulls user and event data from Redshift, BigQuery, Snowflake or Databricks through a SQL data model you write.

The same platform page frames the product as designed for multi-channel consumer brands, and its FAQ describes enterprise consumer brands as the customer. That does not rule out a software company, but it means a B2B buyer should ask for reference customers and ask how an account or workspace entity is represented before spending time on a pilot.

Billing mechanics, using one B2B product as the example

Segment's MTU documentation says its usage model bills against API calls, monthly tracked users and, on some plans, throughput. An MTU is the number of unique userIds plus the unique anonymousIds that were never tied to a userId, counted across all sources in a billing month. The docs give a throughput example of 250 calls and objects per MTU, which on a 10,000 MTU plan allows 2.5 million a month.

mParticle's usage documentation measures something else: events, in units of one million, in three tiers (Connect, Preserve, Personalize), plus add-on meters for real-time products, retention, hosted rules and replays, all drawn from credits you commit to in advance.

Take a product with 3,000 signed-in users a month, each sending 150 events, plus 40,000 anonymous marketing-site visitors who never log in and send 5 events each. These counts are assumptions, so substitute your own export.

• Segment: 3,000 + 40,000 = 43,000 MTUs. Public traffic dominates the count, so a busy blog or docs site raises the bill faster than product growth.
• mParticle: (3,000 x 150) + (40,000 x 5) = 450,000 + 200,000 = 650,000 events, which is below one 1M unit. How a partial unit is rounded is a contract term, so put it in your quote request.
• RudderStack, as a yardstick: 650,000 events sits under the 1M step of its printed ladder, where Growth is $265 a month, or $265 x 12 = $3,180 a year. Billed annually it is $225 a month, or $225 x 12 = $2,700.

Now change the product: 20,000 signed-in users at 150 events each, with no anonymous traffic, is 3,000,000 events, or three mParticle units, against 20,000 Segment MTUs. That is 150 events per MTU, where the first example had 650,000 / 43,000, about 15. The higher that ratio, the more an event meter costs relative to a people meter, and a content-heavy site with little logged-in activity sits at the other end. Neither vendor gives a number, so a quote on your own export is the only way to know which meter is cheaper for you.

Identity and consent: where the two designs diverge

mParticle's IDSync components page says each request resolves to a single profile through one of four endpoints (Identify, Search, Login, Logout) and that an identity priority list decides matches. Its ranking advice is concrete: put immutable customer IDs first, then unique and login identifiers, then mutable ones such as email, and anonymous cookies and device IDs after those. Segment's graph is built around userId, anonymousId and groupId plus custom external IDs, with rules you set in Unify and a persistent ID that several external IDs map to.

Consent differs in shape. Segment's consent management overview says it integrates with your consent management platform, requires the consent object on all events, and sends events only to destination categories the person consented to. It adds two cautions that matter to an email-heavy B2B team: the feature is available on Business tier plans, and Segment does not recommend using it for communication opt-ins and opt-outs. Keep email subscription state in your email tool.

mParticle's data privacy controls guide keeps a consent_state record on the profile: regulation, purpose, a consented flag and a timestamp, where a newer timestamp replaces an older one. Forwarding rules and audience consent filters read it, and OneTrust consent states can be ingested. If your legal team asks "what did this person agree to, and when", mParticle's model answers from the profile, while Segment's answers from the event stream.

Who should choose what

Choose Segment when: your product is sold to companies and you need users linked to accounts. Your engineers already know the Identify, Track and Group calls. You want a plan without a warehouse and unlimited teammates. You want the option to add Unify and Engage later without changing vendors.

Choose mParticle when: mobile and web apps share people and the merge must happen the moment someone logs in. You need to pick between strategies such as profile conversion and profile isolation. Your privacy team wants consent state on the profile and rules that block outputs. You are comfortable with a credit commitment and a contract negotiation.

Choose neither yet when: you send under 250K events a month and no one has asked for identity resolution. RudderStack's free plan covers that volume at $0, which beats waiting on two quotes.

For wider context, the customer data platforms category, the B2B customer data platforms guide and the Segment vs Hightouch comparison cover the surrounding tools.

Moving from Segment to mParticle

mParticle publishes the path, which is its strongest switching argument. Its migration overview says the APIs and event models are similar to Segment's, and that a Segment Migration Professional Services package exists for teams that want help.

1. Decide between a client-side and a server-side implementation, since the guide splits the two. Endpoints in the guide are for the US1 pod; other pods have their own.
2. In mParticle open Setup, then Inputs, then Platforms, and issue a key and secret for each platform that sends data.
3. Swap the SDK. The client-side guide shows the package change for iOS, Android and web side by side.
4. Translate calls using the guide's migration reference, which maps Segment identify, track and related calls to mParticle events.
5. Choose an IDSync strategy and priority list before traffic flows, then import consent as purposes and consent_state records.
6. Run both pipelines on one source, compare counts, and remove the Segment library last.

Moving from mParticle to Segment

We found no vendor guide for this direction, so treat the list as a plan assembled from both vendors' documents.

1. Export your mParticle identity priority list and inventory every output, rule and audience.
2. Map events onto the six spec calls. Account traits move to Group calls, a call type with no counterpart named on the mParticle pages we read.
3. Recreate identity as ID rules in Unify, giving each external ID the uniqueness setting its mParticle priority implied. This needs the Customer Data Platform plan, since the pricing page puts identity-resolved profiles there.
4. Rebuild consent: mParticle consent_state records become the consent object on events plus a consent management platform, and confirm which plan includes it.
5. Replace Warehouse Sync data models with Reverse ETL queries, one destination at a time.

FAQ

Does either vendor publish prices?

No. Segment's pricing page names a Customer Data Pipeline plan and a Customer Data Platform plan, both priced on monthly tracked users with custom volume and invoice payment. mParticle's pricing URL opens a contact form, and its documentation describes credits bought through a commitment.

Which is cheaper for a B2B site with lots of public traffic?

Probably the one that does not count anonymous visitors. Segment counts anonymous IDs that never log in as MTUs, while mParticle counts events, so a marketing site with many casual visitors and few events per visit can favor mParticle. A product with chatty logged-in users can favor Segment. Quotes on the same export settle it.

Can Segment handle consent for email opt-outs?

Its documentation recommends consent management for cookie and data collection preferences and advises against using it for communication opt-ins and opt-outs. Keep subscription status in your email platform and let the CDP carry that field as a trait.

Is a data warehouse required?

Segment's pricing page says no. A warehouse is optional, though both plans offer warehouse interoperability. mParticle's Warehouse Sync reads from a supported warehouse, but the real-time pipeline runs without one.

What if I want a printed price instead?

RudderStack lists a free plan with 250K events a month and Growth steps from $265 at 1M events. See the mParticle vs RudderStack and Segment vs RudderStack comparisons.