Written by Yuri Polukeev, CEO, BNTouch Updated August 14, 2026
Mortgage CRM reviews are worth reading, and the overall star rating is the least useful number on the page. The sub-scores, the review dates and the reviewer’s company size will tell you far more about whether a platform survives contact with your team than a 4.5 versus a 4.3 ever will.
Most buyers skim the headline rating, read the newest five-star review and the newest one-star review, then call it research. That finds the loudest opinions rather than the useful ones. Here is how to read a review page so the half hour you spend on it changes your shortlist.
The overall rating averages away the answer
Review directories such as Capterra and SoftwareAdvice do not ask reviewers for a single score. They ask for several, usually ease of use, customer support, value for money and functionality, and then publish a blended overall number on top of them. Two products can land on the same overall rating for opposite reasons: one is simple to use and thin on capability, the other does nearly everything and takes a month before anyone is comfortable in it.
The blended score hides which of those two you are looking at; the sub-score breakdown shows it immediately. Open it before you read a single written review, because it tells you what complaints to expect and what to ask on the demo.
Sub-score
What a weak score usually signals
Who should care most
Ease of use
The system does the job and the people using it fight it daily
Teams whose loan officers work without an admin nearby
Functionality
Real gaps, or capability that demos well and turns out shallow in month three
Shops with unusual workflows or heavy automation plans
Customer support
Slow replies, ticket queues, no named person who knows your account
Anyone migrating data or connecting an LOS
Value for money
The product works and buyers felt they paid for more than they used
Small teams buying per seat
A strong functionality score next to a weak ease-of-use score does not mean the product is bad, it means the product needs an owner inside your company. If nobody has time to be that person, the warning is about your situation.
A 2019 review describes software that no longer exists
Mortgage CRMs ship changes constantly, and a platform gets rebuilt in pieces over a few years. A complaint about a clumsy mobile experience from four years ago may have been fixed twice since, and praise for fast, personal support in an old review tells you little about the support team after an acquisition or a reorganisation.
Sort by most recent and read in that order. The past twelve months are evidence, one to two years back is context, and anything older belongs in the history section of your notes. What you are hunting for is a cliff, a stretch of warm reviews that turns cold in a particular quarter or the reverse, because that break usually lines up with something real. Ask the vendor what it was. A good answer names the change, and a vague answer is its own data point.
Feature complaints and adoption complaints are not the same problem
A feature complaint says the software will not do a specific thing: it will not split commissions the way we do them, it will not show partner referrals in a view we can sort, it will not push a status back to our point-of-sale. Those are the easiest to handle, because you can verify each one on a demo call and get a yes or a no.
An adoption complaint sounds different. It says setup took three times longer than planned, that the data came over messy and nobody ever cleaned it, that the loan officers were back in their spreadsheets by the second month, or that the campaigns got built once and never touched again. Read enough of them and you notice they rarely blame a missing button; they describe a project that lost momentum.
Those predict your own outcome far better, because the conditions behind them are conditions you also have: limited time, a team mid-pipeline, data spread across three places, and no obvious person to own the build. A feature gap you confirm before you sign, while an adoption failure shows up in month five with the annual invoice already paid. Read them closely for what they say about onboarding and training, the part of the purchase most buyers negotiate least and need most.
Check the seat the reviewer is sitting in
Directories publish company size and job role alongside most reviews, and both change what a review means for you. A 500-person lender cares about permissions across branches, provisioning users, and reporting rolled up by region. A solo originator cares whether past clients hear from him while he is out on a listing appointment. A one-star review about missing branch-level permissions is close to irrelevant to a two-person shop, and a five-star review praising how fast one person got running means little to a forty-officer operation.
Role matters as much as size. An operations manager reviews the build, a loan officer reviews the daily use, and an owner reviews the invoice against what the business got back. All three are legitimate and they answer different questions, so weight the reviews written from the seat closest to yours.
Reviews that came with a gift card
Incentivised reviews are common and mostly disclosed: directories run programs offering a small reward for a verified review, and vendors point customers at them. Incentivised is not the same as fake, and plenty of honest, detailed reviews get written for a gift card, so weight them accurately rather than throw them out. A burst of reviews landing in a narrow window after a long quiet stretch usually means a campaign. Reviews that praise every category evenly and repeat the product name in full tend to have been written with the vendor watching, and a cons field that says “none” will not help you either way.
What you want are reviews describing a specific situation: which system they came from, how long the migration took, what broke, what they use daily. Those are hard to fake.
What a run of support complaints really tells you
One bad support story is a bad week at the vendor, and every company has those. A pattern is different. When complaints repeat across several years, company sizes and roles, you are looking at something structural: an understaffed team, a queue with nobody named on your account, or an escalation path that only works if you know who to email.
Support matters most in the two windows where CRM projects die, implementation and the first real change after go-live, when an integration breaks or a key admin leaves. If the pattern shows slow response during implementation, get onboarding commitments in writing before you sign. While you are there, read how the vendor replies to critical reviews. A reply that names the problem and says what happened next is worth something; a templated apology under every complaint tells you how the company handles friction.
A checklist for reading any vendor’s review page
Open the sub-score breakdown first, and note which category is weakest.
Sort by most recent, and treat anything older than two years as history rather than evidence.
Look for a point where sentiment changes, then ask the vendor what happened.
Filter to your company size, and set aside reviews from operations ten times your size.
Separate feature complaints from adoption complaints, and take the adoption complaints more seriously.
Count how many reviews mention onboarding, migration or training by name.
Discount clustered, evenly glowing reviews with empty cons fields.
Check whether support complaints repeat across years and roles, or sit in one bad stretch.
Write down the three questions the reviews raised and take them into your demo.
Put the same test on us
BNTouch holds a 4.7 out of 5 across 1,300+ reviews, and the honest thing to say about that number is that it should not be why you buy anything. Go find the sub-scores, filter to shops your size, read the last twelve months and see what people say about getting set up. If the pattern holds, the next step is watching the platform run against your real pipeline, so request a demo and bring the three questions the reviews left you with.
Read mortgage CRM reviews for workflow fit, not a generic star average
A customer story is useful when it explains the operating context: who was using the CRM, what work they were trying to coordinate, and which workflow changed. A published BNTouch case study describes Cendera Funding’s historical experience with originator-level CRM use and corporate campaign control. It should be read as one customer’s experience, not as a promise of the same timeline or result for another mortgage company.
When evaluating any CRM review, ask whether the reviewer is like your team, what workflow they used, whether the feedback is dated, and whether you can find the original source. Those questions are more useful than a star average on its own.
Use customer statements as a starting point for your own workflow test
The BNTouch testimonials page includes named customer comments about CRM setup support, database work, partner coordination, communication, and task management. These are first-party customer statements, not an independent ranking or a promise that another mortgage team will have the same experience.
Read the original source, note the role and workflow being discussed, and ask to test the equivalent workflow in a demo. That keeps a review useful when your team is choosing a CRM.
