I have started working since last 6 months as par of a new Analytics division in a large firm. And the air is quite exciting around Analytics as the next big thing. However, what I learnt is that not many people really understand “analytics” even if they are extremely enthusiastic about it. Some researchers have tried to find out the real benefits achieved by organizations by investing into analytics based capabilities and overall just end up confirming the notion that your organization will become more profitable or competitive in some way or the other but other than that it doesn’t really help me or any organization in understanding what it really means to them or is it of any use? Is it worth investing their time and money?
So I decided to look it up by myself. And started understanding the fundamental ideas about information and data. What I never knew was that there was something called as “Information Sciences” i.e. some people actually study information as a subject. It was really intriguing for me. Through some research I found out that Information is just one piece in the game, there is a model called as DIKW Pyramid (http://en.wikipedia.org/wiki/DIKW_Pyramid). DIKW pyramid is Data > Information> Knowledge> Wisdom.
So to briefly explain, information is an intermediate stage and it actually starts from Data, which then is then converted into meaningful information, which increases your knowledge about your business and then eventually with the decisions you take and the learnings your derive out of it eventually bring your wisdom. (Well that was my interpretation of the DIKW model). But what the heck, how can I use this model to find out what I lack? I mean is my problem a lack of data, information, knowledge or wisdom? Can you tell your employees that wisdom is your area of improvement? If yes, how can you work on it? I am not satisfied with that model although it is good from a concept point of view.
The Practical Information Cycle
So, I decided to chalk out my model based on my learning and experiences so far. Here it goes.
It starts with…
The big cloud like shape in the diagram. Your environment will consist of your people, processes, your businesses, products, services, functions like HR, Infra/facilities, Finance, Travel, etc and your Systems like ERP, CRM, or any other platforms your use to manage your organization. Although in the diagram I have mentioned whatever came to my mind first, but it won’t be restricted to this. If you see, I have primarily mentioned the internal environment. But you will also immediately recognize external environment factors like gas prices, stock market data, weather, or govt. policies and stuff. These environmental factors will directly or indirectly affect your business.
2. The Data Block:
This is the place where you store all the raw data which you get out of your internal and external environment. It could be the logs of transactions, or templates filled out by your cost centre managers during your annual plans, currency fluctuations over the last two years, the dump from your ERP system or whatever you can think of. And the arrow that connects your Data block and the environment block is the “Sense/Fetch/Retrieve/Monitor” activities. This is where I believe you should first develop your enterprise capability. Most leaders want to add analytics capability to their organizations, but they have no clue what data they need to get it from. Even if they do know the data they need, they are unwilling to invest in tools or systems that will enable fetching and storing that data. In my opinion, without strengthening your sensing and storing capabilities, no matter what analysis you conduct on your inadequate data, you will end up having partial insights. There are many areas where you can deploy various types of tool-sets to fetch your data. If you are an industry driven enterprise, lot of sensor based platforms have started emerging that will help you fetch data about your machines, power consumed, the speed of your turbines, etc. If you are a web-based enterprise, its needless to say there are many analytics platforms available today – Google Analytics being close to a one-stop solution for being more intelligent about your web environment. If you are a just any SME firm, then improve your CRM, ERP, SCM systems and upgrade them with better tools and products, that go beyond the rigid framework of the existing ERP models or build your own simple tools to fetch the correct data periodically. When I is say correct data it includes clean data, that well organized and well described. One of the greatest challenges today in good analytics is having correct data. You can make use of good ETL tools in order to fetch and clean data.
The data block also deals with where you will store all your meaningful data? how will you keep it secure and for how long? So, all sorts of technologies are available today, with traditional options like Enterprise Data Warehouses to newcomers like Cloud/Big Data platforms. Depending on your requirements there are host of storage options available in the market today. But with new technologies come new challenges and a push for transformation in your organization. That said, it is very difficult to manage change in your organization as most people are happy working with existing technologies and will protest any new changes. Also, there is certain amount of risk involved on both the effectiveness, efficiency and security of the new data sensing and storing technology which needs to be considered.
3. The Information Block:
Before I jump to the information block, I want to discuss about the activities mentioned on the arrow connecting the Data and Information block. For those who don’t know the difference between Data and Information, here it goes – Data is just Data you can’t make sense out of it. Information is what can truly inform you about something. Like your sales are growing at 10% rate or your organization is making 5% fewer profits than last year or acquiring 10 new customers ever hour. That’s information! Data is just some dump or a set of thousand records which look absolutely rubbish and overwhelming when you see it at first. So, Analytics plays a major role here. You take your data and give it to someone who loves to play with that data and finds out what are the top performing products, which are the next areas of improvements or opportunities for your company. Analytics enables companies to back their information with evidence. Also, some of the organizations employ sophisticated techniques like Statistics or Algorithms and stuff most of us don’t know to draw more and more meaningful information out of your data. There are various techniques to do so and it is a branch in itself. Also, various tools and platforms are available for the same. But I believe one needs to have someone very curious with a ‘what-the-hell-is-going-on-in-here’ kind of attitude to help you find good insights with your data. Also, the person should have at least some sense of business or the area he already wants to try to focus on. Get a statistician with no business sense to do your analytic stuff and it will only lead to statistical jargon. But get a young professional who has good business sense, decent background but with an attitude of curiosity and discovery and assign a good statistician or a modeler with him/her and maybe you are going to find some very good insights. And if you do find someone who knows business, stats, algorithms and has curiosity, then nurture him/her and empower him/her.
But too much of information is no information. I mean, if you have a 100 good reports generated everyday, you can absolutely make no sense out of them. Who has the time these days anyway? You want instant gratification don’t you? So find some Dashboard developer who will put it all together neatly on a single page with graphs and trends and meters and all sorts of charts and tables you want. Please sit down with that person, and help that person prototype and refine your dashboard before you start developing on some BI tool like Qlikview, Tableau, Domo, SSRS, Cognos, etc. Even minute changes in the dashboard are onerous and time-consuming. So prototype it on some piece of paper or power point. Plus more ideas will flow once you do that. This is what I meant by Visualization part of the activities.
Information can be represented in many formats: PowerPoint presentations, word reports, excel files or BI dashboards. And since, timing plays a major role in information, make sure you don’t ask your employees to spend considerable time in formatting or jazzing up your reports. Templatize the reports, productize them and improve them on an ongoing basis. The sooner it reaches you, the better and faster decisions you will be able to take.
4. The Action Block
If you see the last line of the previous paragraph, I mentioned about faster decisions. No matter what information you have, if you don’t plan to act on it, or take decisions that will matter, it will only add more knowledge to you. With this knowledge, you will at some point take better decisions or else it will culminate into wisdom (Remember the DIKW model we discussed). So, no matter what, all information will eventually culminate into decisions, actions or some form of control mechanism. This control mechanism or your actions are going to be redirected to your environment (either internal or external) depending on your situation. As a leader, I am sure you would want to assign your decisions to some of your employees to ensure accountability. So accountability really matters along-with good planning. That is why, with good information you plan better and manage accountability better. Accountability does not mean only a single person responsible for the improvement of your environment, it could be an automated system as well. These days, there is a surge in prescriptive analytics where best sets of solutions are chosen automatically through algorithms and executed through various systems. However, such systems are still in development. In any case, having someone on top of your actions and decisions you take is important. That’s what accountability is all about. And it goes without saying that without good planning one can’t take good decisions. So, these two things are critical in your information cycle once you get good info on your hands.
Actions can be broken down at three levels: short-term, medium term, long-term. Also, their level of impact and the number of benefits or downside they bring along. Actually, if you are bad at remembering your actions, you can also have an “action tracker” dashboard which can help you keep track of all actions or decisions.
These actions and decisions eventually get routed to your environment. It is kind of feedback loop. So Operations, Tactical and Strategic plans are a way of doing that, some companies use balanced scorecards for their managers, Some use workflows. Other ways of taking sound decision are through budgeting and forecasting methods or other periodic reviews mechanisms.
Assuming your actions will improve the internal and external environment of your organization, I believe you have completed your information cycle – Environment > Data > Information > Action > Environment and so on…
Now some important questions?
1. Do you really need more analytic capability? or more data fetching mechanisms? or more data storage mechanism?
2. Do you already have enough insights? But are unsure because you do not have data based evidence?
3. Do you have bandwidth to act on the new insights being thrown by your system? After-all you must already be lined up with so many action plans (both strategic and operational)? Or do you just need good data driven information to validate your existing actions plans?
4. Do you really know why you need the analytic capabilities or are you going along with the hype created by tech-evangelists and investing in something which will cost a lot but marginal returns may not be so high?
That was my take on the practical information cycle. Feel free to express yours!
Lot of people do not have any idea about the hidden capabilities of Excel.
Especially when it comes to reducing repetitive tasks.
Also, even if they have heard about it, they have little clue as how then can write those instructions which can do all the manual repetitive work for .
I say, it is not so difficult. Here is a tutorial to begin with for all of those who want to learn a thing or two about being more productive at work through the use of Excel macros.
It has an example of how to create a particular pattern and executing it.
The link for the file is here: https://app.box.com/s/kq4pbegn6od2arty3eyj
and the explanation is given below:
Whether you are a program manager, project manager, finance exec, HR person, sales, or even a techie, you cannot escape from excel and this tutorial will really help you avoid the repeat tasks saving you considerable amount of time. Also, by recording macros you can learn how excel works step by step and if you grasp those instructions you can make excel do anything for you. For example, periodically collecting data from multiple sources and then converting it into meaningful information in a matter of minutes. Or linking multiple macros with each other and automating your excel processes.
If you are a sucker for retro-games like Tetris, Air Fighting, Bowling, Chopper Challenge etc, then hold your breath; I have added a complete games folder in my file Box.
Here is the link:
These games are timeless, no PS3, XBOX, Wii or iOS, Android games can replace the fun.
Give it a shot and get hooked.
This is one heck of a folder to bring some fun to your office life ;)
PS: These games are not created by me.
Excel is probably the best software product ever created.
It has the ability to organize randomness into neat structures, build relationships between values and turn your abstract concepts into tangible presentable files.
Having said this, it is very essential for every person whether he/she is a coder or a carpenter or a manager or housewife, to learn to use Excel; because it enables a person to develop slick organizing skills.
So let me provide you with some of the most essential shortcuts to excel. These shortcuts should be practiced again and again until they become part of your body and muscle memory. Such that you start using them even without thinking of them. So, go ahead and take a printout and put it up your desk because they will be very very useful in your daily life, no matter what kind of profession you are part of. So here it goes:
1. Add a row
Alt + I + R
Not necessary to simultaneously press all 3 keys
2. Delete a row
Select row (Shift + Space”); Delete selection (Ctrl + -).
3. Add a column
Alt + I + C. For rapid column addition
4. Delete column
Select column (Ctrl + Space); Delete selection (Ctrl + -). This will enable rapid column deletion
5. Merge or Unmerge cells
Alt + H + M + C
6. See formula in cell:
Press F2. This helps to check the formula in the cell. Also for edit text in cell.
7. Hyper jump to edge value
Ctrl + any arrow key. This enables very fast movement across edge of data sets.
8. Enable/Disable filters on a table
Ctrl + Shift + L
9. Show Dropdown items for filters
Alt + down arrow
10. Select/Unselect items in drop-down filter sets
That’s about it.
Remember MS Excel is going to be used for a long time to come. By learning, just the above 10 skills, you will save a lot of time for yourself, so that you can leave home early or enjoy with your friends.
Please let me know in case of any more shortcuts, so that I can update the list accordingly.
Hans Rosling made me fall into love with 3D bubble plots. They are brilliant, powerful and wonderful to understand. Unfortunately, MS Excel has not tapped into their potential much. Infact, MS Excel does not see the scope of Infographics, Information Design, Advanced Charts/Tables at all in my opinion. As a result in my opinion a huge huge huge chunk of business lacks brilliant wonderful insights today. Ofcourse, MS Excel has Pivot tables and you can use them as you want, but the power of visualizing the hidden meaning in your data is missing from Pivot tables and also other charting tools.
the 3D bubble plots in excel do not even allow me to lable my bubbles well. It gets so frustrating. Can you believe that you have to use a VBA program for doing something so simple? Or copy paste your labels everytime onto your data.
The beauty of a 3D plot lies in its time based evolution (that will make it 4D then). That’s what Hans Rosling does, you feel overwhelmed because he can show you 4-dimensional data (the X-axis, the Y-axis, the Size of Bubble, and a time frame) beautifully. Not only that he helps to draw a huge number of relations between different dimensions across different items very easily. Using 3D bubble plots + a timebased dimensions, I can theoretically draw insightful relations across all dimensions to the count of over more than 50,000 (if I present 4-D data of 3 items in a time frame of 4 periods)….
Below is gapminder’s home page pic (whic puts in a nutshell whatever I am saying):
Now if you use a pivot table, I don’t think it is going to be possible to work out all this in a flick of few moments. But your eye can show you all the significant deviations out of these many relationships quite easily. This power would have been with everyone, alas, only if Microsoft believed the same…Nevermind Hans Rosling does it and we can ask him to share his tool for free for then whole world to experience the joys of 4D data or hope that someone develops it for us….
Moving on, this style of bubble plotting can be useful in financial securities to visualize the profitability or returns of securities over a period of time across time, across EPS of the security, across, P/E of the security…or whatever you can think of…..you can even categorize your Bubbles across type of category they fall into (like a bond or a stock) by just coloring the bubbles differentially. (Oh well, then makes it a 5D plot, no wonder they are more insightfully beautiful)
So, as an experiment, I applied it to the Strategic Business Units of a firm, and wonder what? I think I have almost struck upon a good model to measure the relative profitabilities between SBUs:
So in my 4D plot, I have Profit Margin on one axis, Net Income on another, Size of bubble would be the revenue of the SBU (time is the 4th dimension which I haven’t shown)…..
Here is the bubble plot I came up with:
Each bubble represents an SBU.
So according to their revenue their size is being shown and hence if each SBU acquires more revenue over time, it will start inflating. Now as it starts starts growing revenue, there will also be an increase in the top level profitability which is the Gross Margin (or Project Margin if you are a project driven firm) and also the bottom level profitability, which is the Net Income. So, with time a bubble will grow in size and also move vertically as well as horizontally. (Please note above bubbles are different SBUs for a single instance of time and not across different time frames for a single SBU, so please don’t get confused). So if you can see, there is a black line cutting across all SBUs. This path is indicative of the average path of movement expected from an SBU (in that time instance). Any SBU above this path indicates that it is performing well on the project margins compared to the net income and any SBU below the path indicates that it is comparing better on Net Income that on Project Margin.
So this got me into thinking, that yes, this model captures the expenses of an SBU pretty well. So, I plotted the lines for Direct Expense, Indirect Expense and Total Expense, as shown in Red color. Each line corresponds to that expense by way of calculating the difference of values between the starting and ending points of the lines. (Please note that length of line does not represent the value of corresponding expense, it just helps to calculate the expense):
-“Total expense” line calculates the difference between the beginning point: “Revenue” and ending point “Net Income”
-“Direct expense” line calculates the difference between the beginning point “Revenue” and ending point “Project Margin”
-“Indirect expense” line calculate the difference between the beginning point “Project Margin” and ending point “Net Income”
Which means the above model very well captures a P&L (across SBUs, and their broad expenses, profit margins, and net income)..Woohoo!! :)
At the same time it shows relative performance.
Now say a bubble is above the line, then that means, the particular SBU is good on profit margin and can move horizontally by reducing its indirect expense. The difference between the end state and current state can provide understanding of by how much it should reduce its indirect expense. Similary the other way round if the bubble is below the line, it needs to improve on project margin.
So this model can capture SBU performances and help in SBU performance benchmarking too. That too across time. And hence, the model can be quite strategically useful and can be used for long period budgets too. Also, it can track month on month performance of SBUs in the short term.
Which means it is a very useful model!!
So go ahead, use it, test it and if you feel its very useful for your business strategy, do let me know or if you have any ideas to improve this model, your comments are always invited.
Well today, I was thinking I will clarify my obsession for dashboards. I know what you must be thinking “Who the hell is so passionate about a dashboard??”. Well, I must say I am not really passionate about dashboards, I am passionate about “Information Designing”. Its the great-grandparent of dashboard designing. Dashboard is just a tool to show information. Oh by the way for those who do not understand the difference between data and information, the answer is that information “informs” and data is just data, it may or may not inform. So “information designing” is putting together data in such a way that it informs you. That’s what even a dashboard does. And in this age of the hyped “information explosion” (which is a wrong term entirely; Yeah because information can never explode. Data can! Information cannot!), there is this strong need for good information designing.
Well, I must admit that I stole the above lines from Mr. Richard Saul Wurman.(Ref: http://dsc.discovery.com/videos/curiosity-richard-saul-wurman-information-anxiety.html#mkcpgn=snag1) He is a pioneer in information designing. He is the founder of TED and has written several books on information designing and has done a lot of creative and interesting stuff . One of his current works is the http://www.192021.org/. Check it out for yourself its pretty interesting! I am sure you will love it.
Another pioneer who passionately loves information and can bring upon the ‘O-holy-mother-of-cow’ reactions from any audience, is Mr. Hans Rosling (Ref: http://www.ted.com/speakers/hans_rosling.html). He has done a lot of videos on TED and they are all listed on the rightside of his profile page. Please go and check it out. You will absolutely love it! He has also created an equally interesting site : http://www.gapminder.org/ . Again its a great site that will knock you off your chair!
Then there is Mr. David McCandless, who created the helicopter game and is also one of the leading information desginer (ref: http://www.informationisbeautiful.net/about/). So all you people, scared of too much data being bombarded on you without really informing you correctly, relax, coz this new branch of science or art whatever you may assume, has picked up quite well since 2009 (Although it was established pretty early on in the 80s itself)and will hopefully come to save you soon.
So, keep this in your mind that Information design is the mother of Dashboards. And that Information design is somewhere between creativity and engineering, so now you know what I am passionate about. Its not the dashboard, but this creatvity and engineering of data to convert it to meaningful information is what I love.
Remember dashboard is just a tool to inform neatly , cogently and correctly (just like a presentation or a video is) . Its not just those sexy speedometers that matter. Its the meaning they throw which counts the most. It has to inform! It has to bring good insights. It has to make a mark on those who read it or else its pretty much useless!
Yesterday I was thinking of how I can squeeze all my time based data into a single graph.
And so I came up with this idea of having a control on your chart to move your time-frame in the same graph itself.
I won’t explain much out here. Instead, I am uploading the item in the adjacent document Box.
or download it from this link:
Download it, play around and let me know if you like it.