AI is exciting and is dominating the headlines. But, if you still have data processes that involve downloading data from one system, messing around with it in Excel, then uploading it into another system, you should be looking at automating that before anything else. And that goes for most accountants!
There's a ton of proven, affordable solutions that can deliver immediate ROI. But for many in the office of finance, they seem technical and hard to understand — but with a few tips you can get started today.
The business case for integration
First, it's good to think about why we might want to do that. Since so many of us process data flows manually, it's easy to forget that it doesn't have to be that way — and that it comes with steep costs.
Time is a huge one. Every accountant wants to be spending time on delivering value to the business; spreadsheet jockeying takes us away from that.
Errors are another. Manual work always invites errors! And then we waste even more time double and triple checking.
Manual processes also hold us back from timely reporting. How useful would the business find it if we could deliver numbers every day instead of every week or month, with no additional overhead?
And don't underestimate the impact on morale. Will people want to work for us if we make them waste time on manual data work? Something to consider as we navigate the shortage of accounting talent.
Exploring automation
Getting automated doesn't have to be scary. First, take an inventory of what you need to integrate — and don't just consider the worst offenders.
What are your systems — and are they on-premise, hosted or cloud? How often do you need the data to move? What time period should each process cover? Are there any data quality checks that need to happen (for missing fields, incorrect formats etc.)?
Most importantly, what data needs to move where, and do we need to change it along the way?
Investigate your options
Then, there are three main options to explore.
Many of your systems will probably have a set of built-in or "native" integrations with other tools that they are commonly used with. They're always worth checking out as they can be very convenient, being within a system you already use, and can often solve a problem for you very neatly — often at a very good price point.
However, they only offer specific use cases; if you want something else, you may be out of luck. They also may not scale with volume — it's best to test with the vendor. Most importantly, they only solve a slice of the problem. You have a whole pile of applications that need to work together, and it's unlikely that built-in integrations exist for everything you need. Even if they do, the ease and cost benefits of this option start to go away when you have a lot of them!
The second option is for a consultancy to create you a custom integration — most commonly the firm that sells you your accounting system. There are a lot of possibilities here, so be sure to ask for more details if your reseller offers this option. This can be a great option to deliver relatively complex requirements at a reasonable cost, but be sure to bottom out a) initial costs, b) ongoing costs and c) have some idea of what costs will be involved if changes are needed — either because your business changes, or the systems do.
That future cost uncertainty can be a downside to this approach, along with potentially requiring a project from your consultant any time you want to make a change. Custom integrations also don't typically have advanced, useful features that you might find in a dedicated tool.
Finally, you can look at dedicated integration software, sometimes called Integration Platform as a Service or iPaaS. These vary hugely from simple self-service tools to high-end enterprise software, which is why it's always important to understand your requirements. These tools offer the broadest flexibility and premium features, allowing choice in notifications, scheduling, data checking and the like. As a result, historically this has been the most expensive option, but it's worthwhile to look at all your options — the one that is most cost effective may surprise you.
There's a lot more to think about, but if you haven't done so, why not make 2024 the year you look at the true costs of manual data processes, and explore your options to automate them?
