Blog | Field Enablement | Strategic Business Systems Consulting

How to Make Your Customer Data a Strategic Asset

Written by Gareth Williams | Jul 31, 2025 6:55:29 PM

Practical steps to transform fragmented customer records into a powerful business resource

Do you have Customers? Clients? Patients? Members? Volunteers? Athletes?

Whatever you call them, your customers are arguably the most important thing for your organization and your records about them are critical to the customer experience and operation of your enterprise.

But most organizations have messy customer data. Different systems handle different processes and have their own list of customers. Different parts of the organization interact differently with customers and so need slightly different pieces of information, or see the customer differently. Maybe different locations have different systems, through rapid expansion or acquisition activity.

Whatever the reason, over time most organizations find their customer data fragments far beyond the ideal of having one, clean customer list that everyone uses.

 

Why do we care?

This is a problem!

  • It makes it harder or impossible to get an overview of the interactions (think sales, support, marketing touches) you have with your customers. Without that it is hard to make data-led decisions about where to apply focus and resources. This reduces effectiveness. Example: if a customer has a problem with the current product you should know that before opening a conversation about a new product
  • It introduces internal friction which reduces efficiency You have people in marketing, sales, finance and operations performing VLOOKUPs to stich the data together to produce some sort of answer to questions that span multiple processes and systems
  • You don't really know the customer, so it is harder to deliver the kind of personalized Customer Experience that will make you more effective. My favourite example here is activating events with special experiences (could just be an exec meeting) for certain customers.

How to fix it

We should treat customer data as a strategic asset for decision-making, customer experience and compliance.

Don't think of the data as something you do a one-time clean up on. Yes, you will have to clean it up and connect it, but in most cases there is deeper thinking required. It's not a chore... you are creating an asset.

Here is what to do:

1) Define a single system as the authoritative list for the enterprise

This is your Customer Master Data. You can use one of your existing systems, such as ERP or CRM for this if your needs are not too complicated or, if you are a large enterprise, you might need a dedicated Master Data Management (MDM) solution. Try not to introduce a different system if your existing ERP or CRM can handle it, especially as the vendor may have matured the MDM elements of those systems recently.

Don't make the mistake I did the first time around and believe the MDM will organize your records for you through clever logic, AI and external sources of reference like D&B Hoovers. Those elements help with acceleration and metrics, but they probably don't tell you enough about how you want to model your most important customers.

That said, a good starting place is to do exactly that. Get your customer data in one place so you can see it and start to manage it.

Following that you can graduate to a true enterprise approach where other systems should receive the authoritative records from the MDM and can update the MDM only through a verification and governance process. You can prevent the governance process from being a bottleneck by defining automated business rules that implement your policies and customer data model.

"Policies and customer data model?", I hear you ask...

2) Implement and Automate Data Governance

Define who is responsible (stewards) for which parts of the customer data  and what good looks like in terms of quality standards. (If you can, give people who are actually called Stuart or Stewart this role to increase enterprise levity).

These days most systems have great API options, so you can automate data flows. But what is power without control?

I have seen the case where the ease of automation has pumped thousands of low-quality duplicates into a CRM system. We need to build quality in.

This comes from:

  • setting up broad enterprise participation from the stewards, with good decision-making powers
  • defining, via rules, what good data looks like
  • embedding those rules in the automation to prevent pollution
  • ongoing monitoring to enable out-of-bounds management


Lower the barriers for good quality and raise the barriers for bad.

For example, for users who only care about performing their task quickly it might be easier to enter a new customer record than find and reuse the right one. This is especially true if the name cannot easily be matched.

In one case we implemented a "Search before create" process in all contributing systems to measurably increase data quality through a) reuse and b) correct setup of new records when needed. We also adjusted the searchable attributes in the CRM to make it easier to find the right record.

The governance process is also a good place to implement elements of your compliance with data privacy regulations such as GDPR and CCPA.

3) Standardize Customer Data Definitions

In large enterprise this is the hardest part. This section discusses three key challenges:

  • How to model large customers in the simplest way
  • What to consider as enterprise-level data and what to just leave in a siloed process/system
  • How data is a symptom of other things you might need to tackle

In B2C or SME it can be more straightforward to define the hierarchy for the customer records. If you don't have customers with a matrixed organization, or don't have a matrixed organization yourself you can minimize this part.

In large B2B enterprise it is harder as large B2Bs tend to have large organizations in the same way their customers do.

Large organizations have lots of legal entities, often driven by their geographic and financial structures. Although external reference lists of these Legal Entities can be bought from vendors like D&B Hoovers they don't always correctly model how your teams are actually interacting with your customers.

Let's take an example. Pearson is "the world's leading education provider". It is a large organization that operates multiple lines of business across the globe. Please note that the following is just an example.

If you are working with Pearson at a similar scale you cannot simply have one record for "Pearson" in your information systems, as there will be multiple interactions.

You need to model a hierarchy with parent/child relationships. This allows each business function to perform its function with the right part of Pearson at the right level of granularity, but also get an overview of all the business at Pearson. Our hierarchy might look something like this:

Example: Pearson Customer Data Hierarchy

Pearson Group (Parent Company)

  • The ultimate parent entity, encompassing all business operations and global subsidiaries

Subsidiaries

    • Examples:

      • Pearson Education Ltd. (UK)

      • Pearson Education Inc. (US)

      • NCS Pearson, Inc. (US)

    • Each business unit operates in distinct regions or markets and may have its own customer sets, regulatory requirements, and business models

AND

Pearson operate five distinct product divisions

    • Assessment & Qualifications
    • Virtual Learning
    • English Language Learning
    • Enterprise Learning and Skills
    • Higher Education

Which may or may not be materially present in each of the legal entities.

This sort of product/region/customer matrix is typical in large enterprise.

If modelled in your customer master data with every distinct combination (for accurate reporting) then you will not be able to see the forest for the trees. The complexity will be unusable.

I have found that the best thing to do is create nodes for the most active combinations and establish a principle of grouping the business together at those nodes in the transactional systems. It is a bit of an approximation, but tends to be more practical and manageable. In this way the customer master data hierarchy matches your actual customer engagement, either as-is, or to-be, as you decide. An example might look like this.

 

This is an example of why you can't leave everything to the logic in the MDM system. The MDM tends to become a dumping ground of all possible records and then becomes the opposite of the clean, singular list of customers you are trying to create as strategic asset for your organization.

AI and analytics will help you define the common nodes, but you have to have business judgement involved. This needs to come from the business strategy and the customer account team.

This level of curated modelling of customer master records is only really worthwhile for your larger or, perhaps I should say, most strategic customers. Smaller, or less strategically-important customers simply won't have the complexity and it is less important if the model is a bit wrong.

The approach I have taken in large enterprise is to curate the MDM model of the strategic customers as described above, then use rules-based, automated curation of the remainder.

When I and the team I was working with did this it led to a night-and-day improvement over letting the MDM work on its own.

What belongs in the Enterprise Master and what doesn't

Again, this differs based on the size of your organization.

If you are an SME you are probably using your ERP or CRM. Or maybe your till system! This is fine. Just put what you need in your customer list... which is probably everything.

In large enterprise you will find that different transactions need different levels of detail

  • Sales needs a high-level view of the customer to keep the pipelines straight
  • Logistics needs the ship-to and sold-to addresses
  • Partner Marketing might operate at a country level
  • Different products might ship from different acquisitions 

If you Master everything then the MDM can become too complicated: a deep hierarchy that can become a bottleneck when governance is applied. Consider here a more federated model, where true enterprise master data is mastered and governed and managed in the MDM, and some detail-level data is only kept in the transactional systems that need it.

In my example above that might mean that the Sales view is considered enterprise-shared master-data, but the logistics details are only in the ERP. Clearly you have to have a common key value to link the two and the MDM approach will provide that.

 

Fragmented GTM approach and processes

Often messy customer data is not simply a sign of lack of governance, of no-one tidying recently. Sometimes it is a smoking gun for unaligned business processes that may be undermining your customer experience.

For example, different departments may have created different "Customer Types" within the CRM to allow a speedy instantiation of a business process. However, sometimes, the customer is truly the same and we are just doing something different with them. The trouble with modelling as different "Customer Type" is that we haven't looked at things from the outside-in, from a CX point of view. We have modelled our internal business processes.

It is strategically better to do the hard work to map the ideal customer journey across multiple internal processes and have a more integrated experience. That, in turn, will show up in more integrated customer records. Going siloed is fast... and you might need to go fast right now. Just realise that you might be building up debt in you enterprise agility and customer experience.

4. Invest in managing customer data as an asset

Now you have created a customer master list - a singular(ish) list of your customers that is clean and everyone uses then you can start to reap the benefits. Watch as automations and comparisons become easier. Watch as dashboards become more useful. See how you can innovate new customer experiences or offers.

It makes sense to keep this asset up to date:

  • Set up ongoing data cleansing, deduplication and validation. This sort of task is ripe for AI-acceleration
  • Enrich the customer data from additional internal and external sources so that it becomes event more useful
  • Make sure your organization knows how to use it, in the transactional systems and the analytics
  • Let your AI see the data (with guardrails) so holistic information is in the context

 

In summary

Most enterprises have messy customer data, and over time and with growth it gets worse

You can reduce internal friction, drive better decisions and offer more granular customer experiences if you invest in organizing the customer data across your organization

Start by consolidating, then organize and govern.

Drive actionable insights and operational improvements from the data such as:

  • Compliance with international export controls
  • Identifying upsell or cross-sell opportunities by understanding the full history and relationship depth with a customer
  • Comparing future pipeline and current shipments at whole-customer level to enable resource allocation and product planning

Don't take your eye off the ball - this is not a one-time clean-up. AI can make the maintenance less resource-intensive and can scale your effort.


Please get in touch for a no-obligation conversation on getting started.

Gareth Williams | LinkedIn

or use the Contact form here on fieldenablement.com

 

Gareth and I collaborated on a large-scale data transformation program touching hundreds of thousands of records, hundreds of business applications, and dozens of stakeholders and business groups. Gareth helped organize and mobilize teams required to design and install operational processes and metrics, as well as execute data cleanup and alignment across multiple business applications. Gareth was a champion of data quality and governance, data simplification, and automation. Gareth's common sense approach and disciplined program management, as well as his ability to influence stakeholders and leadership, were essential to the success of the program.

- Director, Data Management, Large Enterprise Client