"The CRM is not being used properly."

It is one of the most common complaints in sales, marketing and customer experience work. The system exists. The licences are paid for. The dashboards are available. Yet nobody is quite sure whether the CRM shows what is really happening.

A lead might be marked as contacted, but nobody knows whether the customer got the answer they needed. A task might be complete, but no next step has been agreed. A note might exist, but the useful context sits in someone's inbox, Teams chat, call recording, project spreadsheet or memory.

AI does not make this problem disappear. It makes it more important.

The better question is not "will AI replace CRM?" The better question is: what commercial context does the business need, where does that context live, and how can people and AI use it safely?

AI changes the role of CRM by making reliable commercial context more valuable. The CRM may remain the record, but the useful context for follow-up, ownership and customer progress often lives across data, communication and operational systems.

What job should CRM actually do?

The point of CRM was never software compliance. It was supposed to help the business know who is interested, what they need, what has happened, who owns the next step and what should happen next.

That is why so many CRM adoption projects disappoint. They treat the tool as the answer before defining the commercial system underneath it. A team can be trained to log activity and still fail to move opportunities forward. A dashboard can look tidy and still hide slow response, weak follow-up or unclear responsibility.

Break.Beat calls this Value Leakage: value disappearing between customer interest and commercial outcome. Lead Leakage is the most visible form. A good enquiry exists, but the response, follow-up, ownership or visibility is not strong enough to turn interest into progress.

The CRM may be part of that system. It is rarely the whole system.

Where does useful commercial context live?

In project and account work for a seven-figure account for a household-name international technology business, the live picture does not sit neatly in one platform. It moves across delivery plans, client requests, email threads, calendars, Slack or Teams conversations, meeting notes, shared drives, status trackers and individual judgement.

That is not unusual. In large environments, the truth of the work is often distributed. The project plan shows one version. The meeting notes explain another. The client email changes the priority. The delivery tracker shows the status. A spreadsheet provides the working view. The account lead holds the commercial context.

The same pattern exists in customer journeys. A CRM record may tell you that a valuation enquiry arrived. It may not tell you that the vendor mentioned a probate issue on a call, asked about fees in an email, received a generic response, went quiet after a handoff, and then reappeared through another channel.

For property businesses, that context matters. For PropTech businesses, it matters during onboarding, renewal, support and adoption. For service-led businesses, it matters because customers experience one conversation even when the business manages that conversation across several systems.

How does AI change the CRM question?

AI makes it possible to research a prospect, personalise outreach, time a message, record a response, recommend follow-up, review a calendar, check inbox context, summarise meeting notes, scan Slack or Teams updates and produce a daily working view of live priorities.

Some of that can happen without using a traditional CRM as the main interface. A spreadsheet, inbox integration or custom workflow can sometimes provide enough structure for a specific job.

That does not mean the business no longer needs a commercial system. It means the commercial system may no longer be identical to the CRM screen.

Microsoft's 2026 Dynamics 365 article points towards CRM moving into the flow of work, where customer data, productivity data and business logic operate together with governance, security and observability. That does not make CRM irrelevant. It changes the job CRM is being asked to do.

Why is the data layer becoming commercial infrastructure?

At Andrews Property Group, one of the important shifts was not simply campaign activity or CRM usage. It was the move towards customer visibility: conceiving and implementing a data function/team and initiating Single Customer View / data warehouse thinking across Sales, Lettings and Financial Services.

AI can only work usefully with the context it can access. If that context is scattered, duplicated, stale, restricted in the wrong places or exposed too widely, the business has a problem before the automation starts.

That does not mean every business needs an enterprise data warehouse before it can do anything useful. It means the business should think like an operator before it behaves like a tool buyer.

Useful AI workflows need several things underneath them: clean source data, agreed definitions, permissioned access, audit trails, retention rules, security controls, clear ownership and human review points.

What happens when AI uses bad commercial data?

AI can summarise the wrong thing beautifully. It can prioritise based on incomplete evidence. It can personalise from stale context. It can recommend follow-up that ignores a sensitive conversation. It can turn poor CRM hygiene into polished misinformation.

That is why the current rush towards AI automation should make leaders more interested in data quality, not less. Gartner reported in May 2026 that marketing leaders expect AI-driven automation of marketing work to rise from 16% in 2026 to 36% by 2028. The direction is clear. More automation increases the cost of weak context.

Salesforce's 2026 UK marketing research makes the same operational point from another angle: customers increasingly expect two-way conversations, while many marketers struggle to respond promptly because they cannot access the context they need.

That is not only a marketing problem. It is a commercial systems problem.

Is the future CRM versus no CRM?

No. That is too neat, which usually means it is about to become expensive.

In some businesses, the CRM should remain the main operating interface. In others, it may become the system of record but not the daily place where every action happens. In smaller teams, a practical combination of structured data, email, spreadsheets and AI-assisted workflows may be enough for a specific process. In larger organisations, stronger data infrastructure and governance may be essential before AI can be trusted with meaningful work.

The question is not whether CRM survives. It is what job the CRM should do.

Should it be the operating interface? The customer record? The reporting layer? The workflow engine? One input into an AI context layer?

Those are design choices. They should follow the work, the risk, the customer journey and the decisions leaders need to make.

What should leaders check before adding AI to CRM?

Before adding AI, replacing CRM or building custom workflows, review the commercial system underneath the tools.

  • Where does useful customer or commercial context currently live?
  • Which information is reliable enough to automate from?
  • Which information is important but trapped in inboxes, documents or people's heads?
  • Which fields, definitions and stages are inconsistent?
  • Who owns the next step when AI identifies a priority?
  • Which data should be pooled or transformed before it is exposed to automation?
  • Who is allowed to see what?
  • What should AI be allowed to suggest, draft, update or send?
  • Where is human approval required?
  • How will outputs be logged, reviewed and corrected?
  • Which management rhythm will turn the insight into action?

The answers matter more than the software category.

The Break.Beat view

AI does not remove the need for a commercial system. It raises the standard for one.

The CRM may become less visible. The need for reliable commercial context becomes more important.

The businesses that benefit will not simply be the ones with the most automation. They will be the ones that know what data matters, who owns the next step, which judgement should remain human and how progress will be reviewed.

Most organisations do not lose good opportunities because nobody cares. They lose them because the next step is unclear, the context is missing, the system is not trusted or nobody can see where the opportunity has gone.

Before spending more money on CRM, AI tools or automation, look closely at what happens after someone gets in touch.

That is where the value often leaks.

Sources and further reading

FAQs

Will AI replace CRM?

Not in a simple way. AI is more likely to change the role of CRM. The CRM may remain the customer record, while useful commercial context increasingly comes from communication, project, customer and operational data around it.

Do small businesses still need a CRM?

Sometimes. A CRM can still be useful where there is meaningful enquiry volume, several people involved, complex follow-up or management reporting. The business should define the journey, data requirements and ownership rules before buying or replacing software.

What is commercial data?

Commercial data is the information a business needs to turn customer interest into a useful business result. It can include enquiry source, stage, contact history, next action, ownership, value, customer need, project status, service history and management notes.

Why does data infrastructure matter for AI?

AI tools perform better when they have reliable context and clear rules. If the source data is incomplete, duplicated, stale or wrongly permissioned, AI may automate confusion rather than improve decisions.

How does GDPR affect AI workflows?

If AI workflows process personal data, organisations need to consider UK GDPR requirements including lawful basis, transparency, data minimisation, security, individual rights and safeguards around automated decision-making. The ICO's AI and data protection guidance is a useful starting point.