CRM is changing faster than most businesses are moving. Here is why the gap matters, what the destination looks like, and why the work of getting there is more important than ever.
–
By Will Sawney, Founder — Sideways CRM
· April 2026 · 5 min read
–
Most of the leaders I speak to have the same quiet acknowledgement somewhere in the conversation. Their CRM is a mess. It started clean — or at least manageable — and over time it accumulated the scar tissue of every team change, every process tweak, every integration that was added and never fully thought through. It still functions. But it is not working. And they are aware, increasingly acutely, that the gap between where they are and where the technology is heading is widening every month.
This is not a new problem. CRM systems have always decayed without care. But something has shifted in the last eighteen months that changes the stakes. It is no longer just about efficiency or having clean data for reporting. The tools are becoming genuinely intelligent — and an intelligent tool built on a weak foundation does not just underperform. It compounds the mess.
The window to get the foundation right is now. Not because there is a deadline, but because the compounding works in both directions.
What is actually changing
For a long time, CRM sat in a particular category of software: important, necessary, and fundamentally passive. You put data in. You got data out. The intelligence was yours — the system was just the container.
That model is ending. The next generation of CRM is being designed not as a system of record but as a system of action. Platforms like Attio are building towards something qualitatively different: a CRM that understands both structured data (your pipeline stages, your deal values, your custom fields) and unstructured data (the email thread, the call recording, the tone of the conversation) — and that can reason across all of it to surface insights, trigger actions, and automate processes that previously required human judgment at every step.
This is not a feature update. It is a different kind of tool. The analogy that keeps surfacing is the shift Salesforce represented when it moved CRM from on-premise to the cloud. That was also framed as a technology upgrade. It turned out to be a generational change in how businesses managed their revenue operations. What is happening now is that kind of shift, compressed into a much shorter window.
The particular difficulty for established businesses
Startups have a different problem. They are choosing a CRM for the first time, often for a go-to-market motion that is still forming. The stakes are real but the friction is lower — there is no legacy to migrate, no embedded habit to shift.
Established businesses face something harder. They have history. A data model that was built for how they worked three years ago. Integrations that were stitched together in different moments for different reasons. A team that has learned to work around the system rather than with it, because the system has never quite fit.
The conventional wisdom is that this makes transition slower and more painful. And it can. But it also misses something: established businesses have the most to gain. They have the data, the process depth, and the relationships that — properly structured — become extraordinary leverage in an AI-native CRM. The history is not baggage. It is the raw material. The challenge is transforming it.
The history is not baggage. It is the raw material. The challenge is transforming it.
Why the foundation matters more than it ever has
There is a version of the AI-in-CRM story that is reassuring: the AI will figure it out. Dump your data in, ask it questions, and the intelligence will do the rest. I understand why that is appealing. I also think it is wrong, at least for now, and dangerously so if it delays the structural work.
Attio’s CEO Nicolas Sharp put it clearly in laying out their product vision: when an AI model interacts with a well-architected CRM, it is not just seeing a collection of data — it is interfacing with a structured representation of the business. It understands that a particular field represents a currency value, or that a relationship between two records has a specific meaning. That semantic richness is what allows the AI to reason accurately, not just retrieve.
The inverse is equally true. A sophisticated AI layer reasoning over a badly-structured workspace does not produce sophisticated outputs. It produces confident-sounding nonsense — reports that look right but are not, automations that fire on the wrong conditions, summaries that miss the nuance the data actually contained. The AI amplifies what is underneath it. Which means the quality of the foundation is not a legacy concern. It is the precondition for everything the new era promises.
What good architecture actually looks like: A CRM built for the AI era has a data model that mirrors how the business actually operates — not how a generic sales template assumes it does. Custom objects for the specific entities that matter. Relationship structures that capture real-world connections. Fields with semantic meaning, not just labels. Workflows that encode real process logic, not workarounds. This is the work that makes everything else possible.
Why I chose Attio for this moment
I started Sideways CRM to help ambitious businesses build sales technology that actually worked for them — the thesis being that most CRM implementations failed not because of the tool but because of the setup. That has turned out to be even more true than I anticipated, and for reasons I did not fully foresee when I started.
Attio was built from the beginning with a different data philosophy. Most CRM platforms treat data as strings — flexible enough to hold anything, which means they hold everything badly. Attio’s architecture assigns semantic meaning at the property level. That design decision, made before the current AI wave, turns out to be exactly the right foundation for an AI-native future. The flexibility that made it the right choice for complex, bespoke business models is the same quality that makes the AI layer work properly when it is switched on.
I also believe in the team and the trajectory. Attio is building deliberately — not bolting AI onto a legacy architecture but rethinking what CRM should be from first principles. That is a different kind of bet, and the early signals are that they are getting it right.
A word about what is changing for me, too
I would be dishonest if I did not acknowledge the obvious. Some of what I do today will be automated. The more routine configuration work, the maintenance tasks, the things that currently require a specialist to execute — the AI layer in tools like Attio is already beginning to absorb some of that. That will accelerate.
What will not be automated, at least not yet, is the judgement that precedes the work. Understanding what a business actually does, how its revenue motion operates, what the critical relationships and data points are that drive performance — and translating that into a structural architecture that will hold under the pressure of growth and change. That is a design problem, not a configuration problem. And design problems require a combination of experience, pattern recognition, and trust that software does not replace in the near term.
The transition moment is real for me too. I am more useful to clients at the strategic end of the conversation than I have ever been. The execution end is getting more automated. That is, honestly, a better version of the work.
The transition moment is real for me too. I am more useful to clients at the strategic end of the conversation than I have ever been.
What the destination looks like
For businesses that make the transition well, the destination is genuinely different — not an upgraded version of what they had before. A CRM that ingests data from every interaction without manual entry. That surfaces the right insight at the right moment in the sales process. That automates workflows that previously required a human at each decision point. That becomes, in effect, the operational backbone of the go-to-market function rather than just the place where data is stored.
That destination is closer than most people think. The technology is not five years away — it is being built now, and the gap between organisations that have their foundation right and those that do not will become visible within the next eighteen months to two years.
The question is not whether to make the transition. It is whether to make it on your terms, with intention, or to be pulled into it reactively when the gap becomes impossible to ignore.
Will Sawney is the founder of Sideways CRM, a specialist Attio consultancy. He works with revenue leaders to design and build high-performance CRM architectures. If this resonates, get in touch.
