Episode 7 – Discussing Data Strategy with Kazuki Ohta

Suruchi Bhargava [0:04] Hey there, thanks for tuning in! And welcome to the MartechCube podcast. Each episode is a big step forward for enhancing and optimizing your MarTech stack.

I'm your host, Suruchi, and I can't wait to tell you who we have with us today. Our guest is the perfect embodiment of entrepreneurship and a computer science genius.

Here we have- the one and only- Kazuki Ohta, the CEO, and co-founder of Treasure Data.

I am so honored to be talking with you today, Kaz! Treasure Data is indeed a treasure! The best- of-the-class enterprise customer data platform has received several accolades and the best customer reviews. But before we dive in, I just wanted to congratulate you and the entire treasure data team on being recognized as a leader in the IDC market scape lately. So congratulations, and thank you for joining us today!

Kazuki Ohta [1:11] Thank you for having me Suruchi. It's my honor to be on MartechCube

Suruchi Bhargava [1:18] So, to start off, the COVID-19 pandemic compelled many businesses to accelerate digital transformation. But it’s been challenging. What groundwork does an organization need to get started with CDP?

Kazuki Ohta [1:35] I think every organization has to understand the market change, the dramatic change. And I would say, I want to point out three changes that are happening. One obviously is the digital disruption. So according to McKinsey, 60% of the customer journey is now happening in the digital world and this percentage gets tripled after the pandemic and this percentage is higher for GenZ new generation consumers. So any business has to one, understand the customer by data and then two, use the data to better personalize the experience. And second, at the same time 10 years ago it was a big data era where you collect everything about your consumers and customers and do whatever you want with the data. But now there are so many regressions and privacy concerns are happening amongst all the consumers including myself, to be honest. So every company has to comply with something like GDPR in Europe CCPA in California and 150 countries out of 200 countries has put consumer data protection law and it is getting stricter and stricter. So any business has to have a centralized customer repository to manage customer data. Otherwise, you will be fined and sued. So those trends will continue next 5 to 10 years. At the same time, we also have these tactical changes where in 2 years a lot of browsers and mobile devices will become anonymous. Like 75% of web browsers will become anonymous because third-party cookies and IDFA which are like very important mechanisms or framework to track the ID of these devices will be dying. Right? So every marketer or organization has to have a first-party data strategy meaning you have to earn the data by themselves and I think the groundwork you need is you have to admit and understand those three disruptive trends deeply so that you can take actions.

Suruchi Bhargava [3:49] Yeah, like without trust, the entire experience is incomplete.

Kazuki Ohta [3:53] Yes

Suruchi Bhargava [3:54] So the customer’s journey is no longer linear. In the digital age, customers go to different channels for more information. How can CDPs help marketers deal with a multitude of touchpoints and turn customer data into actionable insights?

Kazuki Ohta [4:13] I think in this digital world it's really really hard to understand the customer by talking. Right? Like 10-20 years ago, customers coming to your store or retailer and then understanding customers what they need- that is all gone. Everything is databased and there are multiple channels and multiple devices and multiple platforms, they are interacting with. The only way to understand the customer is now becoming more like data-oriented and what marketers should do is merging the art or credit part with the instinct coming out from the data and that's becoming the central skill set for all the marketers right now.

Suruchi Bhargava [5:00] Yeah, the unification/ integration of data makes the whole customer view clear as day!

Kazuki Ohta [5:07] Yeah

Suruchi Bhargava [5:08] Get it. There are many conversations around a Single Customer View (SCV). It has an organization-wide impact, increasing customer loyalty, and lifetime value. What are the challenges in creating a single customer view, and what are the compliance and privacy concerns associated with it?

Kazuki Ohta [5:22] Yeah. So, based on our survey, a typical enterprise has 170+ more applications and tools which has a piece of information about the customers, like on an average. Okay. So the challenge really is ..okay..if you have 170 systems holding a little bit of customer data you really can't have a single customer view, but, first of all, you have to identify what are these data sources, what is the volume? Are they coming once every day or the batch or can we become real-time, right? And so those are the technical aspects. The other aspect is organizational friction, right? Okay. Your marketing team or customer experience team trying to help a single customer view but lot of these applications and other sources are managed by different teams. So just by having the access or connection or having more comfortableness of having other teams actually giving you the data as a centralized, you know customer data management function is what we are seeing lot of challenges to create a single customer view.

Suruchi Bhargava [6:41] Right. That sums it up. You have had some big clients, like AB InBev and Subaru, how have they used a CDP to dramatically increase customer-centricity across the globe?

Kazuki Ohta [6:57] Yeah, so AB InBev right now consolidated more than 1000+ data sources into one. They are the largest beer and brewery manufacturers on the earth having brands such as Budweiser and Corona light of the world. I am sure a lot of people drink one of their brands. The problem is, each brand in each country is using very very different set of tools and although the customer data was siloed, they were also relying on third-party audience or advertising platforms with third-party cookie dying, the efficiency of the advertisements is going down, and down and that they need to personalize it, rather than going into the mass marketing. So they tapped into treasure data and then we created CDP across 500 brands in more than probably 40 countries by complying in with the low core regulations as well and they are using the CDP for personalized advertisements, paid media lookalike, and also optimizing all of the campaign across multiple channels. And more importantly, using CDP to comply with like all the consumer data protection laws.

Subaru is very interesting. One, they use treasure data to capture 800 million digital and physical customer touchpoints. Meaning the moment they use website go to, you know, mobile site using mobile app advertisement, email open, click, or in the physical experience, they, the prospects go to the dealership or even start driving the car. We're capturing the telematics data. And we created the single customer view from all these data sources. And we apply machine learning models to predict who is going to buy which model, which add-ons, you know, if you think about how you buy the car, you do a lot of research online and then go through dealership and then buy the product. Right? So a lot of digital and physical journey gives you a good overview of the prospect’s buying propensity. So by using this model, 90% of the chance we can predict who is going to buy a car, right? So the dealership sales rep can better spend time on better prospects rather than not. You know, um, better spend on the better prospect. So we increased the dealership conversion rate from 17% to 31% and drove hundreds of millions of dollars of ROI.

Suruchi Bhargava [9:43] Wow. That's quite a way to break down silos and, you know, enhance marketing in real time.

Kazuki Ohta [9:49] Yep

Suruchi Bhargava [9:50] So organizations can’t take piecemeal measures when it comes to data. Data readiness is a critical component of data mapping. With colossal volumes of it available, how can brands identify which sources should be integrated with the CDP platform?

Kazuki Ohta [10:07] I talked about, you know, on an average enterprise has 175 data sources, right. But the same time you can really do everything at all. So what we recommend is first of all, you have to identify a key campaigns or use cases you can run on top of CDP, more law hanging fruit one, and then identify a couple of data sources to start with. I mean, typically what we're seeing is for example, you wanna, send a private, personalized email campaign to do it. You wanna see the past behavior in the website and a couple of customers attributes. So we often ended up with the first project by unifying CRM data and also website and mobile behavior into it, and then running the personalized campaign for the email and get that, you know, quick, first ROI from CDP. And then after that, you will think about more campaigns and then you identify more sources to integrate. So that's how we see our customers and prospects evolve the use of CDP over time.

Suruchi Bhargava [11:24] Okay. Makes sense. Customer data platforms have reframed customer experience. How does Treasure data improve customer lifetime value and beef up engagement? What role is it playing in martech innovation?

Kazuki Ohta [11:40] Yeah, so I think the whole reason why CDP is becoming important is there are 8,000 plus martech tools available in the market, right? No vendor can provide everything, you know, as a suite. So people ended up buying bunch of tools and then each business is using different set of tools. Each country is using different set of tools, but then you're lacking this like a single customer view. So that's why this category called CDP was created to unify all of the data and then activate these profiles into like thousands of tools. And then in fact, Treasure data is now activating 40 billion profiles every month to the entire market and advertisement ecosystem. Just to give you a quick close, we only have 7.5 billion people on the earth and out of which 4.5 are the internet connected one. So that, and then we're activating 40 billion profiles every day.

So you can see how important CDP is becoming for the entire martech ecosystem. Now, going back to your question, how treasure data can improve customer lifetime value. So a lot of case, you know, it's very simple, you know, first of all, we can acquire more customers and then we can do upsale and cross sale. There are certain industries such as retail, banking or insurance. Sometimes retail up sale and cross sale is becoming more important, right over time to create a loyal customer base. And then data driven approach can dramatically increase the loyalty across all of the customers. Right? And then obviously, we can reduce the churn. If you can see some signs of customer complaining or leaving, then you know, we can have multiple campaigns to prevent it. So that creates the entire customer lifetime value average increased.

Suruchi Bhargava [13:56] It's about pivoting. Yes. The make entire experience, more seamless. In the era of conversational eCommerce, CDPs are proving to be the biggest competitive advantage for retailers, the enterprise platform, which is a gap created by legacy platforms, CRMs, DMPs, what do you think are the fine developments retailers can orchestrate with CDPs? What's the role of machine learning models here?

Kazuki Ohta [14:22] Yeah. So a lot of retailers right now are trying to embrace this new era of both digital and physical channels. Right? So traditionally people just go to retail shop and then research and exploring the store and then purchase it. But now the behavior is like, okay, you go to e-commerce side, look around the product. And then you go to retail shop, look at the actual product and then purchase. Or they can even just go with the eCommerce, right? So I think retailer should adapt that the customer journey is now being more, becoming more complex and you have to unify physical behavior and digital behavior into one. And then trying to predict how can you maximize the revenue coming out from like both channels rather than focusing on one or the other. Now the machine learning model can help is first of all, we can apply multitouch attribution model so that you can identify which campaign is helping the most to generate more revenue or selling more product.

We also have a customer lifetime value prediction models like we saw in Subaru, okay. This customer will more likely to become the loyal customer than others. So then you can spend better time or campaign dollars onto those customer base, which will generate more leads. And then for example, in the email’s case, you don't want to spam everyone with these, you know, all of the campaigns, you might have more targeted, personalized email campaign, which actually generate better results by sending fewer emails or messages, right. That also comes into play with this like a prediction model too. Now we have another six to, you know, seven machine learning use cases, but I would say those are the major ones.

Suruchi Bhargava [16:40] Right. Efficient segmentation can really help marketing break barriers. So moving onto the big one, in November last year, Treasure data raised 234 million$ led by Soft bank. Not only enormous, but the largest funding ground for a customer data platform. What does the roadmap for the company look like in the future?

Kazuki Ohta [17:05] Yeah, so I was, um, CTO, chief technology officer, and I became CEO last June 1st. And then we raised, um, 234 million by SoftBank.

Suruchi Bhargava [17:07] Whoa

Kazuki Ohta [17:10] We created this vision called connected customer experience. Meaning, you know, for example, I've been a cable internet, providers customer for 11 years here. And every time I was the highest, TV and internet plan, okay. Every time I call them, I need to wait like 30 minutes and they don't know anything about me after 11 years, they're even asking my plan and they should know it. Right. And they're asking my phone number, even though I'm calling from my phone and they transfer my line to another one and asking my phone number again, it's just like a, such a bad experience in the contact center, right? So with this connected customer experience vision, we created new product lines called CDP for service/ contact center and CDP for sales, which allows contact center, contact center agent, and then sales agent to look up the data within CDP.

So that contact center can solve the problem quicker, route a VIP customer into less crowded queue. And also more importantly, by reflecting the web behavior of that particular customer. They exactly know what type of problem they're trying to solve or what type of product they can. They are now searching to so that you can actually sell at the contact center level so that you can turn contact center from cost center to revenue center. Now CDP for sales allows all of the sales agent to look up the customer data from the tools or devices they have so that you can better sell. Like we saw in the Subaru's dealership sales example, right? And we have so many industries, for example, the salesperson in the retail store or someone selling the medicines, someone selling insurance, retail banking, right? So, and then automotive dealership, anyone who is selling some product should have an access to CDP so that they can take better actions.

And with this connected customer experience vision, we're trying to release more product lines so that we can unlock the power of customer data across all of the enterprise. And then our goal is to every customer facing functions have access to CDP so that they can provide better experience and increasing the customer's lifetime value. Some of the interesting use case we're also hearing is like supply chain management. Meaning if one particular region wants to have certain type of product like the car or the beer, then we should ship the right amount to that region. After pandemic, the supply chain costs has gone up quite rapidly. So sending the right amount to the right region will not only reduce the cost for supply chain increase the revenue opportunity. Right. And we're also started hearing, but using customer data, what's the right amount of staff we need to have in the exchange store or the retail store in the manufacturing process. Right. So that's another human factor to it. So we started hearing all these like interesting use cases. Yes. Our core is about the martech and the marketing department. But over time, our vision is we would like to unlock the power customer data across all of the enterprise again.

Suruchi Bhargava [21:14] Wow. That's perfect. So being the CEO and co-founder of Treasure data, how has the journey been for you since you started in 2011? As a leader, how are you shaping the company culture and connecting people to purpose?

Kazuki Ohta [21:29] Yeah. So that's a great question. You know, when I started the company, it's just three people. I was 25, nothing to lose. I was in Japan. I couldn't speak any English, but nothing to lose. If you wanna start a software business, just come to Silicon valley and figure it out. And then, you know, I was lucky enough to start this company, right. And I majored in computer science when I was in university and my specialty or background was HPC (high performance computing). So my professor has built the world's fastest, super computer, basically half a million computers combined into one. And I was a part of the team to build the file system for that super computer. And obviously that gave me an opportunity to handle large amount of data and also made me realize how important data will be in the world.

So that's why we started this company, treasure data to build the enterprise grade CDP customer data platform. The realization is the customer data is the most precious type of data in the world. And then we decided to focus on it. And, you know, as a leader I've been going through quite a lot of change. You know, I feel like for me as a founder, I need to change myself every three months, right. The role. But at the same time, you know, my job is just always the same. I need to hire great people. I need to make customers happy and then shipping great product. So I always tell this to the entire company, but people, customers and product, that's all it matters. And, you know, I can't empathize enough that like, just keep focusing on having happy employee, happy customer and the best product.

Right. And then connecting people to purpose is, you know, another interesting one. So we have around, uh, 700 people across 20 countries and it's quite an international organization, but more importantly, 400 people joined after pandemic. And for me, it's very important to connect all the people remotely. And a lot of people even haven't met any colleagues face to face, right. But again, I think at Treasure data, we have this huge opportunity where there are 4.5 billion internet connected people on the hours. And we, you know, just a, you know, a brief expansion, but we have half of the profiles inside our CDP platform, across 450 customers we have. So every employee has an opportunity to influence billions of people by working at Treasure Data. And at the same time, we also need to have a social responsibility to protect their data as well. Right. And then in the martech space, data and privacy are the two fastest growing market. We are the intersection of it. So that's why I think a lot of people here at treasure data, including myself, is getting excited every day because it's just so much opportunity out there. The market is growing and we have the leading product and that's the purpose of working for me.

Suruchi Bhargava [25:11] That's literally the leaders guide to creating an exceptional culture. Well, I'm stunned by how much you have to offer to the world, Kaz! You’ve been so great. Thank you for taking the time out to share such an incredible journey and some straightforward takeaways. I'm grateful for the entire Treasure data team to make it happen. Appreciate the insights. Thank You!

Kazuki Ohta [25:35] Thank You!

“Any business has to have a centralized customer repository to manage customer data. Otherwise, you will be fined and sued. So those trends will continue next 5 to 10 years. At the same time, we also have these tactical changes where in 2 years a lot of browsers and mobile devices will become anonymous.”

“First of all, you have to identify key campaigns or use cases you can run on top of CDP, more low-hanging fruits, and then identify a couple of data sources to start with.”

“People end up buying a bunch of tools and then each business is using a different set of tools. Each country is using a different set of tools, but then you're lacking this single customer view. So that's why this category called CDP was created to unify all of the data and then activate these profiles into thousands of tools.”