Why graph technology is a business strategy: Interview with Scott Taylor
To complement this year’s edition of the Graph Landscape, we interviewed seven leading voices from across the graph and data community — practitioners, strategists, and industry leaders — to explore where graph technology stands today and what the next phase of adoption will require.
One of those experts is Scott Taylor, The Data Whisperer.
Scott has spent decades helping organizations understand the strategic value of proper data management. Championing the strategic “why" over the technical “how”, he specializes in helping Data Leaders craft a business-accessible narrative that secures executive sponsorship, stakeholder support, and ongoing funding for data initiatives.
In this conversation with Matthieu Besozzi, Scott explains why graph technology is not a flashy trend but a foundational enabler, one that connects entities, reveals hidden relationships, and supports the analytics and AI initiatives that organizations increasingly depend on. He also shares why storytelling remains the biggest challenge in data management, and why learning to communicate the value of structure is a true leadership superpower.
Read the full interview transcript below and watch the complete video to hear Scott’s perspective in his own words.
Matthieu Besozzi
Okay, so maybe we can start with a quick introduction about yourself. What drew you into the data field?
Scott Taylor
I've been in the data space since pre-Y2K, always helping people understand the power and value of data governance and data management. I go under the moniker these days of The Data Whisperer—data whispering being a way to help calm data down and get it structured, organized, and ready for whatever else an organization wants to use it for.
I now focus on helping data leaders put together a business-accessible narrative to convey the value of what they're trying to do in the data space, to get executive support, stakeholder engagement, and funding. I call that data storytelling for data management: a way to put a story together that you can use to talk to executives or non-technical stakeholders that explains the value in very simple, business-like terminology.
Matthieu Besozzi
Very clear. I believe since you started, you've seen a lot of evolutions and new paradigms. Can you talk about that a little bit?
Scott Taylor
I think I've been through a number of cycles. I would say AI these days is probably the biggest cycle I've ever seen. But starting in the nineties and going through the 2000s, the 2010s, and now the 2020s, there's always something new coming along in the data space.
What I've found, though, is that no matter what it is, it still needs structured, linked, connected core data to make it work. Every company I've ever dealt with is trying to bring value to their relationships through their brands at scale, that’s the way I like to articulate it. And if you want to do something at scale in an enterprise, you need technology. If you have technology, you need hardware, software, and data. If you have data, you need data management.
From GenAI all the way back, if we go back before computers, before electricity, to the concept of a general ledger, you still needed a chart of accounts. So from GenAI to general ledger, you still need data management no matter what. You still need that structure.
Matthieu Besozzi
Yeah, sure. You've seen many cycles, but there's a continuity, which is: How do we structure the unstructured data?
Scott Taylor
Yes, there's always value in unstructured data. But the way to get value out of data, at least from my simplistic viewpoint, is that you've got to put structure on it.
I remember when big data came out and everyone said, “Big data is so fantastic.” What makes big data big? Its lack of structure. What makes big data valuable? Putting some form of structure on it.
Now we’re at a point where big data isn’t as hot anymore, it’s just what everybody is dealing with. But there are still tons of unstructured data that organizations have. Some people say up to 80% of enterprise data is unstructured. The only way to get value out of that is to determine what entities are in that unstructured data. What is this data about?
That’s the place I always like to focus on: What it's about. What are the entities? What are the products, relationships, brands, and whatever core terminology an enterprise is using to drive its business? How do we find that?
And to move into the graph side a little bit: How do we draw relationships between those entities, relationships that people either don’t know exist or have too much trouble tracking themselves?
For example, here’s a customer. What’s their hierarchy? Who are they connected to? Who else do they deal with? What other parties are they engaging with? That, I think, is critically important.
Matthieu Besozzi
I hear you talking about entities and relationships. So let’s talk about graph. How do you see the value of graph technology in this context?
Scott Taylor
If you can’t connect these things and unlock the relationships between them, then you're really missing out on a lot of the value of the data you have.
There are some standard things that a lot of data governance folks do, like hierarchies, taxonomies, and classifications, which are fairly straightforward. But the real nuance and exciting value comes when you can find links and connections that would elude a human.
Graph technology can reveal that this is connected to that in a way you might be able to take advantage of, or that this is connected to something else, which is connected to something else, which suddenly reveals a risk you wouldn’t have been able to see before.
So once you’ve established those core entities, you want to start connecting and graphing them together to unlock a whole new set of insights.
Matthieu Besozzi
Yeah, I totally agree with you. Something we’ve noticed—I’ve been in the graph technology field for more than seven years—is that all these market analyses were predicting a huge boom in graph technology ten years ago.
It took some time to actually take off. I have the feeling it’s starting now, maybe thanks to LLMs and GenAI, but I’d like to hear your perspective. How do you explain this delay in traction for graph databases, despite a very clear value proposition? And do you see this technology getting more traction today?
Scott Taylor
I certainly hope it gets more traction.
When we think about graph, semantic layers, ontologies, all of those things, they’re getting more attention lately. But you and I might think graph technology is sexy. I’m not sure how sexy a businessperson who doesn’t understand any of this would think it is.
The infrastructure, foundational, back-room stuff never gets the spotlight it deserves, while the cool, hot, flashy applications get a disproportionate amount of attention and funding, even though they depend heavily on the structuring elements underneath.
Graph is an enabler. And enablers aren’t always the stars of the show. It’s a means to an end.
Part of the storytelling I share with data leaders when they’re trying to sell this internally is showing that whatever the end result is, providing value to relationships through your brands at scale, cannot happen without the foundational structural pieces in place first.
It’s not chicken and egg, it’s egg and omelet. If you don’t have the data management portion under control, you’re not going to develop analytics and business intelligence successfully. It’s just not going to happen.
The business sees the analytics and the outputs. They don’t see what goes into creating them. It’s like going to a fabulous restaurant and enjoying a spectacular meal. You don’t necessarily see what’s happening in the kitchen, the sous chef and everyone preparing it. And you may not want to know. But you don’t get that great meal without the techniques and preparation happening in the back.
Matthieu Besozzi
I agree with you. What resonates with me is that we need to double down on graph visualization. And second, structuring unstructured data is still an unsolved problem.
Scott Taylor
Sure.
A lot of the graph visualizations I’ve seen make you think, “Okay, I see something, but I don’t know what I’m looking at yet.” How do you simplify that to show clearly that this is related to this, which is related to this, and that this gives you value?
The simpler you can visualize that, especially for non-technical executives, the better. You want them to go from “I have no idea what this is” to “Wow, how did we live without it?”
Visuals help. Simple explanations help. And you have to remind them there’s cause and effect. If you don’t have this, you won’t get that. It’s as simple as that.
At the enterprise level, outputs must be structured and aligned to the operational realities of the company.
Matthieu Besozzi
What do you believe is the main challenge that still needs to be solved in data management?
Scott Taylor
Telling the story.
People come to me all the time, data science students and enterprise CDOs alike, and ask what they should focus on. For me, it’s making sure you can tell a story.
Everyone can tell a story. You tell your parents what you did over the weekend. You talk to your spouse, your kids, your friends. Humans naturally communicate in stories.
But in the data space, being great at data often means being strong in hard skills. Communication is considered a soft skill. Yet if you want to be a leader and drive initiatives, you must be able to communicate.
And here’s the challenge: there are often better storytellers in an organization than those in the data department. Marketing knows how to tell stories. Sales must tell stories or they don’t hit quota. Executives know how to articulate value.
So learning storytelling techniques is critical to succeeding in the data space. And data management needs its own special kind of storytelling, one that clearly explains why managing data is critically important to an enterprise.
Matthieu Besozzi
So this is the one piece of advice you give to leaders: tell a story?
Scott Taylor
Learn how to tell a story. Get better at it. It’s a superpower, it absolutely is.
And focus on why it’s important rather than how to do it. A lot of technical folks love to share the 57 steps they took to get to an answer. Executives don’t care about that. The CFO doesn’t care about every pivot and adjustment.
Get to the point. Get to the why.
I’ve never met a CEO, CFO, or non-technical business leader who cares about how you're going to get the data work done until they understand why it matters to the business.
Matthieu Besozzi
What technology or product has excited you recently?
Scott Taylor
I’ve been watching the AI developments, of course, that’s exciting. But for me, it’s actually the elevation of foundational capabilities: graphs, semantics, ontologies, hierarchies, master data, reference data, metadata, MDM, RDM, PIM, RIM, DAM, all those structural elements.
They’re finally getting more spotlight because of their critical importance in AI. When AI goes bad, it can do serious damage. This isn’t just a report where two numbers don’t tie out. Poorly controlled AI outputs can create real risk.
I think people are realizing that we need the structural foundation in place first. Hopefully, that understanding continues to grow.
Matthieu Besozzi
All right. Thank you, Scott.
A spotlight on graph technology directly in your inbox.
