What accountants can learn from Tesla and SpaceX

I recently finished ‘Elon Musk: by Walter Isaacson’ and true to the author’s previous subjects with world-changing feats including Steve Jobs, Leonardo Da Vinci and Benjamin Franklin, this book lived up to the hype.

It’s unfortunate that Elon’s current profile is overshadowed by his social media antics and political viewpoints but there is no doubting that that out of his contemporaries in Silicon Valley, I’m convinced he has a conviction to serve a higher purpose for humanity other than making money.

Some of his amazing feats I was unaware of were:

• He is one of the founders of what was eventually Paypal, the world’s largest payments processes

• He could easily have retired on yachts and jets for the rest of his life, he risked it all to create SpaceX, which almost went under after three failed launches. It went on to be the first to create reusable rockets and the only private company to send payloads to the International Space Station with a core mission to create a sustainable colony in Mars

• Created the largest internet satellite constellation with Starlink, changing the way internet is accessed in infrastructure deprived locations

• Is one of the co-founders of OpenAI and could be argued that ChatGPT would not exist today were it not for his support to challenge Google’s early AI dominance after its acquisition of Deep Mind

• Founded Neuralink to build a direct interface between machine and human through the implantation of chips

It’s a big read and the book also delves heavily into his family and personal relationships, which appealing to my lower-brow interests, was equally as interesting. That aside, the passages that resonated most with me was his five-step process, which Musk called the 'algorithm'. It's a distillation of lessons learned while trying to increase production at his Tesla factories. He also revealed that the model was applied at SpaceX on when trying to overcome challenges on the design of his rockets.

Elon Musk’s five-step design process

Applying the algorithm to the finance function

1. Question everything

   This is one I resonate with most. Whether deploying a new ERP, transitioning to a shared service or reorganising finance, the first remark is ‘it’s required’. The only requirements are those mandated by laws and regulations. Everything else should be questioned. Even company policies should be challenged if it prevents change that makes the business better off. Too often, we retain unnecessary processes because ‘it’s always been done this way’ and accept that this is the way.

   
   

   Many year’s ago, I was charged with redesigning a finance organisation to split activities between accounting and business partnering. Historically, the latter was largely responsible with month-end close such as extracting data and pivoting spreadsheets to calculate the prepare cost and revenue accruals. The justification was that they knew their data better than everyone else. While this was true, this prevented business partner’s from achieving its main objective – to partner with the business to provide them with guidance to make good commercial decisions.

   
   

2. Scrap unnecessary activities and tasks

   One morning after we had gone live with an ERP deployment, I sat with a client to hear why certain journals needed to be approved by the Senior Finance Business Partner. Some of these journals were loaded directly from batch files containing thousands of lines and when questioned whether these approvals there actually reviewed and whether they’d ever been declined, she insisted that this was the process and it needed to be done due to internal controls. The reality was all journals were approved and even when these postings contained material mistakes, a reclass journal was reentered with another approval. The approval here served no purpose and got in the way during late evenings in month-end when senior finance staff had to login to click a button so that numbers could be posted and reported.

   
   

3. Simplify tasks you must perform

   It’s important that we shouldn’t undertake this step until completing the first two steps. I learned this lesson as a junior analyst where a client had performed cost accrual journals on each working day during month-end. As the dataset on each day varied, I had to write a script to transform the data into an accrual journal for each day. Later on, we had realised that doing a daily cost accrual five times during month-end achieved the same result as doing it once on the last working day, rendering the work to optimise each day to waste.

   
   

4. Find ways to improve dependencies so that tasks are performed more efficiently

   You should only be doing things efficiently if they’re worth doing in the first place. Following our mishap on the cost accrual, we had decided that posting this journal once at the end of the month-end cycle meant we had to take a holistic approach to ensuring we had an accrual that was efficient to generate and posted accurately. This meant:

   This meant:

   • We wanted to minimise the invoices that needed accruing by making sure they were posted before month-end close by automating reminders to approve their invoices and encouraging vendors to send invoices prior to the month-end rush

   • Working with requisitioners to code purchases properly so that related invoices would be enriched with financial data that minimised the time spend from the general ledger team to clarify which accounts and cost centres amounts needed to be approved to

   
   

5. Automate

   We eventually turned on features in the finance system to extract unposted invoices and transform them into a batch of accrual journals that would automatically reverse in the following month, which completely eliminated the manual effort needed to perform this once laborious task.

   
   

   Creating automation through developing applications, building interfaces or enabling system features is time consuming and often not fit for purpose unless the need to do so is clearly defined and the parameters are precise. It was only when the process was completely stripped back and the task was optimised that automation was made possible with positive results.

The issue with many transformation projects

Many consultants don’t come from a practitioner’s background and have never been in the trenches of the finance function yet know all the buzzwords of change, transformation, process optimisation, machine learning and automation.

I’ve had experiences where advisors have come in proposing big projects to deploy automation software, without really understanding the specific problem this software is trying to solve and the precise benefits it brings. Instead, what happens is glitchy solutions that need patching resulting in broken promises, high costs and change fatigue.

What sets Sub One apart?

Sub One is dedicated to making finance better by using our strong foundation as finance practitioners. Like you and your team, we’ve been through the pain of running a finance function and the promise of making it run efficiently and effectively. We understand the complexities of month end close, submitting tax returns, working with auditors, collaborating with shared service centres and familiarising with new finance systems and ERPs.

Sub One is dedicated to making finance better by using our strong foundation as finance practitioners. Like you and your team, we’ve been through the pain of running a finance function and the promise of making it run efficiently and effectively. We understand the complexities of month end close, submitting tax returns, working with auditors, collaborating with shared service centres and familiarising with new finance systems and ERPs.

Our team will work at every level of your finance function to focus on transforming the areas of finance that makes a real difference to your core objectives.

Credits

Header image provided by: Shortform

Elon Musk: by Walter Isaacson is now available on Amazon UK

Back to Articles