Like in most knowledge-based professions these days, we too get this future-inspired question: ‘What’s the point of you as financial analysts when you (like my accountant) will be replaced by Artificial Intelligence’?
Let’s start with perception vs reality
First, your accountant is not going anywhere soon, unless you have decided to become an expert in taxation.
AI developments promise simple solutions for complex problems. That is perception. In reality, complex problems require complex solutions, but the end result can be displayed in a user-friendly manner to give you the impression of simplicity.
With AI, there is a promise of no need for human interference with processes. Indeed, provided they have already been set up by a human with a specific purpose.
With AI, there is a perception that what you see is intelligence that eliminates the need for a human brain to interpret it. That is a dangerous idea. What you see is a report or display of something, usually aggregated data. You still need to know how to interpret that thing to turn it into ‘intelligence’. We know of big companies that implement systems and then start generating reports at which executives stare, waiting for a revelation, as they don’t really understand what they’re looking at. Let alone what to do with it.
What does the above question mean by Artificial Intelligence?
In financial data for small business, because the volume of data is so low, you need at most an integrated accounting system that can collect the data and produce some form of reporting. Many systems do it – see for example, Xero. If the accounting data system also has some time recording and operational data capabilities that can be integrated with finance data, then Bob’s your uncle. That is how far the part of ‘intelligence’ goes.
And while having an intelligent system that streamlines your work by e.g. learning your client names or types of services will speed up your invoicing process, watch out for some learning glitches. Such as applying VAT to all your expenses, including taxes.
How intelligent is Artificial Intelligence?
The name that companies rush to apply to new developments may be misleading. The ability to aggregate vast amounts of data using an algorithm does not make ‘intelligence’ in every sense, but only in some – intelligence as in information. For example, it does not teach you how to see it and what to make of it. Intelligence as in ‘supporting decisions without your input’ requires a human brain with the understanding and ability to interpret what the system shows and asks it for adaptations.
In conclusion, the use of AI systems comes with one hefty condition: that you know what the system does and you know what you want it to do for you. It does not supply you with the knowledge about itself. Think Google SEO: to be able to use the advantages of top ranking in this search engine, you need to ‘know’ Google and how it works. If that were not the case, your company would be in top ranking right now without the help of an SEO marketing company – is that your case? More specifically, you need to know what debt to equity ratio is and whether you need it when it comes to financial intelligence.
All small businesses who implemented a Xero-like system have an accountant to check the input and output and report properly. And some still don’t know their profit per service line, even if the data is there. When it comes to your finance and making sense of it, here is a list of what AI can or cannot do for your small business finance data and analysis:
What Artificial Intelligence can do for you:
- Speed up the input process e.g. your receipts can get processed by scanning into the pre-ordered fields automatically
- Ensure that the data input has no gaps; for example, you can only input data that has all fields filled in.
- Multiply an existing pattern and save time and energy in replicating those patterns using new data for an integrated view
- Speed up your regular analysis process by providing you with clean, complete and accurate data e.g. quarterly review of your finance cut from 1.5 days of work to 2 hours by downloading pre-categorised revenue and expenses
- Easier to archive and consequently dig out past data for longer-term trend analyses and comparison – this is a huge benefit that Excel archives cannot accomplish.
What Artificial Intelligence cannot do for you:
- It cannot correct you if you by-pass the rules: if you decide to get creative with the data input and bypass the “rules” because your reality does not match their model, the system cannot stop you – garbage in, garbage out.
- It cannot analyse something that it has not been pre-programmed to take account of – my favourite example is ‘other expenses’ in accounting that don’t get allocated and affect your gross margin calculations
- It does not know your business specifics – it may know general business principles for example in allocating costs, or even your industry specifics (e.g. the cost model expected for a construction company) but not your suppliers and the names that you use to identify your costs from such suppliers
- It cannot show you something that is relevant for you but which it has not been pre-programmed to show you i.e. it cannot show you that which you did not know you did not know.
- It cannot interpret what it shows for you. It will not explain what the implications are for the future that the trend displayed is in one direction or another – you need to do that yourself.
Basically, to rely on a system to drive your finance data and insights is possible provided you:
a) already know finance and financial principles of budgeting, etc.
b) you have analysed your data and know what system you need and what you want it to show you regularly
c) you know how the system works so well so that you can customise its reporting to your needs, and finally…
d) you know how to interpret the reports to understand what decisions to make as a result of that reporting.
What we find is that the entrepreneurs know very well what their business does and how to serve their customers, but less about finance, financial indicators and financial systems. And that is the point of us.