Professional services firms commonly use client data for segmentation, campaigns, pipeline reporting and renewal activity. The same systems can also reveal where service is difficult: a failed onboarding step, repeated support requests, an underused capability or a handoff that causes clients to explain their situation again.
That does not make every available signal appropriate to combine or act on. Client experience data can include personal information, confidential material, inferred interests and commercially sensitive service records. A useful experience must also be lawful, fair, secure and proportionate.
The opportunity is therefore more disciplined than “join up all the data”. Start with a specific client need, identify the minimum evidence required and design the service and governance together.
From targeting people to improving a journey
A marketing question asks who is likely to respond to a message. An experience question asks where the client is trying to accomplish something and what would make that interaction clearer, safer or easier.
The distinction changes what the firm measures. A portal login can be a campaign engagement signal. It can also reveal that clients repeatedly fail to retrieve a report. A CRM note can support sales planning. It can also prevent a client from having to repeat an accessibility preference to each new team.
Neither purpose is automatically valid because the data exists. The firm should define the use, applicable lawful basis, reasonable expectations, access, retention, security and any required communication to the individual. Legal, privacy, compliance and information-security specialists should advise where personal or confidential information is involved.
The ICO’s current purpose limitation guidance explains the need to specify and document purposes, tell people why information is collected and assess compatibility when reusing it. Its data minimisation guidance requires personal data to be adequate, relevant and limited to what is necessary for the purpose. Check current guidance and obtain advice for the particular processing.
Four practical applications
1. Relevant information based on an expressed need
A client may ask for updates on a sector, regulation or service issue. Recording that preference can help the firm provide useful information without relying on a broad newsletter.
Distinguish an expressed preference from an inferred interest. A person visiting a page several times may be researching for many reasons. Calling them with a sales pitch could feel intrusive and may trigger legal or professional issues. Aggregate content patterns can guide the editorial programme without identifying individuals.
Where individual-level use is appropriate, define:
- what signal will be used;
- why the client would reasonably expect the response;
- the role of consent or another lawful basis as advised;
- frequency and channel;
- how the preference can be corrected or withdrawn; and
- the outcome that indicates the information was useful.
Measure usefulness through client feedback, continued preference and relevant behaviour. Opening an email does not prove that it improved the relationship.
2. Proactive service when circumstances change
A drop in service use can prompt a human to check whether help is needed. It should not be labelled “churn risk” and treated as proof that the relationship is in danger.
The source article described a relationship partner who noticed that a client had stopped using a portal and learned that the client’s finance director had left. The exact causal outcome cannot be verified here. The observation is still useful: a change in behaviour can be a prompt for appropriate curiosity, not an automated conclusion.
Design the prompt so the relationship owner sees context and uncertainty. Set frequency limits and suppress signals where outreach would be inappropriate. Record whether the contact helped, was irrelevant or revealed a service problem. Use that evidence to improve the rule.
For sensitive services, automated profiling or decisions may carry additional obligations. Obtain specialist advice before implementation.
3. Fewer repeated requests across the service
Clients become frustrated when they provide the same company detail, matter history or preference to several parts of one firm. The cause may be fragmented systems, permissions, data quality, operating boundaries or a deliberate confidentiality control.
Map the repeated request before joining records. Establish which team needs the information, for what purpose and whether sharing is appropriate. Give people access only where their role requires it. Preserve information barriers and professional duties.
Sometimes the better fix is a coordinated handoff or a client-controlled record rather than a single firm-wide database. The target outcome is less avoidable repetition with equal or stronger control.
Measure repeated requests, correction, handoff time, errors and client effort. A reduced number of fields may improve both experience and data minimisation.
4. Friction detection before a formal complaint
Support tickets, response time, task failure, abandonment and feedback themes can show where a journey repeatedly creates difficulty. Start with aggregated or de-identified patterns where individual identification is unnecessary.
Trace a pattern to the process. A cluster of portal calls may result from unclear instructions, failed identity checks, inaccessible design or an upstream data delay. The first interpretation is a hypothesis.
Assign an owner and test a change. A better handoff meeting may solve the issue; a new technology platform may be unnecessary. Monitor unintended effects. Fewer tickets could mean the journey improved, or that people gave up seeking help.
Trust requires more than good intent
“Use data in the client’s interest” is a helpful ethical question and an incomplete governance test. Firms can sincerely believe an intervention is helpful while clients find it intrusive, inaccurate or manipulative.
Test the use through five lenses:
- Purpose: What client or service problem is being addressed?
- Expectation: Would the person reasonably understand this use from the information and relationship?
- Proportionality: Is less or less-granular data sufficient?
- Agency: Can the client express, correct or change relevant preferences where appropriate?
- Consequence: What harm could follow from error, inference, disclosure or over-contact?
Consent is not a universal answer. The correct lawful basis and any marketing, profiling or sector-specific rules depend on context. Transparency also needs plain explanation, not a broad privacy notice that technically mentions every possible use.
Good experience can result from collecting less. Deleting obsolete preferences, correcting inaccurate relationship data and limiting access may improve relevance and confidence more than a new predictive model.
Work with the systems already present
An organisation can begin without buying a customer-data platform. Inventory the CRM, service, portal, analytics and feedback data already held. For each source, record owner, purpose, quality, access, retention, legal basis, security and the decisions it currently supports.
Choose one journey with a visible problem. Examples include:
- briefing a relationship lead from approved, relevant records before a meeting;
- showing a client which part of an onboarding request remains incomplete;
- identifying a recurring service failure from aggregated tickets;
- helping a user discover a capability included in the service they already receive; or
- carrying a stated communication or accessibility preference through a valid handoff.
Establish a baseline and a review. Include the client-facing outcome, operational effect, complaints or objections, data quality and control performance. Stop or revise the use if the assumptions fail.
Do not infer an individual’s situation from page views and present the conclusion as knowledge. Human review is valuable only when the reviewer has context, authority and a genuine ability to disagree with the system.
Give the question an owner
Marketing may own the CRM, operations the service data and partners the relationships. The cross-firm question can therefore remain unowned: are we using data to make the service better for clients?
Create a small governance group with service, relationship, data, legal or privacy, security and technology perspectives. Its purpose is not to approve every report. It should prioritise use cases, prevent incompatible reuse, assign owners and review outcomes and harm.
Our client experience data audit maps existing data against relevant information, proactive service, reduced repetition and friction detection. It is designed to identify a bounded first use case and the governance questions that must be resolved before it begins.
The new role of data is not greater surveillance of the client journey. It is better evidence for improving a service, collected and used within clear limits. The quality of those limits is part of the client experience too.
