REanalyze This!: Modeling

Showing posts with label Modeling. Show all posts

Thursday, April 19, 2012

Top 5 Takeaways from INFORMS 2012 Business Analytics Conference

Rob was a Guest Blogger for the INFORMS Analytics 2012 Conference. This was originally posted on the Conference Blog.

As expected, it was a great conference with valuable presentations and plenty of opportunities to network with professional colleagues. After a day to reflect, here are my top 5 takeaways (not necessarily in order). Hopefully they will be useful to both experienced analytical professionals and executives, and those just starting the journey.

(1) Google Insights for Search
Hal Varian, Google’s Chief Economist, gave a compelling keynote that made great use of this free tool. Check it out if you haven’t done so already. It’s a great way to identify market trends based on search frequency. The data goes all the way back to 2004, and you can download the results to combine with other data sources.

(2) It’s not just about the analytics
Congratulations to TNT Express, winners of the 2012 Edelman Award. My biggest takeaway from their presentation had nothing to do with their network optimization models. Rather, it was how they built an entire optimization “ecosystem” centered on their GO (Global Optimisation) Academy.

“(In addition) to adopting OR tools, TNT Express has teamed up with the Tilburg University/TiasNimbas Business School (Netherlands) to create a two-year management development program in transportation network optimization. Called the GO Academy, the training has been delivered to more than 150 managers from TNT since 2008.”

Lots of companies have advanced analytical capabilities, but TNT have added huge value to the organization – and set a great example for other companies -- through leadership, education, training and change management.

(3) The role of Analytics in Decision-Making
We’re always looking for better ways to describe what we do and how we add value through Analytics, and I thought a few of the speakers had particularly interesting angles.

Thomas Olavson of Google described his team’s role as “replacing intuition with data-driven decisions”.
Colin Kessinger of End-to-End Analytics reinforced the role of leadership. “Math supports the decision-making process, but the math doesn’t make the decisions”. Don’t expect an analytical model to give you “the answer”.
Both Colin and Glenn Wegryn of P&G focused on visualization & interactive decision support. According to Colin, “This is what gets you a seat at the executive table”, and Glenn provided the evidence when he described the weekly “immersive business reviews” for P&G’s executive team (with a great picture of the huge hi-def screens that surround the conference room).
Focus on the “so what” or the “why” – not the “what”. Reports are a dime a dozen. How are you going to help drive better decisions?
A neat quote from Glenn Wegryn: “We bear gifts - in the form of rational business insights”. (In my words, "Geeks bearing gifts" -- OK, that's a really bad pun...)

(4) Simpler is Better
The simplest approach that solves your problem is usually the best one. Several speakers hit on this point in different ways.

There is “no correlation between analytic complexity and business value”, according to Glenn Bailey of Manheim.
Chris Fry of Strategic Management Solutions paraphrased Einstein – A model “should be as simple as possible, but no simpler”.
Before you jump into a big, complex analysis with expensive 3^rd-party software, make sure you really understand the problem you’re trying to solve. Colin Kessinger advises that if you can’t build and understand a prototype in Excel, you don’t really understand the problem and you’re not likely to be successful with a complicated off-the-shelf software solution.
Thomas Olavson echoed the sentiment that “small steps lead to big wins”. He favors analytical building blocks and rapid prototypes, and cautions against looking for “grand unified theories”.

(5) 50 Minutes with the 5 minute analyst
It seems only appropriate to make the 5 Minute Analyst my 5^th highlight. In a very interactive conference session, Harrison Schramm of the Naval Postgraduate School shared some fun and thought-provoking “toy problems”. Be sure to check out the 5 Minute Analyst column in Analytics Magazine. The most recent column is available here.

Let me know your thoughts. It's been fun being one of the Conference Bloggers and I look forward to continuing these interesting discussions. Until next time...

Rob

_________________________________________________________________

Rob Ende is Founder and President of REanalyze Inc., a Supply Chain, Inventory Management and Analytics consultancy based on Long Island, New York. Rob can be reached at 631-807-2339 or rende@reanalyzeinc.com.

Thursday, April 5, 2012

Rob Ende Guest Blogging at INFORMS Business Analytics Conference

I am very excited that I've been selected as a guest blogger for the INFORMS 2012 Conference on Business Analytics & Operations Research.

The conference takes place April 15-17 in Huntington Beach, CA, and is the leading venue where world-class organizations showcase the strategic use of business analytics and O.R.

The conference blog can be found here, and the blog will replace the conference home page once the conference begins. I will also mirror my posts right here on REanalyze This!

In addition to taking in some great presentations on the success of Business Analytics in areas such as Supply Chain and Forecasting, I will be co-facilitating a discussion group on best practices in spreadsheet modeling:

Garbage In, Gospel Out? Join Rob Ende, CEO of REanalyze Inc. and SPRIG President Rick Carter, CEO of Equation, in a discussion of tips and tricks on spreadsheet design, development and auditing to prevent errors, reduce risk and ensure quality results.

Please check back for conference updates starting April 15. Until next time...

Rob

_________________________________________________________________

Tuesday, March 13, 2012

Analytics Magazine - a Must Read for Business Analytics

If you manage a business analytics function or analytically-oriented projects, or if you think your organization would benefit from a more analytically-oriented and data-driven approach to decision-making, I highly recommend that you subscribe to Analytics magazine.

Analytics, a digital magazine published by the Institute for Operations Research and the Management Sciences (INFORMS), provides readers with real-life examples of how data, modeling and mathematical analysis is used to drive better business decisions and provide concrete competitive advantage.

I'm a fan of the "Executive Edge" column. The current issue's column features Kathy Chou, VP of Sales Strategy & Operations at Hewlett-Packard, on how HP achieved great rewards by applying analytics to the Sales function, typically one of the most difficult places to make headway with data-driven analysis and decision-making (How Analytics Turned Sales from Art to Science at Hewlett-Packard).

Another recent column from Chris Fry of Strategic Management Solutions talked about some keys to producing real, sustained value from analytics projects (Closing the Gap Between Analytics and Action), including:

Focus on the business process, not just the model
Make the model transparent to users and stakeholders (no "black box")
Highlight the "so-what's" - the conclusions and actions you want the user to take

It's best to view Analytics in the digital reader version and you can download .pdf versions of your favorite articles.

Happy Reading!

Rob

_________________________________________________________________

Wednesday, January 4, 2012

What’s Lurking in your Spreadsheets? (Part 2)

Note: This series of posts is not intended to be a “How To” guide. My purpose is to highlight some common issues and mistakes in spreadsheet development and how to avoid them and the resulting risk to your business. Even if you don’t develop spreadsheet models yourself, hopefully this will help you know what to look for and questions to ask those who DO develop spreadsheets for your business. If you would like more detailed “how to” instructions and examples on any of these topics, I refer you to a good Excel reference book like the Microsoft Excel 2010 Bible by John Walkenbach.

In a previous post I discussed the prevalence of poor spreadsheet modeling practice in most organizations, and the resulting risk to the business. In that post, I covered one very common mistake – using fixed or hard-coded values or constants in formulas.

This time I would like to cover another common type of spreadsheet error: incorrect cell references or ranges. You are probably familiar with Relative and Fixed references (even if you didn’t know what they were called).

“A1” is a Relative reference referring to Column A, Row 1. If you copy or drag a formula to the right, the A1 becomes B1, C1, etc. If you copy or drag it down, it becomes A2, A3, etc. In other words, the reference is “relative” to whatever cell you’re in.
“$A$1” is a Fixed reference. No matter where you copy or drag it, it remains $A$1 and always refers to the same cell. In other words, the reference is “fixed” no matter what cell you’re in. And of course you can fix just the row (e.g., A$1) or just the column (e.g., $A1).

Each type of reference is appropriate in different circumstances. The problem is that it’s easy to use the wrong type in a formula and not realize it. Then when you copy and paste the formula somewhere else, you may end up referring to incorrect cells and getting incorrect results. Things may appear OK but the calculations may be totally wrong!

It’s critical to review all the formulas in a spreadsheet model to make sure the correct types of references are being used. You can also use Excel’s built-in auditing features, e.g., Trace Precedents, to look for mistakes.

Here’s another neat trick: Press Ctrl and the accent key (it looks like ` and is next to the “1” key on the top left of your keyboard). This turns on the Formula View which shows the formulas in every cell on your worksheet. This is a great way to scan your formulas for errors, especially incorrect references.

Perhaps the best way to prevent incorrect references is to use Range Names. To name a cell or a range of cells, just select the cell(s), click in the Name box which can be found toward the upper left corner of the window, type in a name and hit Enter. Then, whenever you refer to that cell or cells in a formula, you can use the Range Name instead of entering the cell reference. Here’s an example, first with a normal reference to cell A2, and second where cell A2 has been named “Margin”.

a. = SUM(B3:B6)*A2

b. = SUM(Q4_Sales)*Margin

Why is (b) so much better than (a)? Three main reasons.

First, when you’re writing a formula you don’t need to remember that the margin is in cell A2; you just type “Margin”.
Second, you don’t have to worry about using the wrong type of cell reference. If you write a formula using “A2” (a Relative reference) instead of “$A$2” (a Fixed reference) and then you drag or copy it to other cells, you will end up with the wrong values. If you write a formula using “Margin”, you can drag or copy at will with no concerns.
Third, it’s much easier to understand what a formula is doing. In example (b), it’s easy to see that you are multiplying Sales by Margin. Not so in example (a).

In future posts I’ll continue to point out other problems to look for in spreadsheet models, along with some best practices to reduce the risk to your business.

Meanwhile, I’ll close with the same question from last time: Do you run critical parts of your business on spreadsheets? If so, have you thought about where and how they are being developed, and what sorts of errors may be lurking inside?

Until next time…

Rob

_________________________________________________________________

Tuesday, December 13, 2011

Can U B like P&G?

I read a great article in the McKinsey Quarterly today entitled Inside P&G’s Digital Revolution, about how Procter and Gamble is using digital technology and advanced analytics to revolutionize its business.

I encourage you to read the entire article (note: you may have to register), but in case you’re pressed for time, the first paragraph sums things up nicely:

“Robert McDonald is a CEO on a mission: to make Procter & Gamble the most technologically enabled business in the world… [He is] overseeing the large-scale application of digital technology and advanced analytics across every aspect of P&G’s operations and activities—from the way the consumer goods giant creates molecules in its R&D labs to how it maintains relationships with retailers, manufactures products, builds brands, and interacts with customers. The prize: better innovation, higher productivity, lower costs, and the promise of faster growth.”

A few of P&G’s keys to success:

Enable better decision-making through real time information (or as close to real time as possible)
Apply advanced modeling & simulation tools to test new product concepts and new manufacturing & supply chain strategies in the virtual world before spending money in the real world.
Hire people with great analytical skills, then continually train them & develop those skills at all levels of the organization

I’ll add a couple of thoughts of my own. First, it starts at the top. P&G’s CEO, Robert McDonald, is not just bought into the strategy, he understands the details and is an active champion for using technology and advanced analytics to drive his business.

Second, the competition is moving fast so there’s no time to waste. Get started now. Don’t wait for perfect data – use what you have today, BUT… do your best to build in data integrity from the beginning so poor data integrity doesn’t come back to bite you later.

Few companies have the resources and capabilities to be like P&G. But almost any company can use P&G as an aspirational ideal – an example to shoot for as you develop your own analytics strategy.

Ask yourself a few questions about your business:

How are you using advanced analytics to drive your business?
Do you know what parts of your business offer the greatest opportunity?
What information do you have today that you can leverage for greater insights and improved business results?
What information that you don’t have today would have the most value? What’s the next step to getting that data (or at least getting closer to it)?

Until next time…

Rob

_________________________________________________________________

Tuesday, December 6, 2011

What’s Lurking in your Spreadsheets?

What’s wrong with this formula from a spreadsheet model?

=A2*.18

Seems simple enough. Imagine you’re calculating an annual inventory holding cost. Cell A2 contains the inventory value in dollars, and you estimate the annual holding cost percentage at 18%. Multiply them together and you’ve got the correct answer, right?

Technically, yes. But…

What if the holding cost percentage changes next year? Will you or an analyst or whoever uses this spreadsheet remember to change the formula?

What if different items have different holding costs? For example, obsolete items might require a higher value due to the risk of write-offs.

What if someone passes this spreadsheet along to a colleague working on a different product line, where the holding cost percentage should be 25% instead of 18%?

Any time a parameter like this is fixed or “hard-coded” in a formula, rather than in an input cell, you are much more likely to get burned in one of these situations. If you are lucky, this is only an informational report. But what if it’s in a financial model that you use to set operating budgets? Or to calculate someone’s bonus?

In this case, the solution is pretty simple. Just put the holding cost percentage into an input cell (or maybe a range of input cells for different products). Highlight the cell with shading and borders for good measure. That way anyone who uses the model can easily see the value and change it if necessary. Even better, give the input cell a name like “HoldingCost” and then you can put it into a formula anywhere in your spreadsheet without having to remember what cell it’s in.

This may seem ridiculously easy, and it is. But you would be amazed at how prevalent poor practices like these are in otherwise very sophisticated companies. Virtually every business uses Excel spreadsheets to a lesser or greater extent. And no matter where I’ve been in my career as an analyst, an executive and a consultant – small and large companies, different functional areas, various industries – I have seen poor modeling practices and outright mistakes in almost every spreadsheet I’ve come across. In fact, some people have remarked that I have a special talent for finding this stuff! Some are obvious formula errors. Others, a result that “just doesn’t look right” based on what I know about the business (note: the published research on spreadsheet errors is a little scary: See this literature review from Powell, Baker & Lawson or this research article from Panko).

When you get right down to it, spreadsheets are software that perform some function or process for your business. But unlike other types of software, which are usually developed by trained software developers with a controlled process, just about anyone can create or modify a spreadsheet. Think about all the rigor and quality control that go into typical software development in your company. Then think about whether your spreadsheets are getting a comparable level of attention. If not, you could be introducing a lot of uncertainty and risk into your business.

In upcoming posts I’ll point out some other problems to look for in your spreadsheet models, along with some best practices to reduce the risk to your business. You may consider starting with a "No fixed values" or "No hard-coded constants" rule for your spreadsheet developers.

In the meantime, here’s something to think about.

Do you run critical parts of your business on spreadsheets? If so, have you thought about where and how they are being developed, and what sorts of errors may be lurking inside?

Until next time…

Rob

_________________________________________________________________

Wednesday, July 6, 2011

Highlights from Llamasoft Conference on Supply Chain Design

Last week I attended Llamasoft’s first annual Summer Conference on Supply Chain Design in Ann Arbor, Michigan, along with a couple of days training on their Supply Chain Guru applications. I met some great people, gained a lot of insight about Llamasoft and their product roadmap, and enjoyed some spectacular early-summer weather in Ann Arbor.

Llamasoft is growing fast, and they are now the only major independent Supply Chain Design software vendor, since the acquisition of LogicTools/LogicNet by ILOG and then IBM, and Optiant's acquisition by Logility.

Founder, President and CEO Don Hicks talked about (OK, I’ll say it…) changing the paradigm for Supply Chain software. Instead of the traditional “Planning versus Execution” framework, Don talks about “Design versus Planning versus Execution”. Planning applications will become increasingly automated and IT driven, cutting across the usual distinctions between strategic, tactical and operational planning. Meanwhile, Llamasoft is squarely targeting the Design space, which will continue to require human input and creativity. To that point, late this year they will add Transportation Optimization to their existing capabilities in Network Optimization & Simulation and Inventory Optimization.

The conference featured some interesting keynote speakers:

Dr. John Gattorna of Australia highlighted his Dynamic Alignment Model, emphasizing the need to develop multiple Supply Chains – with distinct strategies -- that align with different customer behavior models. His book Dynamic Supply Chains is available on Amazon.
Astronaut Buzz Aldrin shared some great stories, images and video from the Apollo program and the 1969 moon landing -- and his strong opinions on how we can and must regain the initiative in space exploration.
Developmental Cognitive Neuroscientist Dr. Bruce Hood talked about how the decision-biases that are hard-wired into our brains can lead people to cling to incorrect beliefs and make decisions counter to all available evidence – something to keep in mind the next time someone questions the “obvious” recommendations from your analysis!

A question for my readers: What are the pros & cons of building your Supply Chain Design capabilities around software, services and solutions from:

…a small but fast growing, independent software company focused exclusively on SC Design (Llamasoft)?
…an established Supply Chain Planning software company integrating an acquired SC Design solution (Logility/Optiant)?
…a huge global conglomerate integrating a SC Design Solution (itself part of a larger acquisition) with extensive software and consulting resources as part of an aggressive Supply Chain Analytics strategy (IBM / ILOG / LogicNet)?

What do you think? I'd love to hear what you have to say!

OK, that’s all for now. Please visit our web site at www.reanalyzeinc.com. Until next time…

Rob

_________________________________________________________________

Rob Ende is Founder and President of REanalyze Inc., a Supply Chain, Inventory Management and Analytics consultancy based on Long Island, New York. Rob can be reached at 631-807-2339 or rende@reanalyzeinc.com.