In his blog our CEO shares personal insights from a career in accounting and tech
Analytics in audit and accounting is not a new phenomenon in principle. Those working in the profession for long enough will recall multiple earlier incarnations.
Indeed, when my Inflo co-founder Graham Clark and I first worked together at PwC we spent many hours running CAATs (Computer Assisted Audit Techniques), the buzz topic of the time.
Cynics may point to such trends being cyclical, but it is different this time.
Primarily that difference is the technology itself. And underpinning that is the movement from desktop tools to cloud applications. On desktop you can do Data Analytics. But on the cloud, you can do Big Data Analytics.
So, what are the dynamics at play behind this movement?
1) Data Volumes
Moving away from analysing data sub-sets to analysing whole ledgers naturally increases the size of data you’re working with. Compounding this, organizations have become more transactional in the way they do business.
Analysing data sets with multiple millions of lines is challenging, if not impossible on desktop with a laptop’s processing power. Graham and I commonly had to leave the software to run overnight and hope we defined the script correctly when we returned the next day to view the results. But now the exact same tasks take seconds with cloud computing.
In its infancy accountants were concerned by the cloud. Most likely, because the rather unhelpful name conjures the image of an easily penetrable object floating around. Not a great image for confidential data… But cloud is now commonly accepted to be far more secure than desktop.
If your client provides you with a full backup of their system, would you rather it was emailed over and saved on the C: drive of multiple staff laptops, or securely transferred and held within cloud database storage protected by Microsoft security experts?
3) Implementation and Integration
Many firms are new to the field, so want to start small. Perhaps trial on a handful of clients first, understand analytics better then implement more broadly. With cloud you can be up and running in minutes using new techniques, without complexities around installing new software or even having to upgrade hardware. And if it isn’t right for you, then you’ve made minimal investment and can move on quickly.
When you do get to scalable use, it’s far easier to integrate cloud applications into your existing processes and technologies. Application Programming Interfaces (APIs) make it far easier to integrate different tools. Central monitoring and review of adoption is also far easier.
4) Skills & Standardisation
One of the biggest challenges with desktop analytics is the specialist skills needed to do sophisticated tasks. I couldn’t personally write the routines I wanted to run at PwC – I needed Graham sat with me to turn my ideas into “scripts”, or code. To scale that model meant a large team of specialist data scientists was needed (more Grahams…), with engagement teams dictating what they wanted. Both costly and risky when a lack of understanding on both sides became apparent.
Now, that challenge is gone. Cloud applications have standard in-built routines and easy user interfaces for teams to run bespoke routines themselves. So, the engagement team can work directly in the application, without the need for a team of data scientists driving it.
5) Benchmarking & Client Expectations
The standardization of routines in cloud analytics contrasts with the often very bespoke nature of desktop analytics. This causes risk in the consistency of work (simple things like 2 teams defining non-working days differently) but also inhibits the broader value you can gain.
Desktop tools are primarily designed to run routines over 1 data set. Cloud applications are designed to run routines over 1 data set AND compare results against all other data sets in the application to benchmark.
Clients find analytics on their own numbers interesting, but they can do that themselves. Clients find benchmarking a clear differentiator and something they can’t do themselves. It positions you completely differently as an advisor. You can see an example of what I mean here.
6) Next Generation
To leverage more advanced capabilities, such as robotic process automation, machine learning, AI and cognitive computing, you need a few key ingredients. 1) a vast quantity of structured data upon which analytics have been performed on, 2) a vast quantity of structured data analytical results, 3) a vast quantity of structured data on the outputs human users turned the analytical results into.
If all the data, results and outputs are scattered all over your business in different places, structures and formats, then you have nothing but a burden. But if you have all this information structured, accessible and perhaps in a format comparable to peers, you have created an asset.
So, are desktop analytics tools still relevant? Well, yes…
Desktop tools still have a place in the world, but it is narrowing to where highly bespoke data analytics are needed highly infrequently on single data sets. If you desire more scalable adoption of analytics across people and clients, to draw value from upgrading to Big Data Analytics and desire a data asset to set you up for the future, you a cloud application.
View our on demand webinar from to hear 3 case studies from firms of various sizes regarding their technology journey. Run in partnership with ICAEW but free to non-members too.
Inflo President & CEO