Attio Review: Which Published Limit Reaches You First

Attio's pricing page settles more fit questions in one table than a week of reading star ratings does. It publishes a record ceiling per plan, an object ceiling per plan, a monthly email allowance, a bulk-send cap, and two separate credit balances that meter the AI features. Those numbers are the shape of the product. A review that skips them and talks about how the interface feels is describing the first hour of use rather than the second year.

This is a read of what Attio publishes about itself, checked against the page that publishes it, and a judgement about which teams the published limits fit. The verdict sits at the end, and it depends on whether you are running a CRM or running outbound out of one.

What Attio actually is

Attio positions itself as a data-native CRM built around custom objects rather than a fixed lead and opportunity model. Records sync from email and calendar automatically, the data model is user-defined, and the automation layer sits on top as Workflows and Sequences. The developer surface is unusually broad for a CRM at this price point: attio.com/pricing lists API and webhook access, an App SDK and an MCP server on every plan including the free one, and the documentation host describes four developer entry points, the App SDK, the REST API, MCP and SQL queries against workspace data.

That last detail matters more than the marketing copy around it. A CRM that exposes SQL and an MCP server on the free tier is a CRM expecting to be automated against, which is a different design intent from a suite that gates its API behind an enterprise contract. The operational consequences of that API are covered on our Attio API page, which reads the rate limits and the complexity-scored list queries. This review deliberately does not repeat them.

The three limits that decide fit, and the one place to check the rates

Attio's own pricing page is the authority on what it costs, and the rates, the two credit meters and the overage blocks are read line by line in Attio pricing. That page also carries the two cautions that matter before any figure travels: the monthly and annual states render into the same document, and the prices are regional, so third-party dollar figures disagree with each other for honest reasons.

What a review can add is which of the published ceilings actually bind, and in what order. Three of them do the work.

Seats gate the plan, not just the bill. The free tier stops at three seats and Plus stops at ten. A team of eleven is on Pro whatever its feature needs are, so headcount rather than sophistication is what moves most companies up the ladder, and it moves them on a schedule hiring decides.

Objects are the real design limit. Attio's appeal is modelling your business rather than accepting somebody else's schema, and the published object ceilings run three, five, twelve and unlimited across the four plans. Every custom thing you want to track, partners, properties, venues, deals of a second shape, spends one. A modelling-first CRM that caps modelling at five objects on its small-team plan is a constraint to check against your intended schema before migration rather than after, because the schema is the thing you migrated for.

Records are generous, and generous is not infinite. The published ceilings are 50,000, 250,000 and one million below Enterprise. For a customer database those are comfortable. For a prospecting database they are not, and that gap is a hint about where this tool's boundary sits.

1. Step 1Seats

   Driven by hiring rather than by any product decision, so it arrives on a schedule you do not control

2. Step 2Permissions

   The first time a contractor, partner or departing rep needs a bounded view

3. Step 3Objects

   When the schema you migrated for outgrows the plan that allowed it

4. Step 4Workspace credits

   When automation runs against records rather than against people

5. Step 5Records

   Which a customer database, as opposed to a prospecting database, rarely reaches

Where Attio stops being the answer for outbound

Attio publishes an email allowance per plan: 200 emails a month on Free, 1,000 a month on Plus, unlimited on Pro and Enterprise. It also publishes a bulk-send cap of ten at a time on the free plan, rising to fifty at a time on every paid plan. Email and calendar sync is one account per user on Free and Plus, two on Pro, three or more on Enterprise.

Read those two lines together and the boundary is unambiguous. Fifty at a time, sent from a user's own connected mailbox, is a relationship-management feature. It is not a cold outbound channel, and it is not designed to be one. Sending cold volume through the same mailbox that holds your customer conversations puts the domain that carries your renewals behind the deliverability of your prospecting, which is the failure this architecture exists to prevent.

The correct division of labour is a CRM that owns the record and a separate sending system that owns the campaign, with a defined rule about which one writes which field. That rule is the whole integration, and the way to settle it is covered in Pipedrive Integrations, whose argument is CRM-agnostic. On the sending side, the category distinction between a sales engagement platform and a lightweight outbound tool is laid out in Sales Engagement Platforms.

House practice sits on the same line. Campaigns run on dedicated sending infrastructure, one message per campaign, with re-approach handled as a new campaign on a new signal rather than as follow-up steps stacked on the original send. A CRM's bulk-send button does not change that, and a CRM with a generous email allowance does not become a sending platform.

Reporting, permissions and the two things that move upmarket

The published report ceilings are three, fifteen, one hundred and more than one hundred. Access permissions are listed as none on Free, basic on Plus, and advanced on Pro and Enterprise. Teams are capped at ten on Plus and unlimited above it. Single sign-on and SCIM sit on Enterprise.

The permissions ladder is the one to check against your actual situation. A CRM with no access permissions is fine for three people who all see everything and becomes a problem the first time a contractor, a partner or a departing rep needs a bounded view. That transition tends to arrive earlier than the seat count that would have forced a plan change anyway.

Check these against your own workspace before migrating

• ~Depends: Count the custom objects your intended schema needs against the plan's object ceiling
• ~Depends: Model monthly workspace-credit consumption for automations that run on records rather than on people
• ~Depends: Confirm the seat count you will reach in twelve months, since seats gate the plan
• ✓Yes: Decide which system owns each shared field before connecting anything
• ✓Yes: Keep cold sending on separate infrastructure rather than on connected user mailboxes
• ~Depends: Check whether you need bounded access permissions before the plan that carries them

What the review corpus is actually made of

The third-party writing on Attio splits into three kinds, and they are worth different amounts.

Operator write-ups, the ones where somebody sets up a workspace and reports what broke, carry the most. Attio's review results are unusually rich in these: forum threads comparing it against another modern CRM, individual practitioners publishing what they found after weeks of real use, and a small number of technical write-ups from integration vendors. What makes them useful is that they name the specific thing that stopped working, which is a claim you can test against your own requirements.

Rating aggregators carry the least, for the reason every aggregate carries least: they average over people whose jobs are not yours. A vendor-directory score is a popularity measurement with a review-collection incentive behind it, and it moves on marketing effort as much as on product quality.

The third kind is the comparison page published by a company selling something adjacent, usually a sync tool, a migration service or a competing CRM. These are often factually careful and structurally incapable of concluding against their own interest. Read them for the facts they check and never for the verdict they reach.

What a review cannot tell you, including this one

Star ratings aggregate strangers whose jobs are not yours. A five-star review from a two-person consultancy modelling three objects and a two-star review from a fifty-seat team that hit the permissions ceiling are both accurate reports of different products. The published limit table is the part that is the same for everybody, which is why this review is mostly a reading of it.

Two things the table cannot settle. The first is migration cost, which is dominated by your data quality rather than by either vendor's tooling: duplicate records, inconsistent company naming and half-finished custom fields all survive an import and become somebody's afternoon. The second is whether a flexible data model is an asset or a liability on your team. A CRM that lets you model anything will be modelled by whoever gets there first, and an unowned schema drifts. That is a governance question rather than a product one, and it belongs to the same function that owns definitions and data quality, described in Revenue Operations.

The short version

Attio fits a team that wants to model its own business, automate against a real API, and keep its CRM and its sending stack as separate systems. It fits less well where a fixed pipeline model would do, where the object ceiling on the small-team plan collides with the intended schema, or where somebody expects the CRM to be the outbound channel.

Judge it on four published numbers rather than on adjectives: the object ceiling, the workspace-credit allowance, the seat cap, and the point on the ladder where access permissions appear. If those four clear your requirements, the interface reviews are worth reading. If any of them does not, no amount of praise for the interface changes the arithmetic.

If the reason you are shopping for a CRM is that outbound is not producing enough conversations to be worth tracking, the CRM is not the constraint. See what a campaign built on your ICP would look like before rebuilding the system that records its output.

Pricing and features verified as of August 2026. Verify current terms with the vendor before relying on them.