more-arw search

How Finance Can Become An Analytics Powerhouse

The Finance function is undergoing a rapid change like everything else in business. In order to maintain its relevance, not only as a stand-alone function, but also as a part of the business it’s time for Finance to power up its analytical capabilities and take over the analysis from the frontline.

I have in previous articles described quite a bit about current trends in Finance not least when writing about the finance transformation nine-box. Those trends are supported by several surveys but for more insight I also encourage you to read “Being The Best: Inside the intelligent finance function” published by KPMG in 2013. Last week we discussed the future of the accounting profession in “Why Accountants Are A Threatened Species” which turned out to be a hot topic and quickly developed quite a heated discussion. This week we will discuss the middle layer of the nine-box and look at analytics.

First we must have one set of numbers

The classic issue for Finance has always been whether or not the numbers that came out of their analysis could be trusted. There is not always just one solution, but flat out making mistakes makes it quite embarrassing to present something to business leaders. The reason for the many mistakes has often been that the analysis has been quite manual rather than an automatic output from a system. Whenever something is manual and requires human intervention it increases the error ratio. Therefore, Finance has invested heavily over the past years in upgrading systems and automating data output to eliminate manual intervention. At the same time many systems have been consolidated into one ERP system leading to only one source for the numbers. Finance can, therefore, in general claim that they now have one set of numbers which are mostly free of errors when presenting to the business.

How Finance should leverage the newly gained confidence

Finance now needs to leverage its newly gained confidence to start consolidating all analytics in the company within the function. The case for this change is fairly clear as it can be free up resources in the frontline to work more with the customers than producing internal analytics. Basically, Finance should create packages for sales people to meet with their customers given them as much knowledge as needed to close the best deal for the company. There is no need to leave anything to the sales person’s imagination as they have full information about the customer and the potential outcomes of any deal they make. How good a deal they will make of course still depends on the individual sales person, however (s)he can’t claim afterwards that they were unaware of the consequences of their choices. Similarly, all management information whether it’s classic financials, operational information or sales statistics should be produced in Finance’s new analytical powerhouse. This will emphasize the “one set of numbers” approach as well as minimize confusion around how a company is actually doing.

How to avoid the politics when Finance gets more power?

In general I will spend much more time discussing this point next week when we look more into the last part of the nine-box regarding business partnering, however, here it should be recognized that making such a shift is not without challenges. The frontline could end up feeling concerned about how their performance is portrayed to management when they are no longer in control of the analysis of the presentation of results themselves. However, this is where the good finance business partner comes into play. (S)he understands that Finance is a support function and that its main purpose is to work well together with the frontline, understand their needs as well help them deliver results. Finance can never be successful if the business is not successful!

Let me know what you think about Finance as the analytical powerhouse of the business and what it takes to get there?

Comments

Dan French
Title: Ceo
Company: Consider Solutions
(Ceo, Consider Solutions) |

Interesting post Anders.

There has been a lot of discussion recently as to whether Finance is losing out on the analytics stage to a new group of 'data scientists' elsewhere in the business partly aggravated by the effect of ongoing cost reductions in the function. However, my observation is that the finance function clearly has a great opportunity to be analytical powerhouse, certainly for the core finance processes. As the analytical technologies become less 'programmatic' and more sophisticated, finance managers will rightly be able to focus more analytics attention on the insights derived, the narrative they support and the actions required, which is where the value is, whilst the mechanics will consume less effort.

Gary Cokins
Title: Founder
Company: Analytics-Based Performance Management L..
(Founder, Analytics-Based Performance Management LLC) |

Anders and Dan ... This is an important topic. My concern where I am less optimistic than Dan is that the finance function is already very far behind the adoption rate of analytics by other line management functions, especially marketing and supply chain. Part of the explanation may be that those other functions have much more "big data" to work with (e.g., point of sale customer transactions and their profiles). But another explanation may be that accountants are less comfortable with statistical techniques such as regression, correlation, clustering, and segmentation analysis.

Gary ... Gary Cokins

Anders Liu-Lindberg
Title: Regional Finance Business Partner
Company: Maersk Line Northern Europe
LinkedIn Profile
(Regional Finance Business Partner, Maersk Line Northern Europe) |

Dan, Gary,

Thank you both for your comments. The advanced analytics teams or date scientists as you call them should naturally be a part of the finance organization and nowhere else. The worst that can happen is that each department creates their own unit. This in my opinion will be a waste of resources.

Greta Roberts
Title: CEO
Company: Talent Analytics, Corp.
(CEO, Talent Analytics, Corp.) |

Anders - great article and I appreciate the comments. However, I have to disagree that functional units of data scientists would be a waste of resources. Just like there are Business Partners for various functions (finance, sales etc), data sets are nuanced and it's important to have dedicated resources to handle data coming from each business unit. Sales, workforce, ops, supply chain etc - they are all a little different and require data scientists to spend time with the data.

Whether this is done within finance and there's a team dedicated to each of sales, HR, ops etc - or, if they're embedded within those areas, either way, those resources will be extremely valuable to the overall success of functional analytics projects.

Great article - and I hope the discussion continues!

Anders Liu-Lindberg
Title: Regional Finance Business Partner
Company: Maersk Line Northern Europe
LinkedIn Profile
(Regional Finance Business Partner, Maersk Line Northern Europe) |

Thank you for your comments Mike,
What I meant was that there needs to be a coordination of the application of these data scientists and that coordination should reside in Finance. Whether they report directly to Finance or are imbedded in the functional units and have a dotted line to Finance doesn't matter too much.

Topic Expert
Alan Hart
Title: Consultant
Company: Pacific Shine Group
(Consultant, Pacific Shine Group) |

I was always of the opinion that finance should own analytics and report to the CFO. There is no reason to have this function performed elsewhere in the organization when technology solutions can be set up to closely tie the ERP data output to the Corporate Performance Management or the Planning, Budgeting and Analysis software solution. With a new generation of this type of software, which is now under full automation control, finance can have access to business intelligence originating in the ERP, CRM and CPM and provide management with insight into the future financial health of the organization, including complete and accurate forecast of the company's balance sheet and other implied financial metrics such as forecasted financial ratios, likelihood of complying with loan covenants, etc.

Finance should continue to own the budget process, which also has evolved into a much more "smart" process with drivers and system built in business rules and complete automation over the calculation process. This generally eliminates most errors and promotes consistency and repeatability of results. Finance has now the tools to perform the needed analytics and have results that are accurate and complete, to the extent of the accuracy of the actual and historical data, and the data obtained from the budgeting system.

That said, finance should become an analytics powerhouse under the CFO's direction (isn't that why CFO stands for "Chief Future Officer"?).

Anonymous
(Independent Consultant) |

Finance should provide all financial analytics. Finance should provide the one version of the truth (inputs) for all financial elements (to the entire company) This is a significant amount of data and other parts of the organization should not be allowed to walk in with their own financial numbers ("we adjusted for this, we took that out", etc).

But from my experience, analytics includes many data elements that are not captured by the finance or accounting function (literally thousands of discrete pieces of data that can be impactful to the organization), and other functions add value by knowing their business best and taking their data and comparing it to known financial data points. Sales, HR, logistics, R&D all have "data" that they use to make informed decision and predict the future. Finance shouldn't be a shared service model for business analytics as it will just slow down the speed at which this information needs to be generated. I agree with the previous post that the value is not in the mechanics but in the art. Finance folks are good at mechanics, but won't ever be perceived as being as good at "sales analytics", "R&D analytics", "HR analytics" as sales people, R&D people or HR people. Ask yourself if you would want them closing the books, and then maybe you can appreciate how they might feel about you taking over a key strategic aspect of their job.

Also, the responsibility for analytics should lie with the group that depends on them (responsibility should always be aligned with authority and impact). If the sales organization messes up the sales analytics, then the sales VP will have to explain it. Having an intermediary won't add value.

Anders Liu-Lindberg
Title: Regional Finance Business Partner
Company: Maersk Line Northern Europe
LinkedIn Profile
(Regional Finance Business Partner, Maersk Line Northern Europe) |

Thank you for your comments,
I don't really agree to the main point though. Finance should partner with the business and gain business knowledge that allows them to perform the right kind of analysis. This analysis is made to challenge the functional units, but naturally also to support them. Finance should provide an unbiased opinion on the state of affairs within all frontline functions.

That said, Finance is not there at this point. While the business partner concept has been known for years few companies have managed to adapt it well. Until the business trusts that Finance understands what is going on the analytics powerhouse will not be effective.

Anonymous
(Independent Consultant) |

This is an interesting topic and needs discussion and leadership at most organizations. Thanks for raising. I agree that most organizations struggle with the new reality of big data. Here is an interesting article I read a couple of years ago.

https://hbr.org/2012/10/data-scientist-the-sexiest-job-of-the-21st-century/

Roger Fried
Title: Senior Data Scientist
Company: Teradata Aster Strategy
(Senior Data Scientist, Teradata Aster Strategy ) |

Anders,

I'm a Finance professional and an Analytics professional, but those two things are very different. Bringing Analytics under Finance may work in some situations. I know of a number of cases where Finance contains its own BI group or other where financial systems are under Finance. But, I would suggest that few Finance executives are really prepared to take an active management of their Analytics team.

Before I took the leap into Analytics I would never have imagined this, but at least three departments, IT, Marketing, and Operations are also be able to make valid arguments to pull Analytics under them. And that is before the separate argument that companies should establish a Chief Analytics Officer.

I don't believe that right now Analytics really has a natural home. The biggest problem with Analytics is not is ability to deliver valuable results, but rather that current organizations are typically not organized in such a manner to effectively understand or operationalize the flood of results that are being produced. In other words, the bottleneck keeping Analytics from achieving its potential ROI is the tendency of many organizations to treat data scientists as if they were accountants or statisticians producing some kind of truth that then slowly gets digested by the organization.

Anders Liu-Lindberg
Title: Regional Finance Business Partner
Company: Maersk Line Northern Europe
LinkedIn Profile
(Regional Finance Business Partner, Maersk Line Northern Europe) |

Thank you for the comments Roger,

Naturally other departments can make their claim on the analytics thrown, however, for Marketing and Operations they will always be limited to their own area of scope. Instead of doing analysis, they should be focused on running the business. As stated in one of my previous comments I agree that few finance departments and executives are ready to take on the challenge today. This shouldn't stop the overall vision though and while companies might hire CAO (in this case Chief Analytics Officer) which I have also written about previously then there is hardly room for more C-suite compensation packages in companies. Therefore, the best guess is that a CAO would still report to the CFO.

When it comes to IT I see them as enablers for doing the analysis, not the drivers. Of course, it depends on the industry but in most traditional industries IT is even further away from understanding the business well than finance. So I don't think it would work well, having IT running the analytics.

Roger Fried
Title: Senior Data Scientist
Company: Teradata Aster Strategy
(Senior Data Scientist, Teradata Aster Strategy ) |

I find it humorous to see that this article combines a recommendation to have a single set of numbers with a recommendation to bring Analytics under Finance. These two recommendations are actually mutually incompatible.

I never cease to be amazed at the torrent of data and different versions of the truth that Analytics can generate. Data Scientists create numbers from text, images from web clicks, numbers from network graphs, and will iterate through literally hundreds of analyses using methods taken from radically different fields before arriving at a simple classification of data.

What matters for Analytics is not a single set of numbers, but instead having reproducible methodologically valid code that can operate within the constraints of the computing environment. That code can be justified via the language of statistics, Bayesian statistics, linear algebra, multivariate analysis, information theory, graph theory, or computer science. The code itself may be written in any number of languages starting with SAS, R, Python, SQL, SQL-MR, Java, Scala, Hive, C, or C++. Different methods are set into competition with each other with the method having the best results being declared the winning in a given circumstance.

This delivers a great deal of value, but it doesn't pass the "one set of numbers" test.

Anders Liu-Lindberg
Title: Regional Finance Business Partner
Company: Maersk Line Northern Europe
LinkedIn Profile
(Regional Finance Business Partner, Maersk Line Northern Europe) |

Thank you again for your further comments Roger,

When referring to one set of numbers this is, of course, a general term. The way I see it multiple scenarios should be analysed and compared. There might even be competing solutions when presenting to management, however, management would also have a clear expectation that one recommendation is put forward.

Further, when referring to one set of numbers, it also relates to the company financials. The revenue is what the revenue is and same goes for costs. There cannot be competing versions of that truth.

David Buley
Title: Chief Financial Officer
Company: Association of Independent Schools of NS..
(Chief Financial Officer, Association of Independent Schools of NSW) |

As expected, the topic has generated some diverse opinions. In a perverse way, everybody is correct because of the fundamental truth that we are all the product of our experience and current environment. We approach the issue from different directions and naturally will spurt off down different mental pathways. If I can recap though, Anders, notes that accountants (in the pure form) may be a threatened species on the basis that many of the lower level jobs will be either automated or outsourced. Thats a no-brainer. Of course it will happen and this will leave potentially thousands of clever, numerate people with no jobs. To fill that void, accountants (and I include Finance people here too as I don't think there really is that much of a difference), should consider how to retool to retain relevance, value, and employability. To me, Data & Analytics is an obvious choice. The growth in analysis and visualisation tools have made that task much easier and it comes at a time when every company on the planet wants to exploit whatever advantage it can. I'm sure Anders was using the 'one set of numbers' phrase in the general sense because the whole purpose of analytics is to combine different, perhaps previously unthought of, data sets and relationships to reveal amazing insights about customer (or employee) behaviour. 'One set of number's is really code for 'the only set of numbers that matters is what comes out of the Finance Dept.' But Finance needs to develop the art of including non-financial numbers into its reports first or other departments will dismiss those reports as incomplete or not telling the full story.

Robert Ewalt
Title: Exam Development Manager
Company: Institute of Certified Management Accoun..
(Exam Development Manager, Institute of Certified Management Accountants) |

In my opinion, finance/accounting people should get involved in the analysis of the data, big or otherwise, throughout the organization, whether or not this data has $ signs. Our training includes understanding "gathering, classifying, summarizing, presenting data" (remember Accounting 201?). The marketing / sales / IT / HR/ operations / legal, etc. people do not have this kind of training. Data scientists or statisticians may know this, but so they know the business implications of the data?A good FP&A person can do it all.

Anders Liu-Lindberg
Title: Regional Finance Business Partner
Company: Maersk Line Northern Europe
LinkedIn Profile
(Regional Finance Business Partner, Maersk Line Northern Europe) |

Thank you for the comment Robert,
I completely agree!

Cory St Martin, CA, CPA
Title: Vice President Finance
Company: CROSSMARK
(Vice President Finance, CROSSMARK) |

Great food for thought and valid points by all.

An analogy I like to use to visualize the issues surrounding analytics and one version of the truth is as follows;

Imagine a large cube with different colors on each side, surrounded by several participants.
When asked what color the cube is all participants answer to the specific color they see.
Each participant is correct when it comes to their side, however only a participant with a bird's eye view can see that the cube is multi-colored.

This is the fundamental challenge with data and analytics, most organizational departments take a single dimensional view by focusing only on what they see in front of them, missing out on the bigger picture.

The CFO sees the bigger picture!

Topics: