CRMs you can export your data from

EspoCRM has been around long enough to have stopped chasing fashions, and its pricing reflects a philosophy the rest of this category has largely abandoned: you get everything, and what you pay for is scale. All three hosted tiers advertise the full feature set and every extension. What separates Basic from Ultimate is file storage and how many records you are allowed to keep.

That makes the arithmetic unusually simple. Basic is US$15 per user per month with a hundred thousand records and a three-user minimum; Enterprise is US$25 with ten million; Ultimate is US$69 with no limit. Underneath sits the open-source edition, which you can install on your own server and run without a licence fee.

The cost is the seat minimums, which make this a poor answer for somebody working alone: the cheapest way into the vendor's cloud is three seats whatever your headcount. Self-hosting removes that and replaces it with an operational burden you carry yourself. Integrations are respectable rather than remarkable, and an external API is on every tier. On data control it is as strong as software gets, because you run it on your own hardware, and the value case is simple arithmetic.

Bigin exists because its own vendor's flagship CRM is too much for most small businesses, and the company decided to say so with a product rather than a discount. What you get is a pipeline, contacts, activities and the integrations of a very large platform, packaged into something a two-person business can set up in an afternoon.

On setup and value it is hard to argue with. There is a free tier, Express costs US$7 a seat each month on annual billing and Premier US$12 — figures that sit at the bottom of this field while still including Gmail and Outlook integration, automation, a developer API and documented data migration. Very little here is withheld to make a cheaper plan feel thin.

The cost is deliberate shallowness. This is a product designed not to grow past a certain size, and the honest consequence is that outgrowing it means changing products rather than changing tiers. For a business that will always be small that is a feature; for one that expects to be forty people in three years it is a migration scheduled in advance.

What defines this product is a pricing decision rather than a technical one. There is one plan at US$15 per user per month, it includes everything the company makes, and there is no tier above it holding the useful parts hostage. Unlimited contacts and companies, unlimited pipelines, custom fields, user permissions and support are all simply in the price.

That makes setup and value unusually easy to reason about. You do not have to model which tier you will land on in eighteen months, because there is only one, and the trial runs thirty days without asking for a card. For a very small business that has been burned by a ladder priced to be climbed, the absence of a ladder is the feature being sold.

What you give up is depth. Automations and an API both exist and are documented, so integrations are a real possibility rather than a dead end, but the product is a well-kept contact database with pipelines attached rather than a sales platform. Because there is only one plan, whatever is missing is missing at every price — there is no tier to climb to when you outgrow it.

OnePageCRM is built on an argument about behaviour rather than about features. Every contact record carries a next action — one thing, with a date — and the interface is a queue of those actions rather than a board of deals. If your problem is that follow-ups quietly stop happening, the product is designed to make that failure visible.

On setup and value it is among the easier decisions here. Pricing starts at US$9.95 per user per month, records are unlimited, and the vendor states there are no paid add-ons, so the figure on the pricing page is the figure on the invoice. Email sync, click-to-call, workflow automation, forecasting and a published API are all part of it rather than upsells.

The tradeoff is the method itself. A product with an opinion this strong is excellent when the opinion matches how you work and irritating when it does not — a team running long, multi-stakeholder deals with no single obvious next step will spend its time fighting the queue. The pipeline tooling is also lighter than the products aimed at larger sales teams, which is a consequence of the same design choice rather than an oversight.

Odoo draws its free line somewhere nobody else does. Rather than capping users or records, it gives you one application at no cost for an unlimited number of people. If the application you pick is the CRM, a company of forty can run its pipeline on Odoo Online without paying anything, which is a genuinely different offer from the two-user free tiers elsewhere in this field.

Above that, pricing is per user for all applications together: EUR 19.90 a month on yearly billing for Standard, or EUR 29.90 for Custom, which adds on-premise deployment, the Studio builder, multi-company support and an external API. The core is open source, so self-hosting is a real route and the data can sit on machines you control.

On value that is hard to argue with: an unlimited-user CRM at no cost. The catch is that the free tier is a doorway rather than a destination. The moment a second app is useful — invoicing, helpdesk, projects, all of which sit right there — the price becomes per user for everything, and the step is deliberate. There is also the configuration burden that comes with any suite: this is more product than a small sales team needs, and making it fit takes work that a single-purpose CRM does not ask for.

Twenty is the open-source CRM that most obviously wants to be used by people who write software. It is licensed under the AGPL, the repository is busy, and the documented way to run it yourself is a Docker Compose file rather than a support ticket. If your objection to this category is that your customer data lives on somebody else's machine, this is the entry that answers it directly.

The pricing inverts the usual shape. Full customisation, custom apps and workflow automation are all on the cheapest paid tier at US$9 per user. What the second tier at US$19 adds is governance rather than capability: row-level permissions, single sign-on, and a custom domain.

On scale headroom it is better equipped than its age suggests, with row-level permissions, single sign-on and an audit log. The tradeoff is the one you would expect of a young project. Integrations are thin — the catalogue does not approach what a decade-old vendor accumulates — self-hosting is only free if your time is, and above the two published tiers sits an enterprise arrangement starting at fifty thousand dollars that is negotiated rather than listed. What you get in exchange is unusual control: run it on hardware you own, read exactly what it does with your records, and take the database with you whenever you want.

Heli is a deliberately small product and makes no attempt to be otherwise. It is a contact list with a memory: people and the companies they work for, a log of every call and email, and projects you can move those people through in stages. The interface is built for the keyboard, with single-key navigation and a shortcut for everything.

The unusual part is where the data sits. You choose a region — the US, the EU, Asia-Pacific — or you keep it on hardware you own, which the AGPL-3.0 licence and the public repository make a real option rather than a promise. Full CSV export is documented, and the vendor states there is no tracking, no advertising and no third-party access.

The cost is scale and evidence in roughly equal measure. There is no team tooling to speak of, no published catalogue of integrations, and nothing approaching the automation that larger products sell. The hosted service is documented as free for one person during beta, and what it costs after that is not published, which is the one place this entry falls short of the standard this site applies everywhere else. And it is new enough that almost every claim above rests on the vendor's own documentation, with no independent coverage to check it against.

Attio starts from the observation that most CRMs decide in advance what a customer relationship looks like, and that the shape rarely fits. Its answer is a flexible data model: objects you define, attributes you choose, and relationships between them that the product does not second-guess. For a company selling something that is not a straightforward B2B deal, that is the whole appeal.

The pricing follows the same instinct. There is a free tier with no expiry, Plus at US$35 per user per month and Pro at US$79, and the things that make the product interesting — custom objects, automation, single sign-on — are not quarantined at the top. A published migration service exists for bringing records across.

The tradeoff is that you are doing the modelling. A product that will not decide what a deal is has, by construction, handed that decision to you, and a team that wanted a pipeline out of the box will spend its first week designing one instead. The entry price above the free tier is also toward the upper half of this field, and the top tier is quoted rather than listed.

Pipedrive is built around one idea carried further than most: a sales pipeline is a board, and everything else is subordinate to moving cards across it. That focus is why the cheapest plan feels usable rather than crippled. You get leads, deals, contacts and a calendar in one workspace, and you can bring your existing records in from a spreadsheet or another system without paying more.

The cost of that focus appears on the second tier. Full mailbox sync, workflow automation and forecast reports are all held back to Growth, which more than doubles the per-seat price. A team that wants its email in the CRM — which is most teams — is not really choosing the entry plan, and it is worth pricing the tier you will actually end up on.

Scale headroom is the quiet strength: an administrator gets role-based access and a change record, and no ceiling on records is published anywhere. What holds up well is the surrounding infrastructure. The API is documented endpoint by endpoint and readable before you buy, the catalogue of integrations is genuinely large, and pricing is published in full at every tier rather than hidden behind a form. Data control is ordinary rather than bad: a published privacy notice and a separate processing agreement, but no choice of where your records are handled.

Close is built for a specific job: outbound sales where somebody picks up the phone. Most of this category treats telephony as an integration you buy separately and wire in. Here the dialler, the call recording and the email sequences ship inside it, which means a rep works a list without leaving the record they are working.

The ladder is published in full and unusually long at the bottom — Solo/Essentials at US$9 a seat on yearly terms, then Growth at US$35, Scale at US$99 and Enterprise at US$139. Two-way email sync, custom objects and single sign-on all appear rather than being reserved for a negotiation.

The catch is that native calling is not free calling. Minutes are metered apart from the licence, and the free trial carries only five dollars of calling credit, so the number on the pricing page is a floor rather than a bill. For a team that lives on the phone the arithmetic still works; for a team that mostly emails, a large part of what you are paying for goes unused.