Duration 30:04
16+
Play
Video

Value Engineering: How to Get Your Big Data Investment.....By Josh Siegel, Director, Hitachi Vantara

Josh Siegel
Director at Hitachi Vantara
  • Video
  • Table of contents
  • Video
Video
Value Engineering: How to Get Your Big Data Investment.....By Josh Siegel, Director, Hitachi Vantara
Available
In cart
Free
Free
Free
Free
Free
Free
Add to favorites
12
I like 0
I dislike 0
Available
In cart
Free
Free
Free
Free
Free
Free
  • Description
  • Transcript
  • Discussion

About speaker

Josh Siegel
Director at Hitachi Vantara

Global Value Engineering Lead for Hitachi Vantara. Josh has over 25 years of experience in business strategy and process, application and data solutions design and professional services delivery. He assists large enterprise clients with strategy and implementation around digital transformations and has worked extensively in Financial Services, Healthcare, Retail, Manufacturing, Life Sciences and other industries.

View the profile

About the talk

You’re a big data professional and you have the perfect plan: you know exactly what technology to deploy, how it works and how to integrate it. The challenge is convincing your colleagues in the line of business to fund this important work. Josh will go over how to build organizational alignment and move these important initiatives forward. Aspects of design thinking, organizational behavior and value engineering will be discussed as well as real world examples of both challenges and solutions.

Share

Best talk on value engineering. My name is Josh Siegel. I am part of an organization called Hitachi vantara, which is a global data and analytics company and service provider or part of the larger Hitachi family of companies. My background is in management consulting. I have 20 years spent a lot of time in the financial services industry and then about 6 years ago made the transition to big data analytics Ai, and I've done that in a cross-industry setting for the

last several years. I'm excited to talk to everybody today. This will not be a technical talk. Often in the Professional Services face when we help customers identify. Data science projects analytics Mi AI projects the process of moving a project forward can be difficult a lot of consensus-building. There's a lot of organizational difficulties related to defining the value of an analytics project and then making sure that you have funding to move it forward. So what I'd like to talk to you about today

is konstrukt that I use now and have used throughout the last 10 years. Bring people together around a vision for what data science ml in a I can do for an Enterprise. No matter how you cut it data scientist cost money, even if their offshore data Engineers cost money Cloud environments cost money governance cost money adoption in Immigration cost money. So if you're really trying to deploy something that's enterprise-wide and and and transformation. How do you get funding

for that? How do you bring organizations together? We all know on this call and I'm sure that data and analytics is a huge challenge right now and that's not a surprise let you read some of these statistics but trillions of dollars spent on remediating bad data 40% Only phone Lee 40% of the fortune Fortune 100 have been successful at creating what is now labeled a really a data-driven culture for making changes and moving away from Legacy systems in Legacy processes. If you're on the phone and your data scientist in the analytics space, you know how much time you spend looking for data instead of

actually doing the analysis on that did and it's not as if companies don't understand that there's a problem companies are throwing gobs of money at this. They're trying to deploy analytics. They're trying to build data-driven organizations. A lot of 90% of them. Are are are going to increase spending in the area that we all focus on. So, how is it that there can be such a big problem? And companies recognize it and throwing Dobbs the money yet, but we're still unable to to Really connect the dots. Lots of money

being spent on AI lots of money being spent on data-driven decision-making and end in improving culture. But according to Mackenzie only 16% of these digital Transformations are actually improving performance. And I'll pause there for a minute because at the end of the day, that's really what we all care about or we should all care of it. How accurate a model is you know, how technically Innovative aai or an ml model is might be very very exciting and we can write a technical white paper on it. But at the end of the day, it doesn't move the needle when

it comes to actual performance one of the things I'll talk about in this talk is how do you measure performance? That's a key component in what's called value engineering? so well, we know that there's a huge problem. We know the companies are trying to fix it. The issue really comes down to there's two sets of broadly speaking two sets of stakeholders around this issue. So let's think of ourselves as data professional data and data analyst data scientist. Boy, do we have problems?

Right you'll see this list. Some of this will really identify with you. I'm guessing there's silos of data. There's there's no Fields. There's lack of metadata. There's there's lack of interest from from data stewards. There's what what I call orphaned analytics different analytics platform scattered around the company each doing probably overlapping things and in a lot of issues standing in the way of of data analyst getting their job done and in providing high-quality analytics to the business.

Those challenges don't necessarily resonate with the business. What they are interested in is the the quarterly results. They're interested in making a Big Splash in the market. They're interested in moving fast and gaining competitive Advantage. They're interested in meeting regulatory deadlines. And as I'm sure you realize these problems are interrelated and they feed each other. The issue is that each of these groups Define the challenge a different way.

So from a data perspective, we need money to create high-quality analytics. But if you're talking to you know, the CMO for the CIO CEO. No VP is of a product in the business itself. They will describe the problem in the opposite way. They will say why I need the analytics to move the business power. So essentially what we have is it is it is it disconnect now the situation I'm describing may not apply to you and your in your daily life. But in my travels over the last 10 20 years this is this is

ubiquitous. This is a situation that occurs over and over again where two or more sets of stakeholders who really want the same thing at the end of the day, but they need or needs to be some some connective tissue between those stakeholders to really to really get them to come together around a problem. The way we think about this is what we call a Stairway to Value if you think about infrastructure data storage enriching that data with metadata or a data catalog. Then finally being able to use that curated data for high-quality analytics

analytics that the business will believe in. entrust entrust to make their decisions on That really leads to the possibility of monetizing your data and in taking the analytics and the Machine learning models that can theoretically be so impactful to a business and actually monetizing. And so what we see is the capabilities are given to the business as you move up the stairway to Value but funding for AI initiative funding for data scientists funding for

for the governance and for the stewardship in for all of the generation that has to happen only comes once that data can be monetized. So it's almost a catch-22 right value for the business can't be created without Predictive Analytics. I would argue more and more we've all moved on that, you know, Beyond business intelligence and and Reporting. This is now a world of Predictive Analytics. In Ai and ML and deep learning. So we can't really create a creative value additional value for the Enterprise

unless we have these things. But but we also then the value isn't defined by the analytic the values defined by what is actually impacted an impactful for the business. So this is where design thinking comes in. This is where value engineering comes in. Value engineering quite simply is the ability to build a business case that connects the modeling that we're all doing to the KP eyes and the success criteria that is impactful to the business. That's really the design thinking component. So when you think about it you can Define. value through a business case.

Anyway you want to What you really need to do is Define the value in such a way that the stakeholder who is responsible for funding analytics initiatives buys into that's the that's the statistic and that is the value kpi that they believe it not necessarily that you do and so it really becomes a function of a three-step process. There is an envisioning process where you bid we bring stakeholders together, you you you cross that divided between how business stakeholders think about analytics and how data professionals think about data and data quality in data analytics. That's the first step

is you build this common common understanding you build common ground. Then it's value engineering you define specifically what is the value that you're going to bring to the table and not at the end of a one year. Not at the end of building a massive data Lake and curating 25 different data sources in building the mother of all AI models. It's incremental value that's important showing that you can make a modest return in a in a really a near-term near-term

time her eyes. Once that's done, then the scale can come. So I can take you through how we think about this cuz I said the the in visioning process is really about stakeholder management. And this is something that I believe it's Mackenzie describes. It's almost like a data translator roll someone who can bring these two sets of stakeholders together. How do you define who will be impacted by your analytic exercise by the data science exercise that you're undertaking? Sometimes that can be tricky if you want to understand who the you know, the quote-unquote buyer

is who's funding this or who potentially could fund it you want to understand who the influencers are you won't understand whose lives will be impacted. If if you're rolling out a model to to predict what customers of an Enterprise are most likely to a trip over the next 3 months. Who needs to really know and understand and believe that this model is accurately defining. Who will attract? And when that happens, what is the Enterprise supposed to do about it? Do they have to retrain customer service reps? You might get a call from a customer who's

likely to a trip do they have to be prepared to offer up more money to keep them? There's all sorts of thinking that goes through what stakeholders to include in this first exercise. Then you'll go through a set of design thinking exercises to Define potential use cases and really work together to link those use cases to a a business outcome and that can be hard depending on what world you're coming from. If you're a technologist, if you're a data professional sometimes the the business

outcome that's important to you is not important to the business. You'll hear a lot of ideas in this first step in so it's important to listen and then to prioritize and then very very important is to build organizational excitement. The modeling that goes on in Enterprises these days is very exciting. It can be transformational. You could be talking about changing a business model. You could be talking about changes that will affect people's lives and livelihoods. It's important to really internalized the change and build that

excitement. Hitachi vantara we do this through a specific process. We call the it's called a digital envisioning process and don't want to get into a commercial about it. But there's a specific methodology and the results really look like this keeping it very simple trying to build build a bridge again between the stakeholders what's impactful to the business on one axis and how easy or hard will it be to implement this idea on the other axis and our whole supposition here are all hypothesis

is that analytics initiatives have to show incremental return? What would I like to call these big bang? You know, we're going to spend 50 60 million dollars on a whole new data infrastructure and analytics infrastructure Pro for a global company those go over like a lead balloon. There's no appetite to spend that kind of money because what return could you possibly get that could that could justify that amount of money? You can get investment of a couple million dollars and show that you can provide a return of half a million dollars

within 6 months or a quarter million dollars within 6 months that shows the business that there is value there and it's worth putting more money into it. So this is how the process works there's a lot of consensus-building and at the end of the day the business defines along with it and data professionals the best use case that will show that incremental return. Now, let's move on to the value engineering. We've talked about a lot of this and I'm watching the clock carefully. I

know we have about 14 minutes. so you've now gotten consensus with the business and you have that shared goal that use case that if data science if analytics can show a real movement on a kpi that everybody agrees with then the funding comes then the governance platform gets acquired. Then the the spin on the cloud analytics capabilities can move from 5,000 a month to $30,000 a month. You've got to show the business that it's worth it. You build out only what you need in this case. So that incremental ISM. If you will don't don't

load 25 data sources don't cure 825 data sources ringing only as much data as you need to show that incremental return. And that's that's what we do. We call it combining the art of the possible. With the reality of the Practical and we use value engineering to do this. So here's an example of how value engineering might work. And I'll distinguished value engineering. It's a term sap and Erp companies use it a lot. They have automated tools to do to showcase

the theoretical return of of deploying that software. This is not that kind of value engineering. This is this is bespoke custom value engineering. This is an example of something we did for a large not-for-profit in the United States. They have certain tears of donors. Those are segments sizes and the average donations those segments give a year. And what we were able to do after we did our analytics r value engineering on exercise on the analytics we did was to show that

we could identify customers are skews me donors in these segments that that I thought were operating like the next segment over but they weren't being treated like that. And so what this allowed the not-for-profit to do is to pay much more attention to the donors who would have been overlooked donors who are giving $25 look just like value donors who can give a hundred and buried within that value donor segment or another mm owners who have the potential to

give $5,000. If only this charity ask them. And so what we did is we showed this charity what the actual impact could be if they deployed this analytic in production and move past this kind of value Engineering in this proof of value Fizz and deploy this into production now to deploy this in production for this charity not a simple exercise. The analytics are rather straightforward, but implementing the change with people in the field can be very difficult.

And so that's why we had to prove to these prove to this charity that there was a significant amount of money that they could game. So that's an example. Here's one more example. This is this is from the Transportation space and amusement park operator. Actually whose issue is downtime of their amusement park rides and having trouble identifying the root cause of that down time and trying to reduce it. How do you how do you identify the value that can be saved in that situation? How do you turn that into a metric that the amusement park operator really

believes and understands and will free up investment to deploy these analytics across the whole park in hundreds and hundreds of rats. I'm having a mentor session later today. We'll be talkin about this example specifically and how we did this. And again, it's it it's a the function not only of what you think the value is and how you define the value but convincing the stakeholders who make the decision to 222 really think the same way. so once you've done that

once you've used value engineering to to build consensus into really Define, what is what is possible and what is incremental available in the near-term implemented in a limited production capacity or you shown theoretically what's possible in the lab now you have support Now you have stakeholders across the business who are who are not only supportive of the analytic the analytics and the processes in the investment. That's that needed to be made to make this production. Scale.

They not only supported. the champion it they wanted to happen. They need it to happen probably oversimplifying here. But you know many of us you run lines of business. We have goals. We have incentives to meet and if you've shown me that deploying this Predictive Analytics will move the needle on the results that I'm on the hook for then again. I'm not just going to support it. I'm going to be your champion. And so you have now you have now open the door to

you know, 222 a really impactful organizational wide operation end in the other thing you've done is you have become self-funded. Right. So if if your project to deploy production skill analytics is going to cost $2000000 in salaries in subscriptions in in Cloud cost. You have deployed the first little bit of that you're bringing in a quarter million dollars a year and you've shown that you can generate real revenue and you can fund some of your cost and you're going to be allowed to continue to grow continue to go. Then it becomes an effort

around marketing your success sharing that with others in your organization allowing your new Champion to share that for you and and not stopping don't take your foot off the gas you own this. This is only the beginning of the success that this modeling and data science can build for the organization. Keep going you have a lot more to do many of our customers. I know who are very excited about a model that works really really well, then don't focus on on The Last Mile,

you know, I'm a model itself provides a great result, but that result has to be implemented. You know, you have to consume the results yet integrated with production systems. You have to continue to tune and reach you in the mall. These are all things that you can't forget about because remember the end of the day the the real thing that we're all striving for is the impact on the organization that kpi that you defined way back and say is you know in Phase 1 of this effort in in those and those KP eyes,

you have to be those that are universally accepted earnings per share Revenue reduction in customer attrition. Those are the metrics that that need to be defined in and done in the value engineering face and then and then delivered on And then you repeat you rinse and repeat and I'll just share very briefly how we think about that. You know, you have a you have a a model that has been effective in your note in and now you're going and implementing that use case

across the Enterprise. Again, this is where you don't Coast you invest you go faster. What's important now becomes? Making sure that the business continues to use data analytics becomes a data-driven Enterprise. We have the technology technology is available. But what sometimes isn't available is is this governance process to to continually innovate and bring ideas on what can we do what might we be able to do run those down go through this cycle again of of proving

the value of value engineering. More than likely you're not going to have a a bottomless pit of data scientist. That's a that's a very hard resource to come up with and you'll have a decision to make you have one model and production let you have to spend time with it. You have to retune it data sources change organizations change. That's a living living breathing thing. You can allocate data scientist to continue to keep that model up-to-date and in production or you could move them to focus on the next model or the next business challenge. So those decisions have to

be made and they have to be made in a way that maximize the investment that the that the organization is make we call that the digital Innovation governance like so what I'm what I'm hopeful that that we reviewed today quickly, but hopefully made an impact with you. Big Data Predictive Analytics, obviously critical critical especially now to defining and bringing about the ability to monetize data Business intelligence no data warehousing orphaned analytics. All

of these Antiquated Technologies are holding companies back think about new entrants in the market all sorts of different Industries with no Legacy technology baggage behind them. They can move much faster and they can move faster because they recognize that you know, they're building technology. They're building a go-to-market organizations using I using these new processes in these new Analytics. This is as much a exercise about stakeholder management as it is about the technology. So focused on that that's your first goal is to bring everybody

together to build a shared purpose and to build a shared goal, which is to move forward with metrics that are important to the business. once you do that and you can't I spend on data management data infrastructure data analytics. Once you can't I spend on those things to business value you're off and running and as I said, you don't only have a a, you know a supporter. You have a champion. Your move incrementally your show and she'll show return an improvement as you go and if you can master that you've now found a way to become self-funding you found a way to

become Central to your organization's growth. And you found a way to use value engineering. to connect Predictive Analytics AIML all the great things that we know we can do today, but that not everybody understands you found a way to connect that to to multiple stakeholders care as much as you do but to think about things in a different way and and now we've we've you know where and we're in a position to really change organizations were in a position to really create

value for for the Enterprise is that we work for So I appreciate your time. I feel very good about it. Because I'm I'm I'm ending exactly on time leaving is going to be very proud of me. Feel free to get in touch through to email or LinkedIn have a couple more mentoring sessions one today and one tomorrow for those of you with specific questions. I'm excited to be available to you if if you have any questions.

Cackle comments for the website

Buy this talk

Access to the talk “Value Engineering: How to Get Your Big Data Investment.....By Josh Siegel, Director, Hitachi Vantara”
Available
In cart
Free
Free
Free
Free
Free
Free

Ticket

Get access to all videos “Global Artificial Intelligence Virtual Conference”
Available
In cart
Free
Free
Free
Free
Free
Free
Ticket

Similar talks

Sudip Chakrabarti
Venture Investor at Madrona
+ 4 speakers
Ian Foley
Venture Partner at Level Ventures
+ 4 speakers
Anupam Rastogi
Partner at Emergent Ventures
+ 4 speakers
Vik Sasi
Partner at Dreamers VC
+ 4 speakers
Dennis Liu
Investor at Ford
+ 4 speakers
Available
In cart
Free
Free
Free
Free
Free
Free

Buy this video

Video

Access to the talk “Value Engineering: How to Get Your Big Data Investment.....By Josh Siegel, Director, Hitachi Vantara”
Available
In cart
Free
Free
Free
Free
Free
Free

Conference Cast

With ConferenceCast.tv, you get access to our library of the world's best conference talks.

Conference Cast
561 conferences
22100 speakers
8257 hours of content