Short answer: choosing a small-business CRM in 2026 is not a ranking problem, it’s a verification order. Five steps, in this sequence: (1) write down who uses it and what must be in it; (2) budget with the right denominators — per-seat, commitment, offer vs normal price; (3) verify each must-have at the entry tier, not the top tier; (4) check seat rules and data-export logistics before you commit; (5) run a structured trial with your own data. The expensive failures this checklist is built to prevent are the ones that skip steps 3–4: a plan that lacks the feature your team named — or locks your data behind export windows — was never actually cheaper.

Step 1 — Write the needs list before opening any pricing page

Two lists, in writing:

  • People: how many will actually log in weekly? (Not headcount — active users. Seats are usually billed per seat, and some vendors bill seats whether or not anyone uses them.)
  • Must-haves: name them as capabilities, not categories. “Email integration” is not a capability — “replies to customer emails log automatically against the contact” is. Also: pipeline stages, quotes/invoices, marketing sends, service tickets, reporting, mobile use, integrations with tools you already run.

Everything later in this checklist refers back to these two lists. If a list item can’t be traced to a real workflow (“we might need automations someday”), it’s a nice-to-have — keep it off the must-have list.

Step 2 — Budget with the right denominators

Per-seat monthly prices are the headline, but three denominators change the real number:

  • Commitment form. Annual billing usually means one upfront payment and a 12-month commitment; monthly billing can be month-to-month — or still carry an annual commitment while paid monthly. Ask which one you’re signing.
  • Offer vs normal price. Vendors run new-customer promotions. Example from the official pages we checked on 2026-10-24 (re-verified 2026-10-30 and 11-01): HubSpot’s entry tier — Starter (labeled “CRM Suite Starter” in our 2026-10-24 card capture; sold as a single “Starter” card since then) showed $7/seat/month with upfront payment and an annual commitment (or $10 month-to-month) against a normal price of $20/seat/month — offer terms “new customers only,” limited time, and renewal pricing not stated on the pages we checked. Salesforce Starter Suite showed a single $25/user/month “billed monthly or annually” (the page doesn’t state whether the effective rate differs by period). Pipedrive Lite showed $14/seat/month billed annually, with its monthly rate not verifiable in our check. Budget on the number you can confirm, not the number that’s discounted.
  • What sits outside the seat price. Implementation/onboarding fees (HubSpot requires a one-time $1,500 onboarding from the Professional tier of its Sales Hub line), add-ons (Pipedrive’s display at $13.33–$41 “starting from” on its US pricing page — the page states neither the unit nor the billing period; its KB says add-ons follow the plan’s billing cycle), premium support percentages (Salesforce Premier: 30% of net license fees), and taxes (Pipedrive excludes VAT; the pricing pages we checked for HubSpot and Salesforce don’t state tax treatment, though HubSpot’s Starter product page footer notes prices are “subject to applicable tax”).

Our published entry-level pricing checklist carries the full dated table for these three vendors — billing terms, seat rules, and the open questions each official page leaves unstated (snapshot 2026-10-24, re-verified 2026-10-30 and 11-01). Treat any price you read anywhere — including here — as needing re-verification the week you buy.

Step 3 — Verify each must-have at the entry tier

Vendors’ marketing pages describe features that exist somewhere in the lineup. The entry tier is what you’ll actually buy first. Official pages we checked on 2026-10-24 showed how differently three entry tiers handle one common must-have, email sync:

  • HubSpot Starter: includes email sync (the pricing feature table shows unbranded 1-to-1 email at Starter, and the Starter product page — re-verified 2026-11-01 — lists email sync among what Starter adds over the free tools). Caveat: users on free View-Only seats cannot connect personal email — a team of five that all need sync needs five paid seats.
  • Pipedrive Lite: the plan card lists pipeline management, AI report creation, 500+ integrations, and meeting intelligence — but full email sync starts at Growth ($39/seat/month annual): the Growth card is marked “Lite plan +” and lists “Full email sync with tracking” (“Fully sync Gmail, Outlook or your company email”), re-confirmed in rendered captures on 2026-10-30 and 11-01. If email sync is on your must-have list, Pipedrive’s entry price for you is the Growth price, not the Lite price.
  • Salesforce Starter Suite: the product page lists marketing email (2,000 sends/month, expandable) — but Gmail/Outlook inbox sync is not confirmed on the official pages we checked. Put “confirm inbox sync in a trial” on your evaluation checklist rather than assuming it.

The method generalizes: for each must-have on your list, find it in the vendor’s plan-comparison table at your tier (not a blog’s feature grid, not the top-tier marketing page), and note anything the vendor’s own pages don’t state — that’s what trials are for.

Step 4 — Check seat rules and data logistics before you commit

Two things are expensive to discover after signing:

  • Seat mechanics. Can you add seats mid-term and at what price? Can you remove them mid-term, or only at renewal? (HubSpot’s seat KB, for example, documents that seat counts can’t be reduced below the subscription’s minimum commitment or until the next renewal date.) Is billing strictly per seat? (Pipedrive’s KB: charged per seat “whether it’s assigned to a user or not.”)
  • Exit logistics. Before you commit, confirm you can get your data out on your terms. The vendors document very different export regimes: Salesforce deletes Data Export files 48 hours after completion and cannot recover them (and weekly exports exist only in Enterprise-and-above editions); Pipedrive keeps generated export files for 28 days, restricts exports to global admins or users with the right permissions (and only data visible to them), and requires linked activities, notes, and files to be exported separately; HubSpot’s export links expire after 30 days with a 300-exports-per-24-hours quota. If a vendor makes exit hard or undocumented, treat that as a selection criterion, not a footnote. Our published CRM data export and migration checklist covers this in detail.

Step 5 — Run a structured trial (not a demo)

The three vendors covered here all offer a free trial or free tier (Pipedrive: 14 days, no credit card; HubSpot and Salesforce have free tiers capped at 2 users). A trial is only informative if you structure it:

  1. Import a real sample (even 20–50 records of your own, with messy names, multiple currencies, and an odd stage or two) — not the demo data.
  2. Connect the tools on your must-have list (especially email) and run your real workflow for a week.
  3. Test the reports you actually present, not the built-in sample dashboards.
  4. Try exporting what you imported — from day one, before your data grows.
  5. Score against the written must-have list from Step 1, per seat cost from Step 2, and the seat/export facts from Step 4.

To make Step 5 concrete, a fill-in trial scorecard accompanies this checklist: twelve criteria mapped to Steps 1–4 (email sync, price form, seat rules, data exit, and more), with no vendor scores prefilled — download it, delete the rows that aren’t your must-haves, set your own weights, and score what you actually see in the trial.

What to ignore while you choose

  • Unsourced numbers — market-share percentages, “X% higher adoption,” user counts without a vendor citation, price comparisons without dates, and platform pages with no links to back their claims. If a claim matters to your decision, it should trace to a vendor’s own page or a named, dated source; otherwise it’s decoration.
  • Rankings. “Best CRM” lists — including ones with this word in the URL — can’t know your must-have list. Use them to build a candidate set of 3–4, then run Steps 1–5.
  • Feature counts. A “1,000+ integrations”-style count (illustrative — the number itself is the problem) says nothing about the two integrations you need.

Where to check prices and plans (official pages only)

Related on this site: entry-level pricing for a 5-person team (the dated three-vendor price table) · data export & migration checklist (export windows and cutover steps) · HubSpot vs Pipedrive (head-to-head for the two vendors this checklist uses as examples).

Disclosure: this page contains no affiliate links.