The bandaid doesn’t work
The CRM was invented to solve a management problem, not a selling problem. Once you see that, everything else about it makes sense — including why it disappoints you.
Go back to the original question that created the category. A sales organization grows past the point where one leader can hold it in their head. How does anyone know what fifty reps did this week? Which deals are real? What was promised to which customer? Before software, the answers lived in call sheets, Rolodexes, and hallway conversations. That does not scale.
The available answer, given the technology of the time, was a database. And a database has one non-negotiable requirement: something has to put the data in. In the 1990s, only one capture mechanism existed for sales activity — the reps themselves. So the design wrote itself. Sellers would do the selling, and then sellers would also do the record-keeping about the selling.
Management got visibility. Reps got a second job.
The CRM was never a good idea. It was the best idea available. The industry made its most expensive revenue-producing employees into part-time data-entry clerks, because no other clerk existed.
The tax was the design, not a side effect
The industry’s largest CRM vendor publishes an annual State of Sales study, and its own findings describe the cost of that design. The fifth edition found reps spend just 28 percent of their week actually selling. The most recent edition, published in 2026, still puts non-selling work at 60 percent of the rep’s week. Forrester tracked 3,031 sales reps across industries and found the average rep loses nearly two full days every week to administrative work alone.
Three decades of software iteration have not moved these numbers. That is the tell. The numbers are not a bug in the software. They are the labor model the software requires.
Why the data is bad — and why your reps are right
The second cost is quieter and worse. The data itself is bad, and it is bad for a rational reason. Every field a rep fills in is unpaid labor that pays the rep back nothing. The record does not help them close the deal in front of them. It helps management build a report.
A rep facing quota pressure will always — correctly — choose the customer call over the status field. Multiply that choice across a team and across a quarter, and the CRM becomes what every sales leader privately knows it is: a graveyard with a login page. Bad data isn’t a discipline problem. It’s a rational response to unpaid labor. Which is why enforcement has never fixed it, and never will.
The compromise carries three flaws no policy can remove. The data-entry tax is structural, so it always competes with selling — and selling always wins. The system records the past, so even perfect data tells a rep nothing about what to do next. And everything built on top of the record — the pipeline review, the forecast, the board deck — inherits whatever fiction sits underneath it.
The constraint just expired
Here is what changed. Software can now listen to a call and produce a structured record of it. It can read an inbox and a calendar and capture every contact and commitment without a keystroke. The execution layer no longer has to be human — which means the 1990s compromise no longer has to be accepted.
The full white paper — Your CRM Sucks: And It’s Not the Software — takes the argument all the way: the three systems your CRM conflates, the five fixes that never fix it, the accuracy loop that replaces enforcement, and the per-rep math on what the old design costs your team every year. Free download.
