Close CRM Buyer's Checklist: Five Contradictions and Six Questions

A buying checklist earns its place by asking what the vendor's pages do not answer. This one is built from reading Close's pricing page, four of its support articles, its published file for AI systems and all seventy-five of its Trustpilot reviews, and it is ordered by what each item can cost you.

OUR TAKE: Three things decide this purchase and none is a feature. Which tier has your automation, what your telephony will actually cost, and what your real AI credit ceiling is at your seat count. Settle those and the rest is detail.

## Before you talk to anyone

Decide whether you need workflows. They do not exist on the two cheaper tiers^[1]. If automation is why you are buying a CRM, your entry price is $99 a user per month, not $9 or $35.

Count the people who will use it daily. There is no published seat minimum, and the trial has no user cap, so the number is yours to establish rather than to guess.

Measure your call volume. Hours per person per day, and average call length. Telephony is billed separately at about two cents a minute with every call rounded up to a whole minute^[1][4]. A team of eight on the phone two hours a day spends roughly as much on calls as on half the licence.

Count your custom fields. The cap is 250 on every tier including the most expensive^[1]. No upgrade raises it.

Count the people who will send email from their own address. Connected accounts cap at three on the lower tiers and ten above them^[1]. Above ten, this is a wall.

## The three numbers that make or break the budget

Before the detail, the shape. A Close CRM budget has three lines and most quotes show one.

The licence, which is predictable and is the number everyone compares.

The telephony, which is proportional to how much your team talks and is billed at the carrier's cost rather than marked up^[1]. On a calling team it is roughly half the licence again, and on a heavy outbound team it can match it.

The credits, which are included up to a ceiling and metered beyond it, with no spending cap^[3].

A budget that shows only the first is not conservative, it is incomplete. Build all three or expect the gap to arrive as a surprise in month two.

Questions to put in writing

What is our included AI credit ceiling at our seat count? The answer should be one number for the account. The per-user scaling stops at ten users, and that ceiling appears only in the support documentation^[3].

What is the consumption rate per operation, in credits? The vendor publishes an allowance and not a rate, so two accounts on the same tier can consume wildly differently.

Can a spending limit be set? The documentation says there is no global cap and no per-agent cap^[3]. Ask what alerting exists instead.

Is there a seat minimum on our tier? None is published anywhere we could read. Silence rather than a stated absence.

What is the ten-seat discount worth? The vendor states that a discount exists for ten or more seats on a twelve-month commitment and does not publish its size^[2].

Is the transcription fee per organization or per billing account? The pricing page says one thing and the documentation says the other^[4], and for a multi-entity customer the difference is real money.

## The two mistakes that cost the most

Both are made before anyone opens the product, and both are reversible only slowly.

Buying the wrong tier to save money. The two cheaper tiers have no workflows^[1]. A team that buys the second tier to be prudent spends a quarter doing follow-up by hand, concludes the CRM is not helping, and either upgrades or leaves. The saving was $64 a seat and the cost was a quarter.

Budgeting the licence and not the usage. On a calling team the telephony is roughly half the licence again, and above ten seats the credits stop being included^[3]. A budget built on the per-seat rate alone is wrong by a margin that shows up in month two and gets blamed on the vendor.

Neither is a trap. Both are published, one on the pricing page and one in the support documentation. They are only expensive if nobody reads the second.

Verify during the trial, not after

Make real calls for two days and read the meter. Then multiply by ten working days. That is your telephony bill and nothing on the pricing page provides it.

Run the AI features you expect to use and read the credit balance. Same arithmetic, different meter.

Import a representative slice of your data and watch the field count. 250 is the ceiling and it is per account rather than per pipeline.

Have the people who will use it build their own view. Setup averages 4.9 stars in the review corpus, so if your team struggles in week one the problem is likely your process rather than the product, and that is worth discovering before you buy.

Test the downgrade path mentally. Moving to a cheaper tier is blocked until usage falls below that tier's limits^[2], so a decision to over-buy is harder to reverse than it looks.

## What a quote should contain

If a quote arrives with one number on it, send it back.

A complete quote for this product has four lines. The licence, at a named tier and a named billing term. An estimate of telephony based on your stated call volume, with the per-minute rounding acknowledged. Your included AI credit ceiling as an account figure. And any add-ons, named, with the transcription fee identified as per account rather than per organization.

Four lines is not an unreasonable thing to ask for. It is what the product actually bills, and a vendor who will not break it out is a vendor whose invoice will surprise you.

If the quote includes a ten-seat discount, get the percentage in writing rather than the resulting figure, so you can check it holds at renewal.

The five contradictions to raise with the vendor

All five are between the vendor's own pages, and each is settled by one email.

The AI allowance. Advertised per user on the pricing page, capped at ten users in the documentation^[3].

Unlimited contacts. Listed as something the second tier adds, while the same page's matrix ticks it on the entry tier too^[1]. The real difference is leads alone.

Data enrichment. Ticked as included on all four tiers, metered at five cents per field each time in the documentation, with the vendor's own warning about bulk use^[4].

The annual badge. Up to 50%, against a real range from 52.6% at the bottom to 6.7% at the top^[1].

The transcription fee. Per organization on the pricing page, per billing account in the documentation^[4].

## What to check on the invoice, not the quote

The first invoice is the only document that tells you the truth, and three lines on it deserve a look.

The telephony total against your measured trial figure. A large gap usually means call rounding rather than more calling, and that is worth knowing in month one rather than month six.

The credit line, if there is one. If you are paying for credits you expected to be included, you are above the ten-user ceiling and nobody told you^[3].

And any enrichment charge, which will not be labelled as a surprise because the interface presents the feature as included^[4].

Check those three on the first invoice and you will not need to check them again.

Red flags that should slow the purchase down

A quote built on the entry tier for a team. It allows one user.

A budget that counts the licence and not the telephony. On a calling team the licence is roughly half the bill.

An assumption that the AI allowance scales with headcount. Above ten seats it does not.

And annual billing taken for the discount on the upper tiers, where it is 9.2% and 6.7% against a monthly option with no contract at all^[2].

What this product does well, so you know what you are protecting

A checklist reads as a list of problems, so the other half deserves stating.

Setup is the highest-rated theme in the review corpus at 4.9 stars. The interface sits at 4.1 across 14% of reviews. Complexity is raised by 3% of reviewers, who rate it 5.0. Price sentiment is 4.3, the only positive price sentiment we have recorded in this category.

Telephony is passed through at carrier cost rather than marked up, which is rare and worth saying plainly. Monthly billing carries no contract, there is a thirty-day refund window, and contracts are raised by exactly one reviewer in the corpus.

And the vendor publishes a dated plain-text file for AI systems carrying every price verbatim, which is the most accurate description of its commercial terms anywhere on its site.

If you are already a customer

Four checks, in descending order of what they tend to return.

Your seat count against ten. Above it, the included credit allowance is a fixed account number and you have been buying top-ups you assumed were included^[3]. This is the single most common overspend on this product.

Average call length against the rounding. Under a minute, and you are paying a premium set by a billing rule. The fix is operational.

Enrichment usage. Metered at five cents a field every time, presented as an included feature^[4]. Bulk actions across a list are where it bites.

Your tier against your actual use. If you are on the fourth tier for a feature you no longer use, remember the downgrade is blocked until usage falls below the lower tier's limits^[2], so plan it rather than attempting it.

And one for renewal: if you have passed ten seats, ask what the twelve-month discount is worth. It exists and its size is not published^[2].

What good looks like after three months

A short list, because a checklist should say what success is rather than only what to avoid.

Your monthly bill matches the model you built during the trial, within about ten per cent. If it does not, the gap is almost always telephony rounding or credits above the ceiling.

Your team is in the product daily without anyone chasing them. This is the thing reviewers consistently report and the main reason to choose this over a general CRM.

You have not hit 250 custom fields, and you know how close you are.

And you know, as a number, what a call costs you. Most companies using a CRM with a phone in it never work this out, and it is the one metric this product makes available.

## The one-page summary

Buy the tier that has your workflows. Measure the phone before you sign. Model the credits at the account ceiling rather than the per-user figure. Take annual billing only on the cheaper tiers.

Then ask the five questions above in writing, and keep the answers.