What is a Data Scientist? I know I am not one…

Below is a lovely infographic of what it takes to be a data scientist.  I can say that I have plenty of the domain knowledge, soft skills and a good amount of communication skills, but bayesian inference – please.

data-scientist

I accomplished my MBA, but I will humbly state that statistics almost made me quit.

The point here is to fully understand that depth of skills and reason these people are rare and in short supply.  Art-template

If you find one – hold on to them because the are like Marino Rivera in your bullpen.  rivera

The closest thing to a sure win against your competition as there could be.

So what is one to do if they cannot find this ‘perfect’ person.  Well a team of 3 would work very well I say.

  1. with the deep business knowledge, real world experience and ability to go toe to toe with c-level executivespower-rangersa number two with the deep statistical math capabilities – a phenom who looks and can talk numbers the way I can about single barrel bourbon or a ribeye
  2. the third is the database/developer who alone with Python and R, is a wise with Hadoop

Number 1 or number 3 could fulfill the cloud and data migration experience or that could be a fourth role.

If you are in retail, financial services or healthcare right now this is a NECESSARY role in your organization do don’t wait.  Other industries for the moment might be able to wait until 2020 before needing this role on a consulting basis every quarter or so.

dont-wait

 

 

 

Video article: big or small data no difference

If you look through this blog you will see analytics pop up often as I am a strong proponent that analytic is needed everywhere in every business.

However, as Ms. Steeley Reed states in this video, just measuring and reporting on everything is not effective for your business.  She speaks of make the data actionable.  images

More so, she points out you must measure and report on what impacts the customer and I will add visualize that data to make it actionable.Capture

Having actionable data focus means that you move away from reporting on everything and focus on both what impacts your customer (internal/external).  Then present that data in a visual manner to allow the employee, manager or executive to take action.

Therefore, big data or small comes into play here.  Start small with using the readily available data to create actionable data visualizations.  Then add to that data from the harder to get to structured and that unstructured data that requires data governance.

Be mindful here of data that is available in real-time video, social media and so forth that you can stream into visualizations to provide improved customer engagement immediately.

  • Tracking customer wait lines in videos streams to start offering discounts on product/service items that are quicker to deliver

 

  • Display support line hold times and hang ups to contact those customers to offer them alternatives to phone support

 

  • Health providers leverage mobile technology to track out-patients health conditions to detect needed adjustments treatments

Many other to be mindful of as you progress.

Business functions separated by a common language.

Two nations divided by a common language“, this quote normally attributed to Oscar Wilde is about Britain and the USA.  Brit and the USSince my daughter has become engaged to a Brit, this is a frequent occurrence in my family.

 

 

 

So when I was reading the latest Gartner CIO survey results and read the below chart –

Top 10 Business Priorities

Rank

Top 10 IT Priorities

Rank

Increasing enterprise growth
1
Analytics and business intelligence
1
Delivering operational results
2
Mobile technologies
2
Reducing enterprise costs
3
Cloud computing (SaaS, IaaS, PaaS)
3
Attracting and retaining new customers
4
Collaboration technologies (workflow)
4
Improving IT applications and infrastructure
5
Legacy modernization
5
Creating new products and services (innovation)
6
IT management
6
Improving efficiency
7
CRM
7
Attracting and retaining the workforce
8
Virtualization
8
Implementing analytics and big data
9
Security
9
Expanding into new markets and geographies
10
ERP Applications
10

My first thoughts were to the quote above.  When the line of business, be it finance, marketing or another department talks to IT or vice versa you can see from the chart above that they are speaking English, but not the same language.

On the left business is saying what they want to DO and on the right IT is saying what they should BUY and IMPLEMENT.

The issue is who is ensuring that the right side from IT will DO what business wants on the left side.

That is an important exercise that both line of business and IT must do is not just at budget time, but throughout the year.

Let’s start with Increasing enterprise growth.  growthWhat is IT doing on the right to increase growth?

1. Analytics and BI can identify untapped markets, great customer sales potential and run tests to improve market success of new products/services.

2. Analytics, Collaboration technologies and ERP apps tied to Mobile technologies can provide field sales the ability to outshine competitors in front of customers.

Delivering operational results – results

1. Again analytics and BI can play a large role in provide visibility to internal processes, tie that to mobile technologies to speed decision-making and leveraging optimized ERP applications to improved processes.

2. The most obvious is that Cloud computing, Virtualization, and Legacy modernization can reduce IT cost and thus help the bottom line.

Reducing enterprise costs – see above  reduce

We can go through the entire list, but you get the idea.  The important note here is that both IT and the lines of business should have their strategies and tactics matrixed to clearly demonstrate alignment is clear.  Also, they need to communicate, measure and promote this alignment throughout the company during your fiscal year and across years for longer range projects.

Performing a constant translation is important, as this year it is Cloud, Analytics, and BYOD requiring translation from the IT to the LOB side.  Depending on  your industry there are several business terms IT will need the business to ensure they have clear translations.

Book Review – Super Crunchers

Author Ian Ayres

Not the fastest reads as the author and the many people he sites like to use their university vocabulary.  I met the author at a speaking event where he spoke of the book and with a number of others I received a signed copy.  While the book held my interest, he is a much better speaker.  If you have the chance catch him giving a talk.  1 supercrunchersIn short, this book explains the use of number crunching can dramatically improve everything from medical to marketing successes.  It does focus on the fact based decision making versus intuition and past experience alone, which the book and frankly my experience is a major culture change for the majority of businesses.

–       Keeping an Eye Out for Chance on page 69  1 math

  • Along with analytical insights from crunching numbers that exist, the tools available today allow experiments to be run to quickly determine a successful course of action.

–       Follow the herd versus evidence based tactics on page 89

  • Regardless of level of education and experience we develop biases and that will drive our decisions.
  • Using evidence based information systems corrects for bias.

–       An average of 17 years for new knowledge generated to be put into practice on page 91

  • In the medical field, it can take that long for a new discovery to become procedural due to habits and bias.

–       1 cokeCoke-can predication on page 114

  • We can predict that a coke can will spurt if shaken and that is how we often rely on a experts opinion.  We treat their knowledge like it is absolute, yet have they continued to refine it, add to it and question their own expertise?

–       But What if it’s Wrong on page 184

  • The culture of the current day super crunchers are different than the ‘older’ analyst, they like their social media they share everything, publishing the data.     1 y generation share

Is this book for everyone, no, I recommend mainly for Chief Information Officers and Marketing Officers.  Like I said above definitely try and hear Ayres speak.

Ah, you don’t need that analytics stuff, Excel is just fine.

1a ExcelIt must be true that business does not need analytics as the majority of businesses are still predominately using Excel for both a majority of reporting and performing analysis for decision making.

So, those CIO’s that Gartner, Forester, and CIO.com find to survey each year are somehow mistaken.  1 senior executive

Just like those 7 out of 10 dentist.  In this case, the 97,500 IT leaders believe that Excel is just fine and frankly is as much as their poor feeble users can understand and use.

1 nj

Have you picked up on my Jersey sarcasm yet?  

Well business people, just like 20 years ago you did not think you needed the internet or email, you are wrong – D-E-A-D wrong that Excel is enough in TODAY’s business climate of information driven decision making.

Go ahead, roll your eyes.  1 rolling eyes

You to Finance, Sales, Marketing, or Operations people.

How many decisions are made each day based on old, limited or even plain wrong information?  Decisions of which projects to approve/cancel, products to launch, reinvest in or remove from the market amongst other major changes to your business that could either drive revenue/profit to meet/exceed your strategy or fail to do both.

You are doing all of this with Excel?  Really? Maybe a reporting or query tool that is the hands of your IT group.

1 bad processYou know the process.  You have a business question.  You do not have access to the data at your                   finger tips.  You ask IT for the answer in the form of a report, a query or both.  You wait, I don’t know a day, a week, 3 months or forever.  The problem there is did you ask the right question?  How likely are you to go back to IT if it took a week to get the report with the answer to the wrong question for a new report?  I ran IT teams with this process.  We tried to optimize this process, but never eliminate it.  Wrong was I.

And you IT people stop treating end-users as if they are dumb, slow or completely resistent to change.  They are running the  business BUT  – they are handicapped by having them run it with a narrow limited view of said business.

So why the hell is it not every business leader, not just CIO’s, making analytics the top goal RIGHT NOW.

It is new.

What if we get it wrong?

You are doing okay business wise, why change.

Damn it man/woman you are successful business people running multi-million if not billion dollar businesses with hundreds to tens of thousands of customers.  You have like 3,000 different business processes, several 100 to 1,000’s of people working for you and millions upon millions of dollars that you are letting just go to your competitors because you won’t take the simple challenge of getting the analytical insights into your business/industry that you damn well know will make you and your business smarter, faster and bigger.

1 obsoleteExcel does not have the answer, it won’t go away, but it is a over 25 years old, thus a shiny antique of an application.  There are cool modern tools that you can start off small, simple and build from there great decision making capability.

Do it today.  1 now later

 

Analytics ingredients

Working with IBM you quickly learn their Analytics mottos.
  • How are we doing?

  • Why are we doing that?
  • What should we be doing next?
Insights lead to smarter decisions which create better outcomes
In other words, those who know more about their business are able to grow and profit exponentially better.  
You know this is the goal, dream, vision or whatever you call it of most managers and above in business.  If I only had the answers to the questions I ask and also the ones I don’t ask (insights) I could blow through my business goals.  Taste that fine wine of success plus plus.
How to get that though? The ‘right’ way is to holistically look at your enterprise, all the data sources, run a project to improve the quality of that data and keep it high quality, then work with each functional area and outside industry experts to determine the decision making processes, what they should be, design the data warehouse, build up your analytic metrics, dashboards, add planning and so on.  If you read that last run on sentence really fast that is what you likely have been proposed in the past.
I have been through that and the 15 month internal selling cycle to convince other board that this is how to get the magic sauce.  Then re-convincing the each month of the year long project to keep the resources focused on the effort.  I see the political campaigns for president of the US right now and can sympathize.
One mistaken statement, one deadline missed, and crisis after crisis and the campaign is a rocky ride.
  • You can do it with less pain.
  • You can do it in less time.
  • You can do it with little drama.
If you are IT or CFO, you need to stop thinking of winning a battle, never mind a war and think like a virus or a weed if you like.    You need to get planted somewhere.  Anywhere at all.  In that nice friendly place and then help it grow by both seeding it with funding for additional licenses, performance and most importantly giving up control over how it grows and where.
Yes, you provide the money and have a decreasing say it how it is spent.
Purchase an analytic solution, heck buy two.  Install it in a small group area who has been begging for it.  Give them funding for training (a good package like Cognos TM1 wink wink is intuitive with its Excel interface or BI with the single workspace), dedicated development for a short window and hardware performance for the first 45 days.  Well the hardware performance is permanent.
Then leave them alone, kinda.  Back away and count to 10. In business terms it could be more like 30 days.
Feed them with praise for how they are using it and fund more licenses even for those people you wonder why they need it.  Listen for an area that it could land next and do another launch area there to repeat the above.
Then expose the examples of the insights, smarter decisions and better outcomes to the sunlight of executive visibility to help it grow further.       

Book Review – The Deciding Factor Book Review – The Deciding Factor

  Author Larry Rosenberger and John Nash

 

 

 

As I focus more on analytics in my professional life I have started going through the major analytics books.  I previously wrote about the Chief Performance Officer book, but have learned a great deal more in the past couple of months.  The Deciding Factor takes the non-tool approach to explain the power of analytics and especially power of expanding the use of analytics from the executive offices to the day-to-day operational staff to both improve their decision-making and better execute the strategy.

–       Capital One’s story of how they used analytics – page 18    

  • How impactful analytics have creating a product set and growing a boring business into an exciting marketing effort.
  • Continued with Progressive on page 21

–       Tesco’s use of analytics to define customer categories on page 41

  • No explanation here except the main categories used of time-poor, food-rich, can’t cook and won’t cook – just love these

–       Best Buy’s push analytics down to their sales force on page 47

  • Training their sales force using specific lifestyle questions to define the product sale match with the customers needs and discover up sales potential

–       Fair Isaacs chart on page 80

  • Chart displaying how to ensure that decision making with analytics to ensure the strategy executes well across all channels
  • Accomplished with a basic rules based management system

–       Work flow vs. decision flow on page 127

  • The impact of how a work flow is impacting by a decision flow and why you need both

–       The difference between business intelligence and decision management on page 141

  • In short, BI focus on how we are doing now and why and decision management builds on the why and provides options on what’s next

If you have a BI solution and are using it to improve your reporting great, but to truly power your organization greater insights move BI down into your operational staff levels and design a rules based decision making system.