How to determine which business processes should be automated

Process automation is becoming increasingly prevalent. Seemingly a thing of the past, I’ve found that businesses are now starting to realise the benefits from process automation, and are looking to make other processes even more efficient with the use of technology. There is also an incredible number of startups structuring their businesses for scale using the right software tools (which may seem obvious, but previously wasn’t cost effective to do). This has all become possible with value based pricing (the model enabling startups to purchase the same software as larger organisations, at a fraction of the price).

With fear of becoming a case-study-dropper, but also being eager to share with you the value automation can add, here are a few case studies to motivate you:

  1. Coles are spending over $700 million over the next five years on automation (see here).
  2. Meridian Lawyers will spend approximately $1 million on automation over the next three years (see here).

Before we delve into how you can determine which business processes should be automated, a quick note on morals. I take a strong personal stance on not replacing staff unnecessarily, but instead using automation as a means of augmenting their efforts. Enabling people to achieve more with the time helps their ideas to flourish and keeps their jobs interesting. This leads to (in my experience) much better results to the business. So, please use this advice wisely, and don’t pull the same bulls!@# as Telstra or NAB.

Without further ado, let’s start by determining an automatable process.

Step 1: List the ideal outcomes of automation

We’ll start by flagging any potential business outcomes you’re looking to achieve with improved efficiency. These shouldn’t be too general (a mistake, for instance, would be to make processes more efficient). Let’s instead focus on organisational sectors specifically:

  1. Marketing: Reduce the time taken to migrate leads between platforms.
  2. Sales: Increase visibility of the sales pipeline.
  3. Accounting: Decrease the time taken to invoice clients each month.

These are still somewhat general, but provide enough for us to work with.

Step 2: Outline the inefficiencies that already exist

Next, we’ll identify inefficiencies within the current process. From each of the outcomes identified above, we’ll provide a problem that currently exists:

  1. Marketing: Leads need to be manually migrated from our email marketing tool to our CRM.
  2. Sales: Information presides across multiple platforms, preventing simplified reporting.
  3. Accounting: Billable hours need to be manually converted into invoices and sent to clients.

Flag any difficulties that have led to this, and ensure you really spell out the process involved:

1. Marketing
Our email marketing tool is Mailchimp, and our CRM is Hubspot. These two systems do not seamlessly integrate, so the data needs to be exported via CSV from Mailchimp and imported to Hubspot.

2. Sales
We have tools such as Google Analytics for marketing data. However, our sales data sits across LinkedIn Sales Navigator, Hubspot and other smaller data tools.

3. Accounting
Our time tracking software generates invoices from billed time. That said, each client has commercial rules that need to be applied against their account, and these then need to be manually invoiced to the client using a custom email format.

Step 3: Convert the inefficiencies to processes

We then convert each of the above inefficiencies into a process that would theoretically resolve the issue:

  1. Marketing: When a new subscriber is captured within Mailchimp, also create a lead in Hubspot.
  2. Sales: Amalgamate the sales data from LinkedIn, Hubspot and other tools into a single platform for simplified reporting.
  3. Accounting: At the end of each month, create and send an invoice to each client applying the business’ commercial rules for that client.

It’s important to focus here on the business outcomes that are relevant to you, whilst also noting that by this stage we have converted an outcome to a process that can theoretically be automated.

It’s nice to live in Neverland, let’s now try to actualise these ideas.

Step 4: Rank processes based on the overall cost benefit

I tend to focus on cost as a factor of time required to setup the automatable process. Naturally, benefit will be the time saved per month across the whole department. Use a table to draft the cost and benefit for each process we just created. Finally, create a column that demonstrates the payoff period (the number of months before the process is completely automated).

It’s worthwhile to identify that the “benefit” column can take some time to confirm, as ultimately you will need the input of a consultant to draft an outline of the time required to automate this process. However, what you now have is a list of priorities that provide a ranked ROI. Logically, we would proceed to automate these tasks as follows:

  1. Marketing
  2. Accounting
  3. Sales

The final considerations before proceeding are:

  1. Consider the pain points of your team, automating the tasks they find to be cumbersome. There isn’t always a direct correlation between time spent on the task and the frustration it causes for your team.
  2. Don’t automate tasks just to replace staff. This will go pear shaped (see employee morale, public relations).

All the best!

CEO of Station Five