dataeconomy

The Economics of Data
Influencer conversations on data economics impact on business.
   6 years ago
#dataeconomyThe Economics of DataInfluencer conversations Data Economics: Driving application and business value to your bottom line
Keith A Thuerk
It seems w/ all the different types of data that their will be several different Analytic solutions in an Enterprise? Anyone else agree?
Tony Pearson
Agreed. Video surveillance data provides different analytics resource than POS transactions
Chris Saul
I'm not sure different data types is what I would measure by. It's more data usage and purpose
Dave Vellante
there are so many - more to come perhaps - vizualization, traditional BI, Excel, machine data systems
Guillermina Sainz
Agree, most, if not all, business data has value. That value changes over time and in relation to business mission, goals and operations
Keith A Thuerk I think you and I are on the same wave length. These too will evolve.
Chris Saul
So you could postulate an analytics system that used (say) BOTH video and POS data for some reason. Driven by usage of the data and not specifically because those happen to be different data types
David Floyer
In my view, there cannot be one set of analytics that will cover analyzing all questions and all types of data - there will always be a wide set of potential solutions
Karla Magana Renoud
Agree. Before you can get value, you need to first identify and define the business goal you are trying to achieve. Build your strategy around your business need, and then seek for the right analytics approach.
David Floyer
Data is both an asset and a liability!
Keith A Thuerk
Does data being kept long term need another review board to see how it might spawn different revenue streams?
Tony Pearson
I usually explain Information lifecycle management (ILM) as the process of managing data as it goes from asset, to expense, to finally liability.
Dave Vellante
the old discussion about auto-classification is relevant here - i.e. if you can't autoclassify you can't scale
fernando gomez
so it will create expenses and revenue if managed in a smart way
fernando gomez
but, is all kind of DATA really an asset?
AliyeErgulen
What data can we dispose of? Organizations need to keep data for 3 reasons: 1) it has business value 2) it has legal obligations 3) it has compliance mandates. All else is eligible for disposal. #dataeconomy
Tim Crawford
2& 3 are really the same. But you are assuming they know #1.
AliyeErgulen right. It is an organizational effort. When line of business units creates information, thye know what they are creating, why they are creating, and they might know to what extend they need the information..
Tony Pearson
But the problem is that data may not have business value today, but may in the future...
AliyeErgulen If we get that information about information it would bring insight about its value to the organization which could enable IT how to manage data based on its value.
Stuart Miniman
the line from #splunkconf has been "spinning data exhaust into gold"
Melissa Morales
Who actually makes the rules for what you can and can't dispose?
John Furrier
is there standards on what to keep or not keep with regard to data?
Melissa Morales
without guidelines it makes it difficult for organizations to get rid of anything
Brad Johns
Seems to me that different kinds of data has different value characteristics; Some data is kept to reduce risk (legal or business), some data is retained to help reduce costs; and some is kept to help drive revenue.
Keith A Thuerk
What metric determines data is kept or not?
Dave Vellante 1/ legal or compliance (binary); 2/ value (revenue, productivity...) 3/ retention risk
Dave Vellante
it's the age old question Brad...is information an asset or a liability - answer: Both!
Dave Vellante
How will innovations #bigdata and flash storage change application delivery and directly impact organizational value/productivity?
Tim Crawford
Flash storage increases costs and performance, but is not (today) a direct benefit to #bigdata. Analytics are a bigger issue than tech perf.
Levi Norman Flash if taken in the form of holistic view does NOT increase TCO...it reduces it...
Karla Magana Renoud
Its not enough that you can extract the information, but you need to do it at the point of impact. Timing is everything! and the right infrastructure matters.
Tony Pearson
#Bigdata could identify inefficiencies that help productivity
Melissa Morales
all-flash storage systems will boost critical application performance, gain efficiencies and strategically deploy resources for data management. All great end results for end users #IBMStorage
Scott Howser
Tim Crawford (@tcrawford) makes a good point. I would argue the real inhibitor to #bigdata is providing the business users with access to the data, and burgeoning ecosystem of associated analytic tools.
John Furrier totally agree making access easy to use is key for non techies (aka future data scientists)
Levi Norman
If you can gain meaningful insight to your pool of data and sort through it quickly, you can make effective decisions in a timely and efficient way...this makes for a more competitive/efficient business...
Guillermina Sainz
Flash solutions can provide real-time decision support for operational info & help improve performance of mission-critical workloads, such as credit card processing, stock exchange transactions, etc #dataeconomy
Lee Johns
Ultimately lower $/GB and $//IOP and improved latency will drive convergence in IT to center on application delivery,
Tim Crawford
Let's not forget that there is a theoretical side of data management and practical side of it. They are not the same. Lots of theory being discussed.
Stuart Miniman
care to share some examples/proof points of practical solutions that you've seen?
Tim Crawford From a marketing/ #bigdata perspective, connecting analytics to revenue increases.
Karen Hsu From a fraud perspective, you can identify patterns of fraud in big data (combining POS, transaction, authorization data). You can then prevent future fraud based on knowledge those patterns.
Karen Hsu From an operational perspective, you can identify patterns of failure in your products and then proactively replace failing components before being notified by the customer (internal or external).
Dave Vellante
I think this is a good point Tim-- So how can organizations focus on the practical and add value - vs spinning the wheels?
Tim Crawford First, start with a clear objective. Understand the business implications and follow the trail. No value = no path.
Karen Hsu
I agree. Practical use cases include: customer analytics, operational analytics, fraud and compliance and EDW optimization
Crowd Doc
How much data is bigdata?
Levi Norman
meaning what cross over volume?
CrowdChat Yes, what volume is bigdata? It changes every year with hardware advancements. There is so much more you can cramp into single machines now a days.
Chris Saul
I don't think there is an absolute number. ESG talks about how "big data" is a LOT more data than you are accustomed to
Chris Saul
So, for a small company that typically has about 100TB, "big data" might be 300TB
Chris Saul
Whereas for a large corporation, "big data" might be in the PB range
Brad Johns
On a theoretical note; just finished a good book on Big Data called "Big Data A Revolution", in which Big Data is described as the environment where the sample set for is "All".
Chris Saul
Translation: big data applies to everyone
Chris Saul
I don't think the hardware is relevant. It changes the economics of what is doable. But that's about it
Karla Magana Renoud
IDC is defining it as "deployments where the data collected (not stored) is over 100 TBs". Data sets that are not very large but are experiencing rapid growth rates of 60% or more annually are also considered Big Data under their definition.
Chris Saul
That IDC definition seems very weak. Lots of companies have >100TB but are not doing "big data"
Karla Magana Renoud I see it as a way to remove the concept that big data is only for massive companies. A small company may be experiencing big data conditions and could seek out a big data solution that fits its needs.
Chris Saul The ESG definition I quoted earlier is better for showing small companies that big data applies to them. Just saying "100TB" devlaues the term because then just about everyone is doing "big data"
Karen Hsu
As a best practice, we have seen companies:
- estimate impact on the business through an ROI and TCO analysis
- calculate ROI for each project
- reuse the ROI metrics (e.g. % cost savings or revenue increase) to estimate impact on future business
Karen Hsu
Have you seen this too?
Tim Crawford
Yes for other areas, but challenging in real practice WRT data.
Tony Pearson
Many clients I talk to have a hard time quantifying the ROI and TCO of their IT projects.
Keith A Thuerk So, do you feel that is more an IT issue or a Project Office issue?
fernando gomez
for that, we need to define the site infrastructure
Tony Pearson
Good job everyone! Special thanks to Dave, Mike and John. This was a great #dataeconomy Crowdchat!
Dave Vellante
What do people think about the notion of a "Data Czar" - i.e. a Chief Data Officer who drives data strategy, quality, governance...independent of the CIO
Keith A Thuerk
How does one ensure the data is all valid and not manipulated (it's all digital)?
Dave Vellante this is an issue of data quality - someone needs to own information quality
John Furrier
This in my opinion will be a new job that will be created more about compliance and API mgt but my concern is does this stunt innovation organically
Melissa Morales
agree that it needs a focus but with ownership and governance
Keith A Thuerk
Why would the solution be just one source is better than a shared validation method?
Levi Norman
My belief is that would create additional layers rather than enabling existing entities to expand and evolve...
Tony Pearson
Why should the "Data Czar" be independent of the CIO?
Dave Vellante the argument is the CIO has too much on his/her plate and is incentivized to keep applications running - not drive data strategies
Dave Vellante but many feel the CIO's job is to be the Data Czar
Tim Crawford
Do we need a Data Czar role for quality? Quality need to permeate throughout the layers that touch data. Otherwise, there is little/ no context.
fernando gomez
that would increase costs
Karen Hsu
If the organization can have czar, that's great. In many companies, we've seen a central group (e.g. center of excellence) agree on the specific roles and processes to ensure data quality, governance etc
Karen Hsu For Design and Production
AliyeErgulen
CDO and CIO are both required to be at the table along with Legal, RIM, Business, Privacy, and Security..