Thursday, September 28, 2017

Finance as the Gatekeeper in the era of Big Data

We have data everywhere  and with the advent of new technologies we  are now getting the ability to collect the data , access and use it.  Data brings with itself  the ability to mine new information and provide new insights into trends etc. But with big data also comes into question more risk of the data being compromised to outside sources, data quality issues,inaccurate data etc. The more we use data in decision making, the more important it becomes that  their are controls and compliance around the collection, storage, access and use of that data.

According to  Deloitte's CFO signals Q2 2017 report  more than half of the financial professional surveyed say they are not yet moving beyond the piloting phase for emerging technologies . Of those who cited aggressive use of new technologies 77% reported improved consistency and controls was the top improvement area while improvement in analytical/decision support was next at 75%,

Quoting Isaac Tucker, VP Product Management , Blackline, "The challenges with data will always relate to volume and complexity, and that is only going to increase as companies increase their adoption of technology. As more and more data is collected, someone has to make sure that everything is correct, and the buck stops with the finance chief.”

FP&A Center of Excellence
Issues in data management, access, quality, use etc can cause of lot of problems.  This is further amplified in the area of financial planning and analytics as this data is used for reporting KPIs, developing forecasts, making decisions for the future strategic direction of the company etc. Missing or inaccurate data can lead to incorrect decision making, delays  and missed opportunities  and in some cases compliance  issues with reporting agencies.  As I mentioned in my last blog post , Data Scientist as a Job Function within FP&A,  their is a need for the an FP&A COE or Center of Excellence where data scientists  can work together with FP&A professionals to manage data assets  and develop predictive models for forecasting. It is this centralized group that can develop data governance and compliance. Today much of the financial data  and models are distributed in excel sheets, ERP systems  and the general ledger etc.  This requires  FP&A analysts to spend time consolidating data from different sources, ensuring that the data is accurate and reconciling different versions of truth   which means that today FP&A analyst spend 90 - 95% of their time cleaning the data and getting it ready for analysis and 5-10% actually " doing " the analysis. Data governance becomes a secondary requirement in this activity especially when no central or standard controls exist.

In the end , it is the financial data that gets reported in the market  and  majority of the KPIs used to measure progress and performance  are based on financial data.  All the marketing and sales analytics should reconcile with financial analytics making finance "the gatekeeper" for data governance and compliance for the company.

Thursday, September 21, 2017

Data Scientist as Job Function within FP&A

So does a data scientist have a career path in Financial Planning & Analysis ? Why does the FP&A function need someone who has specialized in  data analysis , statistics  and data modeling. Why can't the finance function just partner with IT or data analytics team in the company to get what they need ? There are many such questions that will cross your mind when we talk about modifying the organizational structure and actually introducing someone who does not have a background in finance or accounting in the field of finance.

Every function today is dabbling in big data and analytics whether it is sales and marketing looking at customer analytics or HR looking at employee performance analytics or supply chain optimizing transportation and production   through supply chain analytics.    Mary Driscoll in her  article " CFOs want analysts trained in finance data science" refers to the study by APQC , her business research and benchmarking firm, that found that 95% of the financial professionals surveyed pegged data science as important to some degree. She aptly says" If the business is becoming data-driven, financial forecasting has to be driver-based and nuanced. And that means teasing apart probable economic consequences across the chain of value creation."

Today majority of the FP&A professionals develop forecasting models on excel and manage the models independently  with each business unit/ area developing their own models . Their is minimal synchronization or standardization of the models leading to a  lot of manual work of reviewing the methodologies so the outcomes can be compared the same way.  Differences exist due to nuances in business unit, difference of skills sets, differences in data quality  etc. These differences become greater through the years due to  lack of standardization practices.

Even starting small such  hiring an excel expert in FP&A teams who is responsible for maintaining and standardizing the model on excel, adding VB capabilities  etc can go a long way in helping the analyst actually deliver on the " analytics" .   It would free up the time of the analysts to actually  do more value creation work.  these excel experts can become the standard bearer for standardization of the models and data quality.  But as  data becomes bigger , excel based forecasting models cannot keep up with the predictive and prescriptive modeling needs.  Finance functions will need data science experts to work along FP&A analysts to  develop such models. These experts  would know the source of the data  and how to apply various techniques  to develop models  but will work in conjunction with FP&A analysts who " understand" the data as well as the business needs and will guide the data scientists to develop the right models and explore the correct relationships.

One way  is to create a center of excellence for FP&A  which will develop and standardize the models, interact with IT for data needs  and develop information assets  and explore the latest technologies and techniques in the area of financial forecasting.

The FP&A professionals today need to become the agent of change  and advocate for the need of interaction with data experts. Finance function also needs to develop a career path for finance data experts/ scientists and work with universities to develop the training programs needed.

So where is your company in this area - do you have a data expert in your team. Looking forward to hearing your feedback and comments. 

Wednesday, September 13, 2017

Process Change in Budgeting and Forecasting

As mentioned in my first blog post,  there are  five main challenges  that face traditional FP&A functions today. I expanded on the first one , " ERP systems adapting to Big Data", in my last week's blog post.  In this post I will be expanding on  the process changes we need in planning, budgeting and forecasting  to support the transformation through big data.

I read a great article by KPMG - "Planning, Budgeting and Forecasting: An Eye on the Future". It gave a good definition of what is planning, budgeting and forecasting.

Planning is defined as  a top-down strategic activity that defines the strategic aims of the enterprise and high level activities required to achieve the goals of the organisation.

Budgeting is defined as an activity that enables resource allocation to be aligned to strategic goals and targets set across the entire organisation.

Lastly, forecasting is defined as an activity that tracks the expected performance of the business, so that timely decisions can be taken to address shortfalls against target, or maximize an emerging opportunity.

 A company should do all three of the above activities  to plan, implement and measure. With agile technology development , constantly changing technology  and digitization  the way we do these essential activities of planning budgeting and forecasting within FP&A needs to change.  Today these are static activities - we plan  in our strategic planning cycles for the next 3 to 5 years , develop a detailed budget for the next 1-2 years and forecast on a monthly or quarterly basis for the current year. But by the time a budget  is developed and approved , it is already obsolete due to to the changes  that have taken place  in the market or within the company more recently.  We then develop waterfall charts to explain the difference between plans, budgets and forecasts. A lot of  the time of the FP&A analysts is spent in chasing down the data to explain these waterfalls instead or really digging into the insights from KPIs and partnering with business to take strategic decisions.

Many articles today are suggesting we should move to dynamic planning. Brian Kalish describes dynamic planning  is his article " Dynamic Planning For A Dynamic World: Are You Ready For Change?" as " Dynamic planning enables companies to evaluate risks, seize new opportunities, adjust to new challenges, react quickly and properly to threats, adapt to changing technology, and make decisions that help it thrive" . Dynamic planning would still require long term goals  to be set but it would allow the  teams to revisit the ways that would achieve those goals in shorter time horizons. If dynamic planing is done  with flexibility and agility to would allow companies to do course changes quickly instead of making big bang investments  and then being forced to be married to it for some time. Rolling forecasts  is another way that more flexibility can be added to the process of  budgeting and forecasting. But  in some industries  or companies, long term investments have to be made  which will require a long term risk taking and investment. For example , for a manufacturing company investment in a new site is a long term investment based on the current and future predictions of demand at a certain point in time. Once a commitment is made to do an investment, huge amount of capital is invested in it and it cannot be abandoned after 1 year because demand changed.

We know  the current static process is not sustainable to the quickly changing technologies  and availability of huge amounts of data, but we still   have to plan  for  long term as well as short terms planning horizons.  Changes to the current static process have to be made keeping in  mind the needs of the industry.  Balance has to be maintained in the processes to take into account quickly changing technology vs longer term investment such as in fixed assets or R&D.

I believe we need to understand  and divide the budget into short term / dynamic variables vs long term / static variables. The short tern dynamic variables  items should be reviewed on shorter time horizons  ( monthly or quarterly) and regularly allowed to be updated  to achieve the goals while the long term static variables should be reviewed on an annual basis to achieve the goals.  This would require a different way to look at the P&L and balance sheet  and partnership with the business in understanding the short term and long term decision making impacts. I call it the hybrid planning process.


Thank you for reading. Please leave your comments here or  on linkedin. I welcome your feedback and comments.

Thursday, September 7, 2017

ERP Systems Adapting to Big Data

I am not an expert in IT nor do I know a lot about ERP systems. But I am a user - specifically a user in the area of financial planning  and analytics.   As mentioned in my last blog post , one of the challenges for FP&A  in the world of Big data is the adaption of current ERP systems to the dynamically changing data / information.  This blog post tackles this issue from the point of a user.

Many companies today have already invested in a very expensive ERP system as a back bone of the financial systems.  These have been adapted  and modified to the needs to the company and industry. But the ERP systems in the current state  do not adapt well to the ever changing world of data and information.  The Aberdeen study notes that " Many organizations find that by the time they come up with a plan they are happy with , the plan is no longer feasible due to events that occurred while the plan was being composed.  They also find that they are ineffective at evaluating current trends to predict future performance and the impact of transformative initiatives".

Finance teams spend weeks or months developing a bottoms up budget  and a plan for the upcoming year . Historical data is used along with latest economic indicators  but by that time the budget is collated for an overall look of the company's performance expectations   and  approved, majority of that analysis is obsolete as the latest sales/ performance figures  need to be included, Their is a change in the market / economic indicators or new data / information is available which was not available even a few months ago.

As an analyst , I need systems that will help  me provide insights to the data / analytics rather than spend time collecting the data and then creating waterfalls to an earlier version of the budget regarding whats changed.   To  help develop an effective and efficient  organization that can maximize the benefit of  financial planning , the following  factors should be considered:

  1. ERP systems should be stable  and well defined but have the flexibility  to adapt to changes business needs quickly and efficiently. 
  2. There be the single source of truth - as analysts we spend a lot of time trying to reconcile different sources of data . Patchwork databases are created and maintained to collect data that cannot be collected in the existing systems leading to different sources of data. 
  3. Planning systems developed on top of ERP  should be flexible and adapted to the lowest / most detailed level reporting and analysis that will be needed.  The reporting systems should not be created  for the leadership only because they will ask the questions from the rest of the organization. The reporting system should be adapted to the most the detailed level analysis that can / should be provided  and then summed up from there. The first dashboards developed should be for an analyst supporting the business. 
  4. Technology updates and improvements should be done holistically - end to end - Business cannot update for their needs  without  working with finance. In the end it id the financial numbers that used for performance tracking / reporting. 
  5. Systems should have the ability to incorporate , store and use data from various sources such as government data, surveys , 3rd party data etc  so it can be used in conjunction with internal data to develop predictive analysis.  Today a lot of this is done in excel sheets as analysts download data from internal sources as well as external sources to  develop insights and then load the updated plans . These trends, changes and insights are then explained on power point decks. 
According to the CFO research done in collaboration with SAP "~90%  respondents  feel, to maximize the measurable financial benefit of financial planning and business analysis, the finance function needs to spend less time on simply moving the data around—that is, spending time, attention, and resources on manually migrating and reconciling data from system to system . ( Here is the link to the article)


Ofcourse  the real change in technology also comes with the changes in  processes also. My next blog I will talk about processes changes  that are needed with big data.