Strategies for Resolving Companies and the Hierarchy: Key Takeaways from LMA Tech West & Southwest 2025

Key Takeaways from LMA Tech West & Southwest 2025

I just wrapped up a fantastic session at the LMA Tech West & Southwest Conference with two of my favorite people to talk data with: Carrie Remhof, Director of Firm Intelligence at Troutman Pepper, and Scott Winter, VP of Client Development at Index. We tackled a topic that's suddenly become mission-critical: company hierarchy resolution.

If you're investing in AI for client intelligence, this is your foundational question: Can your systems actually understand that those 47 different "Amazon" entities in your database are all related?

Because here's the thing—AI doesn't fix bad data. It amplifies it.

Why This Matters Right Now

Every firm is racing to implement AI for relationship mapping, predictive cross-selling, and market intelligence. But if your AI thinks JP Morgan, J.P. Morgan Bank NA, and Highbridge Capital Management are three unrelated entities, your insights will be fundamentally flawed.

As Carrie put it, reflecting on nearly two decades working with firms on this challenge: "Data is my life. I love it. My boyfriend doesn't like all the metrics I'm tracking, but..." She laughed, but she's serious. This problem has evolved from annoying to strategic.

Scott brought the vendor reality check: "Just in the last two weeks, I've seen a Magic Circle firm and a top AmLaw firm trying to combine their CRM with their experience system. It's a very difficult task."

The stakes? Every AI use case you want—relationship intelligence, predictive models, conflict checking, automated pitch generation—depends on understanding corporate hierarchies. Without that foundation, your AI investment delivers garbage.

Understanding Why This Is Hard

Before we dove into solutions, we walked through why company resolution breaks down.

The usual suspects:

Style variations: "JP Morgan" vs. "JPMorgan Chase & Co."
Brand confusion: Users search "Oreos" but need "Mondelez International"
Typos: "Goldman Saks" (yes, really—even at Fortune 100 companies)
Subsidiaries: Is Waymo its own company or part of Alphabet?
Outdated data: Scott pointed out that "Tesla was entered 10 years ago as just an automobile manufacturer. Now they're in solar, autonomy, energy storage—none of that's reflected in their industry codes."

Traditional matching fails on these edge cases. But AI and modern enrichment services have changed what's possible.

Three Strategic Approaches

We broke down three distinct ways to tackle company hierarchy resolution, each with specific use cases.

Approach 1: Conventional Resolution

Maintain every legal entity exactly as it appears—maximum precision.

When it works: You need forensic accuracy for conflicts and regulatory work. You have dedicated data stewardship resources. Legal entity precision is critical.

The reality: As I noted in our discussion, this approach is "very granular and hard to maintain. Matching people to entities becomes nearly impossible."

Scott illustrated the challenge with Amazon's organizational complexity: "When you pull in a contact from an email exchange, where should she go? Amazon has this massive leadership structure across multiple business units. Deciding exactly which entity someone belongs to—that's tough when you're extremely granular."

Best for: Small firms with limited client portfolios or specific practices requiring precise subsidiary tracking.

Approach 2: Group/Brand Resolution

Organize related entities under recognizable parent brands while maintaining subsidiary details in the background.

When it works: You're balancing precision with usability. Relationship intelligence is a priority. Attorney adoption matters. You're building for both AI and human users.

Scott explained the usability case: "When a lawyer is building their business plan and wants to target Amazon, they don't want to choose between AWS US East 1 or Amazon Fulfillment Services Inc. They just want Amazon."

I emphasized that this approach "really helps marketing and BD tell the story. When you bring information to your lawyers in a way that's useful for them, it gives you a seat at the table."

The trade-off: Your firmographics will be broader, and you won't always represent the exact legal entity. But for most firms, that's acceptable given the usability gains.

Best for: Large firms prioritizing relationship mapping and cross-selling intelligence.

Approach 3: Domain-Based Resolution

Use email domains and AI to automatically match contacts to companies—the fastest, most automated approach.

When it works: Speed matters. You have heavy ERM usage. Data stewardship resources are limited. You need to maximize contact data volume.

Scott walked through the mechanics: "When you set up domain associations, anyone emailing you from amazon.co.uk automatically rolls up to Amazon UK. It's pretty automated."

This approach shines for ERM systems capturing contacts from email signatures and metadata.

The trade-off: Less precision in corporate structure representation. You might assign contacts to the wrong subsidiary.

But as Scott noted: "This can directionally get you to where you need to go. When you're looking for a relationship, you follow the breadcrumbs rather than trying to organize every needle in the haystack."

Best for: Firms prioritizing automation and scale, particularly for mid-to-lower tier accounts.

The Breakthrough: The Tiered Strategy

Here's where our conversation got really practical. Rather than choosing one approach for everything, sophisticated firms use all three based on client importance.

"Not every company and contact is created equal. As data people, we get stuck thinking everything needs the same level of care. You can band your companies by importance—ranked companies get more attention, while the other 80,000 can have lighter rules."

The tiered approach we recommend:

Tier 1: Top 100 Clients → Conventional Resolution
Full subsidiary mapping, manual review, precise firmographics, detailed ownership tracking.
Tier 2: Target Accounts (101-500) → Group/Brand Resolution
Automated grouping under parent brands, quarterly refinement, balanced detail for cross-selling.
Tier 3: Remaining Database → Domain-Based Resolution
Fully automated contact matching, annual review for tier promotion, focus on coverage and scale.

I shared context from working with 20+ firms at my previous role: "When I ask people how many companies are in their CRM, I get answers like 50,000, 100,000... someone said 485,000. Crazy, right? So rationalizing those and putting them in buckets by importance is really critical."

Real-World Implementation: Troutman Pepper's Story

Carrie shared her firm's merger experience—a masterclass in pragmatic implementation under pressure.

"When we announced our merger last summer and became Troutman Pepper Locke on January 1st, I immediately thought, 'How are we going to match our clients?' We think we're all unique and boutique, right? But I found I had both Troutman customers and Locke customers—how do I get them to match?"

The pragmatic solution: "I am not going to fix my finance system at all. I'm going to accept what my finance system gives me. It's going to give me a lot of client numbers that start with the number six—brilliant strategy on our part, right?"

Working with Scott's team, they used AI-powered matching to identify overlaps and assign unique identifiers (D&B numbers) to create universal match keys.

The key mindset shift: "We realized we're going to have duplicates, and I've accepted duplicates. But how do I augment my data so I can find things together?"

Scott reinforced this philosophy: "When you write new information into your systems, you're not overwriting existing names. This is appended data. If you have 'Litigation, McDonald's' as a client name, that can remain because that's what attorneys are looking for. You add additional fields with the official name so you can group for reporting."

The result? Unified reporting across both legacy firms within six months, without spending years on manual cleanup.

Managing Partner Expectations: The Critical Success Factor

Perhaps the most valuable part of our discussion was about managing stakeholder expectations—something every data leader struggles with.

Carrie's approach: "One of the most important things is managing expectations with your firm. Because you'll choose a method and move forward, but there will always be a dozen partners who didn't read the email or didn't fully understand the decision. As long as you can consistently explain why you chose this approach, you're good. If we roll up to the brand entity, we state that clearly."

My reframing technique: "I've learned that when attorneys point out issues—'This isn't right! Where did you get this from?'—five years ago I'd panic. Now? I'm excited because it means they're actually using the system. I flip it back on them: 'Great! Tell me what I missed. Where can I get that information that's living in your head? Help me bring that in.'"

This turns criticism into collaboration. I also remind attorneys: "Some of our sources aren't as quick as you are with news. The client knows best. We need your input."

Setting realistic expectations: Carrie's key message to partners is brilliant: "This is directional information that's better than what we had before. In your legal work, you need to dot every I and cross every T. But for business development and marketing, directional accuracy is still incredibly valuable—even if it's not as perfect as your work product."

She acknowledges their world of precision while explaining the practical reality of business intelligence.

The Technology Ecosystem

We discussed the practical reality of working with third-party data providers.

Carrie was refreshingly blunt: "None of these vendors are perfect—just like I'm not perfect. So use them together. I've had D&B for eight years. Next year I'm adding LexisNexis company financials. They have different slices of the data, and I'm going to show both."

"In the past, I wanted one single source of truth. But there is no single source of truth for company data. So I'm putting them next to each other. By the time someone gets to my company record, they're smart enough to see what D&B says versus what LexisNexis says—and overall, it's helpful."

Scott echoed this: "Even S&P and D&B have duplicate companies with minor variations. There are a lot of hands in the cookie jar."

I emphasized the practical path: "Domain-based resolution with brand grouping is probably the most pragmatic approach for large firms. Using third-party data services or ERM to extract domains and enrich them—that's where these technology tools really shine."

The Data Lake Connection

Scott brought the conversation to the broader context: "The data lake—joining these systems together—is one of the biggest challenges firms are facing right now. Once you get entity resolution done, the silos can start breaking down. You can see the big picture and actually execute your data strategy."

He explained what becomes possible: "When you have the right organizational structure in place, there's so much third-party power you can leverage. You can connect to judges, expert witnesses, pull hierarchy data automatically—things you don't want cluttering your CRM but are incredibly valuable in your data lake for identifying white space."

This is where company resolution moves from "data cleanup project" to "strategic infrastructure for AI."

The Action Plan

We wrapped with practical next steps, emphasizing action over perfection.

Make a Decision and Start

Carrie put it simply: "So many firms just spin and overthink. Sometimes you have to make a plan and start. If you're going to build a house, you can't keep designing it for five years."

Scott agreed: "You have to be practical. Accept your finance data. Accept that you'll have 12 versions of the same entity because partners want their origination credits. Just accept it. Firms have spent three decades trying to fix it from the start—don't make that mistake."

Augment, Don't Fix

Carrie continued: "I'm going to have duplicates, but I'm going to enrich my data with third-party data. I'll add extra fields and metadata so I can report on it. Recognize there are multiple slices to how you capture client data. First comes finance, then you augment with brand, financial parent, ownership—but don't overload it."

Focus on What Matters

"As data people, we tend to be perfectionists. We want to fix everything perfectly," Carrie admitted. "But not everything needs the same level of care. Focus on the big things. I'm happy at 80% because I'm farther along than before. Create business processes to maintain things going forward."

Don't Let Perfect Be the Enemy of Good

This became our rallying cry.

"It's very tough in law firms because there's this demand for perfection. You have to set the stage. This is better than what we had before. This is directional information that's still incredibly useful, even if it's not litigation-grade precision."

Scott brought the reality check: "It's hard to live in law firm data people's shoes. But the way you speak with partners about setting expectations—that's how you make progress."

The Bottom Line

Company hierarchy resolution isn't a "someday" project. It's the foundation that determines whether your AI and analytics investments deliver value or frustration.

The competitive advantage is real. Firms with unified company intelligence can answer questions their competitors can't:

"Show me all our work for JP Morgan's financial services subsidiaries and identify white space"
"Who in our firm knows decision-makers at any Tesla entity?"
"Find all private equity-backed healthcare companies where we have existing relationships"

The firms leading in AI-powered client intelligence didn't wait for perfect data. They chose a pragmatic resolution strategy, started with their most important relationships, and began feeding better data to their AI tools immediately.

They're refining based on actual usage while their competitors are still planning.

Your Next Step

As I wrapped up our session: "Make a decision, get started, and don't be afraid to tap your network and ask questions. There are people who have done this—learn from them."

The data foundation for AI isn't optional anymore. The question isn't whether to invest in company hierarchy resolution—it's how quickly you can deploy it.

Reach out to me, Carrie, or Scott on LinkedIn if you want to continue the conversation. We're always happy to talk shop about making data actually work for law firms.