Choosing a CRM: run the data-export and exit-cost tests first

Start with two tests, not a demo. Before you compare pipelines or per-seat prices, ask each CRM candidate to give you your data back and tell you what leaving will cost. A CRM that the team stops using is expensive; a CRM you cannot leave without costly manual work is expensive in a different way. Test both before you commit.

The data-export test and the 3-year exit-cost test are a professional methodology proposed by Niro Digital, not an externally validated benchmark. This article does not present a published Niro Digital CRM implementation case study or a measured migration outcome; the framework is guidance, not implementation-result evidence. Niro Digital builds custom CRMs and also implements CRMs; the method below works even if you never use us. Every list price and export term is dated as of August 2026.

Before comparing features, run a data-export test on each candidate as a precaution — how heavily it weighs depends on the candidate's documented export tooling and cancellation terms — and include field-mapping and re-entry in a 3-year total-cost calculation. Only after both gates pass do you score the survivors, run a two-week adoption pilot, and shortlist two or three tools to demo. We won't rank one CRM best for everyone.

CRM comparisons often start with dashboards, pipelines, and integrations. Those matter, but they answer the wrong question first. The first question is whether you can leave with your data intact, and what leaving costs. We treat data portability as a decision criterion because one of the company's published values is Radical Transparency: "No hidden fees, no padded metrics, and no ownership of your data. You own your ad accounts and your code." A CRM holds your deal history and customer contacts. If you cannot export those cleanly, you do not own them in any practical sense.

Gate 1 is the data-export test: record what format your contacts, deals, and custom fields come out in, and what limits apply. Gate 2 is the three-year exit-cost test: price the full cost of staying and leaving. Drop any candidate that fails either gate before you spend a day in demos.

Run the test on every candidate. For each tool, request or locate current official documentation that covers exporting contacts, deals, and custom fields. Record four things: the export format; whether field mappings and unique IDs survive; the frequency limits on manual and automatic exports; and what the vendor's terms say happens to your data after cancellation or non-payment. Then test restoration from the export file into a blank spreadsheet or another tool. Generating an export is not enough.

Salesforce shows why the frequency line matters. Salesforce's Data Export Service produces a set of CSV files, which the vendor describes as usable 'either for backup or for importing into a different system'. A manual export can run once every 7 days or every 29 days, and automatic exports can be scheduled at weekly or monthly intervals. Weekly exports are available only in Enterprise, Performance and Unlimited Editions; Professional and Developer Editions are limited to monthly intervals. A separate client-side tool, Data Loader, also exists. The current Data Export FAQ is dated 2026-04-02. Gearset and Horilla both describe the Data Export Service as the standard mechanism for backup and migration, and Gearset urges testing restoration from export files — not just generating them.

For HubSpot and Pipedrive, we do not have the same export evidence in this article, so we will not invent it. Run the same four checks against their current help-centre documentation and record the result in the table. A blank cell means you have not verified it yet.

Build the model in five line items: per-seat list price, mandatory onboarding fees, add-ons, migration labour, and exit cost. Use list prices, then add your own hourly rate for mapping and re-entry.

Use this fill-in formula for the licence line: three-year licence cost = seats × price per seat per billing period × billing periods in 36 months. Then complete the table:

Taxes and vendor-specific charges must be confirmed in the buyer's quote rather than assumed.

The list prices show why the billing term and onboarding fee matter. HubSpot Sales Hub has a Free plan at $0/month for up to 2 users, no credit card required. A six-rep team is past that limit. Sales Hub Professional starts at $90/mo/seat billed annually or $100/mo/seat billed monthly, and requires a mandatory one-time Professional Onboarding fee of $1,500. Sales Hub Enterprise starts at $150/mo/seat and requires a mandatory one-time Enterprise Onboarding fee of $3,500. The Starter tier is displayed on HubSpot's pricing page as "$7/mo/seat" alongside "$20/mo/seat"; the page presents $7 as the starting price with $20 shown against the billing-period selector. Treat that as an ambiguity to resolve with a quote, not a settled price.

Pipedrive Lite is US$14 per seat per month billed annually, one payment of US$168/seat/year. Pipedrive Growth is US$39 per seat per month billed annually, US$468/seat/year. There is no permanent free plan; every plan carries a 14-day trial with no credit card required. A third-party tracker, Axis Consulting, lists full annual prices of US$14/US$39/US$59/US$79 and monthly list prices of US$24/US$49/US$79/US$99.

Implementation labour is separate. Niro Digital publishes indicative price ranges for implementation work on its pricing page; the ranges reflect typical projects. Check that page before you budget setup and migration rather than quoting from memory. For migration labour, record your estimated hours for cleaning the spreadsheet, mapping each column to a CRM field, re-entering history that does not import cleanly, and verifying the result. Multiply by the real hourly cost of the person doing the work.

Exit cost is the line most comparison pages omit. It includes the labour of exporting and re-entering into the next tool, and any downtime while the switch happens. If you cannot export a usable file without manual rework, put a high number on this line and make it visible to your director. The exact post-cancellation terms for each vendor are not in this article; record them during Gate 1.

Custom software development — CRMs, dashboards, employee tools, logistics systems shaped around how the business actually works — is one of our four main services, alongside AI automation, recruiting advertising and lead generation advertising. Niro Digital designs and builds custom CRM systems and supports businesses with CRM implementation as one of its core services. Because we sit on both sides of that line, the rules below are written to work if you never call us.

We suggest preferring an off-the-shelf CRM when your deal stages match the standard pipeline the tool already assumes, the budget suits predictable per-seat licence costs, the existing integrations are enough, and a two-week pilot will tell you what you need.

Consider a custom build when the business runs on deal stages a generic tool would have to be forced to model, compliance or data-locality constraints limit SaaS, or deep integrations would require fighting the generic tool's data model. Custom changes the cost structure: instead of licence-plus-onboarding you carry build-plus-maintenance, so score it on the same criteria with the cost line replaced. Before you score the Custom build column, get a written estimate covering six items: scoped build price, expected delivery timeline, ongoing maintenance scope and price, integration work, data migration, and a named internal owner. Until that estimate is in writing, leave the column unscored. These are our decision rules, not a universal standard.

For the service itself, see custom software development.

Do this after both gates. Any candidate that failed the export or exit-cost gate is dropped before scoring, even if the dashboard impressed you.

Build the stage list from your current spreadsheet and inbox before scoring. Review how deals actually move: list the real states a deal passes through, define the event that moves a deal from one state to the next, and identify the required owner and next-follow-up field at each state. Use those defined stages as the evidence for the scorecard and for the pilot in the next section.

Fill in weights that sum to 100% for the criteria that matter to your team. Choose your own rows if the six suggested below miss something. Score each remaining candidate from 1 to 5 on each criterion, multiply each score by its weight as a decimal, and sum each column. The top two or three go to deep demo.

The most expensive CRM is the one nobody opens. Test adoption the way you tested portability: before committing, choose two or three reps who will actually use the tool. Import a small slice of real deals — not dummy data. Run your real stages through the pipeline for two weeks. Define what "used" means in advance: logged activity, updated stages, no spreadsheet fallback. If the tool cannot hold your actual pipeline without forcing reps to work around it, that is a failure.

Take the two or three candidates that passed both gates and scored highest on the weighted scorecard. For each survivor, write down the specific export evidence (format and frequency, with the vendor document date) and the three-year total cost you calculated. That pair is what a director can question and an implementer can validate.

Two limits apply. The data-export test and the 3-year exit-cost test are a professional methodology proposed by Niro Digital, not an externally validated benchmark. List prices and export terms are as of August 2026 and can change; re-verify them before you put a number in the budget.

At this point the next step is not another feature tour. It is validation and implementation planning: a consultant who does both off-the-shelf and custom work can stress-test your weights, plan the data migration, and produce a realistic implementation estimate whether the shortlist ends in an off-the-shelf tool or a custom build. Book an intro consultation about CRM implementation.