people passing boxes in a row

Data Talks, Episode 14: Improving Match Rates With Multipass Matching

Increasing match results and opportunities

Host: George L'Heureux, Principal Consultant, Data Strategy
Guest: Janine Moreira, Data Strategy Consultant

We like to think about identity resolution, AKA matching, as akin to the foundation of a house – it’s what keeps the house standing. Matching is the foundation of your data connected to Dun & Bradstreet data. Through matching, all of your disparate identifiers and fields of a contact that exists on multiple platforms can be mapped to a single record. Multipass matching takes this a step further by utilizing different input permutations of client data to discover even more match candidates, ultimately resulting in increased match results and more identified opportunities.

If you have a lot of potential contact information and don’t know which is right, you don’t have to keep guessing. Using a multipass match with Dun & Bradstreet gives you an even greater chance of finding and matching to the correct business entity.

 

Read full transcript

George L’Heureux:
Hello, everyone. This is Data Talks, presented by Dun & Bradstreet. I'm your host, George L'Heureux. I'm a principal consultant for data strategy here in the advisory services team at Dun & Bradstreet. In advisory services, our team is dedicated to helping our clients maximize the value of their relationship with Dun & Bradstreet through expert advice and consultation. On Data Talks, I get to chat every episode with one of the expert advisors at Dun & Bradstreet about a topic that can help the consumers of our data and services to get more value. Today's guest expert is Janine Moreira, a data strategy consultant at Dun & Bradstreet. Janine, tell me how long you've been with the company.

Janine Moreira:
I've been with Dun & Bradstreet for almost 16 years now. I've had various roles within the organization. I started in our fulfillment organization and then I worked in different areas, such as global data content, and most recently, our data advisory services.

George L’Heureux:
Can you tell me a little bit about your role as a data strategy consultant inside data advisory services?

Janine Moreira:
At D&B, our mission here is to really help our clients improve their business performance through data and insights. As a data strategy consultant, what we're really here to do is be subject matter experts for our customers around the different use cases of their data, whether it be from D&B identity resolution, master data initiatives, hierarchy assignment, or even various global data points. We're really here to help our clients make sure they understand the data, are maximizing the value of the data in the most effective and efficient ways possible.

George L’Heureux:
One of the things that you mentioned in that list of items that we consult with our customers about was identity resolution, what we call "matching" colloquially. One of the things that we talk about with matching and that you're frequently discussing with our customers is this idea of a multipass matching strategy. That's what we wanted to discuss today. Can you tell me a little bit about what is multipass matching?

Janine Moreira:
Just to take a step back, when I think about D&B matching or D&B identity resolution process, I really think of it like the foundation of a house. It's sometimes overlooked, but it's the most important thing, right? Everyone looks at the exterior of a house. They look at the shingles, they look at the windows and the front porch, and they're like, "Oh, how beautiful that house is." The foundation is never really realized. The foundation of the house is what keeps that house standing, and so when I think about D&B identity resolution, I think of it as the foundation of your data connected to our data. Really, that's the purpose of identity resolution. When we talk about multipassing, what we're really talking about is the ability to take different input permutations that clients may have of their data and take that through our identity resolution process to ideally return not just more matches, but the best match candidates, so more performance, higher-performing matching.

George L’Heureux:
We're really talking about trying to, and this is important to me because I'm in the process of building a house right now, so that whole idea of foundation is important. This is really bolstering the strength of that foundation of matching. When we talk about doing these different permutations, what are the extra data points or permutations that customers might have that we can help them through multipass matching with?

Janine Moreira:
When we think of matching, I think sometimes we're stuck in this common thought of, "Okay, we want name, we want address, we want city, state, and phone and country. Those are the critical elements that we need to match." But that's just one set of matching input permutations that a customer may have. When we talk about D&B matching, we talk about the fact that we have over 420 million records in our global data cloud, which is a massive breadth of data.

But the other important note is that we have over billions of prospective data points to match against because when we talk about the D&B D-U-N-S Number and the value that it brings and the fingerprint of a business that it is is that we may have additional points of information on that record, so we may have not only business name, we'll have multiple trade styles' names available maybe. We could have the primary operating address, as well as the registered address. We could have former address or former phone information. We may have CEO or former CEO information. All of these are data points that we have when we take your input record and we take the data that you're providing to us to try and have the best outcome of match, so we match against all these additional data points.

When we talk about multipassing, it's a very similar concept. You may also have multiple data points, maybe not realized in the existing data set that you're looking at today, but maybe you have a supplier file and you have a company name and you have the remittance address and phone number. But if you work with your legal or your finance, you may have the registered address, the primary operating address, you may have the registered or legal business name. Those could be additional data points that you provide to us that we can evaluate in our matching process.

George L’Heureux:
It's really about trying to take as much of the good data that a customer might have and compare it to what we have in our data cloud and our match reference file, and give really the greatest opportunity for one of those to match and give us a strong connection to a company in our data cloud. When I think about what I've seen in data in some of my previous roles throughout my career, I think of company name fields that have things like "The ABC Company formerly known as George's Donuts," right, and that's the kind of thing that you're talking about, right?

Janine Moreira:
Exactly right. We may see customers have two names in one name field, right, exactly what you're saying, "Janine Moreira DBA, Janine's Subway Shop." Ideally, those would be two separate names and we could attempt to match using both those names, right, and see, where are we getting the best match candidate returned?

George L’Heureux:
We do just that, we take all these permutations and we try matching with each of them. Then what happens internally? Can you step through that?

Janine Moreira:
Depending on your use case and your criteria, we would absolutely want to evaluate with you the best outcome. We have some standard matching logic for multipass specifically for the US. Globally, we do have some best practices, but ideally, what we like to do is talk with our clients and define their use case, define how the data's going to be used to determine really how we want to approach that multipass. It could be done a few different ways.

To your point, we talked about different information. Clients may also have non-traditional data points that they want to match with us, such as URL or domain, email addresses. They may have things like registration numbers or tax IDs, so we can talk about a flow of, "Okay, when your traditional name and address information's set doesn't meet your performance threshold, how can we look to evaluate these alternative data points to see if we can reach an achieved match candidate for you that would meet your use case needs?"

George L’Heureux:
When you talk about performance thresholds, you're talking about things like the confidence code or MatchGrade string patterns and things like that.

Janine Moreira:
That's exactly right. We try not to put it in a box of confidence codes. We certainly have clients, many clients who use confidence codes as a point of a performance threshold, but there are so many other things, such as match grades and match data profile points that we can use to create those performance thresholds with clients.

George L’Heureux:
In the end, though, really, what we're talking about is the same goal that we have with what we might call "traditional matching, traditional identity resolution," and that's trying to get to the idea of one best match, right?

Janine Moreira:
That's exactly right: Who can we return to you as the best match candidate that most closely resembled who you were asking about?

George L’Heureux:
Hopefully, that is an improved response, it's an improved answer because now we've got these multiple data points on the client's side that we're able to compare to the billions of match reference points on our side, whereas we might've had a fairly strong match before, that additional information could potentially take us to an even stronger match, an even more confident match.

Janine Moreira:
That's exactly right. You may also see not only just increased performance on your matches, but also traditionally where you may have only been setting us one set of name and address and there's this percentage of records that don't meet your thresholds that you have today, now, when you try and attempt to rematch it, potentially on these other data points you have, you may watch this percentage of records that we're not being matched at the level that you needed and just on the cutting room floor now meeting the acceptance threshold that you have, therefore increasing your ability for automation, right, auto-accepted matches and the insight that you're getting back.

George L’Heureux:
I would imagine that a lot of times, customers aren't even aware of the information that they have that they're not providing to us. They might have additional phone numbers or additional business names that they're not even thinking about, and so therefore not even sending to us. What tends to be the typical process for customers to understand, "Oh, I've got more data than I thought that I could use for matching?"

Janine Moreira:
When we initiate any kind of a matching project or matching consultation with our clients, ideally, we certainly talk to them about what I call it "the bare minimum," right? We need name, we need address, city, state, phone, country, preferably. The more you give us, the better match outcome that you're going to have. But we also talk about these additional data points with them so that they can start to explore internally within their own organization if there are different pieces of information that they can obtain. Whether it's through their finance, where they may be able to get legal name and legal address of the business, or with their marketing organization, where they may have CEO and/or additional contact information and/or additional names.

We try to explain to them, "We don't want to put you in this, This is all we can take, name and address." We certainly have standard layouts where we guide a customer, but we have to have those client conversations where we talk to them about, "If you have these additional data points, talk to us. Let's talk about where you're getting them from, how much fill you have, and let's evaluate how we can use these in our match performance."

George L’Heureux:
The interesting thing, I think, having been on the client side before, is when you have conversations like that, not only do they address the item at hand, the idea of, "Do we have additional things we can use for matching?" but oftentimes, they open up a whole new world of discussion for the client around other things that they could be doing with these data that they're rediscovering.

Janine Moreira:
Yeah, absolutely. There's a lot of conversations that come about through talking about what they have available. We'll bring up URLs or domains and how they may be able to use those matches, so maybe it's not a traditional name-and-address match they're returning back at a high confidence code or a MatchGrade string acceptance that they have, but maybe using that URL, they can get within the right family, and that can be used for specific use cases within their organization. There's really no data that has to be left on the floor. We can talk about all the different ways they can use the insight we're returning to them as well as what are the things that they can do to help improve that.

Why wouldn't a customer take advantage of this? Why wouldn't they take advantage of a multipass matching strategy?

Janine Moreira:
Sometimes it's just the effort. It takes time sometimes internally within their organizations to talk about where they could get these different data points. Other times, processes are so ingrained that they have to do a lot of heavy lifting to achieve how do they productionalize getting these additional data points into the system. Those are all conversations that we'll have with clients. Really, ideally, whenever we talk about multipassing, we're typically testing it out first with clients so we can show them the value that it brings. That's one of the things that we'll do on the data advisory teams is really work with our customers to prove that concept out.

George L’Heureux:
I think we've all gone by the idea that the more potentially good data that you can give to Dun & Bradstreet for matching, the better your match results are going to be. I mean, it feels like it might be considered almost an investment or beneficial to go hunting around for this data to some extent or another, right?

Janine Moreira:
Absolutely. If you think about it, if you're getting an 80% match rate today, and by passing some additional data points, you increase that to 85%. Now, you have a 5% lift. You think about potentially, if it's a supplier style, how much spend was in that 5% lift I have, I mean, potentially, you could be increasing quite a bit of resolution on some of your highest spend suppliers or some of your potential biggest prospects that you have, so there's a lot of value in doing it.

George L’Heureux:
To your point, I mean, first, I love talking about it in terms of customers, prospects, or leads, right? I mean, instead of just matches, those are things that really have a meaningful impact on your top line or your bottom line as a company. By thinking about them in terms of that's an opportunity for revenue, that's an opportunity for avoiding risk rather than just matches, I think it changes the discussion a bit.

Janine Moreira:
It does. Every time we talk to our clients, "matching" is certainly a familiar term to them, but you're exactly right. We want to make sure we're understanding, "Okay, how many clients are we resolving customers that you do business with or want to do business with? How many prospects are we identifying for you?" Again, all of that is the foundation, which is what D&B identity resolution is built for.

George L’Heureux:
Janine, as we start to wrap up here, let me ask you this: Someone's watching this or listening to this conversation about multi-task matching, are there a couple of key points that you'd say you would want them to walk away from this conversation with?

Janine Moreira:
I would definitely want them to walk away with talk to your D&B team. If you're evaluating what your portfolio looks like today, talk to your D&B team about things we can do to help you. Have that conversation. Think about the data points that you have today where you may say, "Well, I just have a remittance address." Are there other areas in the organization where you may be able complementary fill that record with more, whether it's the legal address or the legal name? Are there URLs that you have available? Then talk to your D&B team about what D&B may able be able to do with that insight and help you improve your performance and your ability to connect more customers to the D&B data or more prospects to D&B data.

George L’Heureux:
Well, Janine, I really appreciate you sitting down and taking the opportunity to chat with me today about multipass matching and the benefits it can have and how our clients could really start to take advantage of that to improve their match rates and potentially their bottom line.

Janine Moreira:
Absolutely. Thank you very much, George. Have a good day.

George L’Heureux:
Our guest today has been Janine Moreira, a data strategy consultant at Dun & Bradstreet, and this has been Data Talks. I hope you've enjoyed today's episode. If you have, I encourage you to share it with a friend or a colleague. Let them know about the show. For more information about multipass matching or other things that we've discussed on today's episode, I encourage you to visit www.dnb.com or to talk to your company's Dun & Bradstreet specialist today. I'm George L'Heureux. Thanks for joining us. Until next time.