datamaster summit 2020

Tamr Customer Mastering Overview—Demo Days

 

Mingo Sanchez

Senior Sales Engineer

Gain unparalleled insights in your data to power customer 360. Learn how you can break down data silos to achieve high quality, consistent customer data in order to improve regulatory operations, lower their risk and enhance customer experience.

Master customer data across direct and indirect channels and manage customer hierarchies, identify cross- and up-sell opportunities and create a single view of each customer.

Transcript

Tags: business decisions, Datum, Data, Demo, Customer, Account manager, data sources, Salesforce, Subject-matter expert, Customer relationship management, Tamr, Operational system, Marketo, SAP, Lenovo, Address, Unique identifier, record, Different Recordings, Analytics

00:01 – 00:34

Mingo Sanchez

Hello, everyone, and thank you so much for joining today’s Tamr overview. My name is Mingo Sanchez. I’m a senior sales engineer at Tamr and today we have something really exciting planned for you. We’re going to be going through the whole process of how Tamr helps its customers master their customers. One of our most common use cases at Tamr, I think almost anyone can relate to the fact that customers are one of the most fundamental energy types that any organization does business with. If you don’t understand who your customers are, you can’t possibly make the most informed business decisions possible.

00:36 – 01:15

Mingo Sanchez

Now, one thing that we’ve noticed as an organization over the past several years is that big data is everywhere, but a lot of companies think that big data is a problem when in actuality having bad data is the more pressing challenge that they’re faced with. When you have all this data siloed across many different sources, across different teams, different regions, it’s impossible to bring together all those different sources and ultimately have that fundamental single source of truth that you can use to answer really simple business questions like Who are my customers who are my most important customers? How can I be cross-selling and up selling to those customers?

01:15 – 01:37

Mingo Sanchez

And ultimately, when you have data and operational systems that can’t be trusted, you can’t make informed business decisions now at Tamr. We think that human subject matter experts are the ones who understand your data best, so you don’t want some complex rules coded by a software engineering team or data science team making those fundamental decisions about what data to trust for your downstream purposes.

01:38 – 01:59

Mingo Sanchez

So rather than taking that legacy approach, we spend months or even years training highly technical teams to write these rules that ultimately break down when you bring in new sources of data. We instead take a different bottom up approach that we call data ops, where you’re constantly integrating those sources, getting feedback from those data experts, and then making that data better and better over time.

02:03 – 02:45

Mingo Sanchez

Now, fundamentally, all of this is in the service of three fundamental things driving the growth of your business, reducing the risk that your business faces when making those decisions every single day. And perhaps most importantly, lowering the total cost of ownership of that overall solution. Now, with Tamr Human got a machine learning approach. We’re able to help customers see 90 percent plus reductions in the manual effort that they need to spend on these data cleansing initiatives. And there are lots of different use cases at Tamr that we focus on working pretty much every industry, every use case you can imagine. But today’s demonstration is really going to focus on one of the most fundamental ones which is mastering that list of your customers.

02:48 – 03:36

Mingo Sanchez

Now, in today’s demo, we’ll be walking through the use case of Shield Inc. Now in today’s demo, we’ll be walking through the example of Shield Inc, as Shield Inc has gone through a lot of changes recently that ground through mergers and acquisitions and ultimately have many different CRM systems that they need to bring together so that they can make those informed business decisions. We’ll be touching on three different things in today’s presentation. First and foremost, we’ll be starting with those upstream source systems that those account managers are living and breathing every single day after walking through those upstream systems before and after Tamr. We’re going to show you a little bit of the magic of how Tamr is actually able to clean up those data sources. And finally, we’re going to show some examples of downstream analytics that are made possible only through using Tamr. Let’s dive in.

03:38 – 04:33

Mingo Sanchez

So now you can see that we’re in our cream of choice in this case, Salesforce now for the purposes of this demo, I’m going to be an account manager trying to figure out where I can find cross-sell and upsell opportunities to ultimately drive our business even further. Now, as a manager for Lenovo Set suspect that there are other Lenovo sites where I can find a recipe for success and ultimately sell more products to our customers. But scrolling through here in Salesforce, you can see that there’s a lot of information missing. Now I as a person know that Lenovo is an active customer of ours, but you can see Salesforce doesn’t have that most up to date information. And similarly, we don’t have a ton of information about the types of products they’re purchasing or ultimately how much they’re spending with us. Now I know that that data exists somewhere, but it’s up to me to go and find where all those different data sources lie prior to using something like Tamr.

04:34 – 05:11

Mingo Sanchez

And as I click into this account hierarchy view, you can see that there are lots of different sites for Lenovo within our Salesforce instance. And while some of these are indeed unique sites, it’s pretty clear to me that these two sites in particular are actually the same as one another. They’re slight differences in how those addresses are formatted and how the country code is format. They’re causing these to not be match together, something that rules based system didn’t catch before. Wouldn’t it be great if you had a system that you can use to provide that feedback to match those records together and then have it learn so that you can find similar cases going forward? Well, that’s exactly what you’re going to get with Tamr.

05:14 – 05:37

Mingo Sanchez

So here you can see a different view within Salesforce after we’ve gone through that data unification process with Tamr. Notice how there’s a lot more information in here, such as the fact that they’re doing business with us currently, as well as all the different products that they’re purchasing from us and how much they’re spending with us. So we can make much better decisions about how to do business with this customer going forward to scrolling back to the top here.

05:37 – 05:58

Mingo Sanchez

I really want to draw your attention to this new field that we have the Tamr idea. And that’s really part of the secret sauce of what Tamr has to offer is that users are able to provide that feedback link together all these records and then Tamr assigns a unique ID that persists over time so that you have all of that lineage linking together those different customer records.

05:58 – 06:34

Mingo Sanchez

Let’s dive into how that actually works. So now you can see that we’re in the Tamr UI, we’re no longer in our CRM, and here you can see a record for that customer, Lenovo. And you can see that same Tamr I.D. that we showed within Salesforce. So all of this information that you’re seeing in this record in Tamr is what’s being used to populate our CRM. Notice how we have a lot of information in here, such as that fully colon street address, as well as address information and even external information like a dance number.

06:35 – 07:11

Mingo Sanchez

Now, this information didn’t all come from one place. It’s actually by linking together many different records associated with this customer that we’re able to get this consolidated, trusted view. And if I click into this record, you can see at the bottom here that we have records coming from many different sources such as Marketo and Salesforce and SAP. Now, how did we actually link those records together? Well, that’s where the human comes in. It’s on the right hand side of this page. You can see that we’ve loaded and records from many different systems like those systems we mentioned before Marketo, Salesforce and SAP, and so on and so forth.

07:12 – 07:42

Mingo Sanchez

Now, as I scroll through here, two things probably immediately jump out at you. One is that we’re often missing information in some of these fields and other times we have different formats for the same types of values. So all things that would cause those matching rules to break down. Now, Tamr, you don’t need to have all of your data format the exact same way to adhere to a strict set of rules, but rather Tamr works with the data as it exists in those upstream systems so that you can ultimately jump from what you see here on the right.

07:44 – 08:14

Mingo Sanchez

To these groups or clusters of records that you see here on the left, now, going back to our Lenovo example, you can see the Tamr has grouped together these eight records that all belong to that same customer, Lenovo. And that’s despite the fact that we have many different representations of the name, different ways that these street addresses and cities are formatted if they’re even present at all, and many different customer IDs that are all formatted in different ways. Because, as we mentioned before, these records are coming from systems that weren’t designed to talk with one another.

08:16 – 09:00

Mingo Sanchez

Now, a user didn’t need to go in and manually find all eight of these records and put them together. All that we needed to do was give just a few examples to Tamr. And then it was able to learn from the subject matter experts and then do the rest with its machine learning. Now, was that process actually look like? Well, it’s really straightforward. Instead of asking users to find every single record associated with a customer, all they need to do is look at these pairs of records. And fundamentally and fundamentally answer is really simple questions about are these records the same as one another? Yes or no? And based on those examples, Tamr is going to learn all those same patterns that a person would be applying in their head.

09:05 – 09:33

Mingo Sanchez

So, for example, here you can see that we have two records, one for Lenovo Whitsett, one for Lenovo Co.. So obviously not the exact same name we can see. There’s a lot of similarity in these addresses and other fields like the zip code and the state, and Tamr is able to pick up on that as well. Where as a general rule of thumb, if a person looking at two records could tell whether or not they should be match together, Tamr is going to be able to learn that same information just need to give a few examples upfront first.

09:35 – 09:57

Mingo Sanchez

Now, if I were to answer this question and say that these records are the same, we wouldn’t be teaching Tamr just about these two specific records then said Tamr would take that as one data point. And when it looks at all of that feedback provided by your teams of experts in aggregate, Tamr is going to be able to learn all those patterns and ultimately take that knowledge that people have in their heads and apply that at machine scale.

10:01 – 10:36

Mingo Sanchez

Now, it’s great that you can bring together all these records with Tamr. But where do we go from here? Well, we’ve already talked about how we can link together all those different records from our upstream systems and assign that unique ID. And that’s fantastic, and we can also use that to create a consolidated view to feed those downstream systems. But those downstream systems don’t just need to be data stores, they can actually be analytics systems themselves. So here you can see that we’ve switched over to a dashboard in our analytics system of choice in this case, looker, and we’re able to get those views about Lenovo.

10:37 – 11:03

Mingo Sanchez

So you can see if I click into this record that we have all of these different sites that previously we had no visibility into. And not only that, but we can see all of our sales to Onavo by these different locations. So while in the wet set location, we have a ton of products that we’ve sold. You can see that we haven’t yet applied that same pattern to other sites like in Wuhan or Chicago. So now we have a pattern for success going forward so that we can grow our business.

11:07 – 11:51

Mingo Sanchez

Now, in conclusion, we should have three really important things in today’s demo. We’ve shown as operational systems that our account managers are using day in and day out, and we’ve shown how Tamr can make those systems better. And we’ve also shown what that process looks like in Tamr so that your data experts can ultimately train a machine to do all that same work that a person would have to do manually without using something like Tamr. And finally, we’ve shown some example analytics for how once you have that cleansed data, you can make much better decisions about how to operate your business. If you’re interested in learning more about Tamr, I strongly encourage you to visit our website Tamr dot com or reach out directly to someone from our team. We’d love to connect with you. Thank you so much for your time and have a great rest of your day.