The Key to Maintaining High Data Accuracy

By rah | rah | 5 Apr 2021

Whenever dealing with finance, accuracy is of paramount importance. There is an urban legend that the 2008 Banking Crisis was provoked by an error on an Excel file and more recently the UK's track and trace programme (which has been flawed from the very beginning) failed because the data required more rows than the Excel file was able to produce.

In the longer term this might actually be a problem on Celsius - I am still a long way off but in the space of just a few months I have already used more than 8000 rows - not sure how they are going to resolve that - there is plenty of column space.

So what is the key to maintaining true data accuracy?


That's right - reconciliations. Banks and other financial institutions do it as part of their BAU (Business as Usual) processes. Many of you probably already know this, but just in case.

Usually the process involves matching an authenticated form of the accounts such as a bank statement with your own records. It can be done in either direction and the point is to make them agree (hence reconciliation). It is good practice to do a reconciliation regularly otherwise tracking discrepancies may mean a lengthy trek back through a lot of data.

There are a number of authentic reasons why they might not match up which have to be factored into any calculation. Historically such factors can include

  • cheques that have been paid but not cashed out yet (I know they are becoming less frequent nowadays)
  • payments that have gone out later / earlier because of factors such as weekends
  • cash withdrawals or deposits that have yet to be added to the banking records (especially if done manually and checked quickly)

Widespread computerisation / digitisation has widely done away with much of this, but it is still good practice to check.

Anything beyond this list usually highlights an accounting mistake.

One factor that may be of particular interest to the crypto community is that Celsius Reward payments usually appear on a Monday / Tuesday, but are actually applied on a Friday. So they could easily be missed. Celsius do helpfully send an email when the Rewards have appeared.

The most common accounting mistakes are

  • missed transactions (not adding them)
  • duplicated transactions (adding them more than once)
  • transposed transactions (figures are added but somehow mistakenly - eg £5.43 as £5.34)

The latter are the hardest to spot and a good indicator is a small inaccuracy which doesn't sound like a transaction in its own right. In the above example is £0.11.

In the first two cases once a discrepancy has been spotted it can often be located with a quick search. It must however be remembered that at times there can be more than one mistake which may create a wide range of differences because they compound each other (in other words add together) or counter each other (that is one is a credit and one is a debit and so they work against each other with the inaccuracy being the net difference).

In all cases both records can be matched and in doing so an accurate record can be maintained.

A Quick Crypto Example

At the end of March I checked my DAI balance on Coinbase and found that it didn't match what my spreadsheet said. So I had to match up. First of all I identified the discrepancy and then I tried to find it. I wasn't able to find it so simply so it meant I had to go through the data from my previous reconciliation point when I knew the figures matched - as it happens I do it mid-month too so it wasn't two far.

I went through Coinbase's record and compared it to my own and very quickly found the reasons and yes there were mistakes in my record. There were two. For some reason I'd added one daily DAI Reward twice and for some unknown reason I had added a rounded figure for a transaction (eg 7.62 instead of 7.61843545)

And now my figures are right again.

If done regularly it is quite simple and it maintains accuracy and this data accuracy is essential to analysis and understanding of your big picture.

Stay safe and stay well!

How do you rate this article?



I love reading and technology as well as history. I teach English and Business to professional clients as well as soft skills with a focus on communications. I am a big fan of both Sheffield Wednesday and Lincoln City Football clubs


Experienced Business Owner and Coach and Tutor who now trades in Crypto. It is proving to be an interesting journey with so much technical language involved. Follow me as I learn the trade (and how to trade). Made some howling mistakes to begin with, but still learning and will share what I learn as I learn it for the benefit of the community. - RAH

Send a $0.01 microtip in crypto to the author, and earn yourself as you read!

20% to author / 80% to me.
We pay the tips from our rewards pool.