Automation without AI — a beginner’s guide

The whole point of automation is to reduce the amount of time spent on your organisation's common repetitive tasks. For example, if you’re a charity, taking a donation or making a Gift Aid claim would be a good candidate for automation. So what are the tasks you spend the most time doing each day?

Bear in mind that just because you do one particular task a lot, that doesn’t mean it’s a great candidate for automation. Automation really only works if the repetitive task is truly repetitive, meaning it follows the same predictable set of steps each and every time. Again, taking a donation is a good example, as is selling a ticket or updating a customer’s details. The fundamentals of these transactions never really change, so they’re easy to automate.

If you’re an architect, you clearly can’t automate your core services in the same way: the processes and creativity required to design and deliver buildings change too much from project to project. On the other hand, you likely still have lots of truly repetitive tasks that automation can help with — sharing updated drafts with clients, for example, and gathering their feedback.

Finally, it’s worth noting that just because you can automate something, that doesn’t mean you should. Make sure you ask the right questions about whether your automation is a good idea.

Once you’ve found a good task to automate, you don’t need to jump straight to an expensive and prolonged development project – in fact, you probably shouldn’t, in case you spend all that time and money automating the wrong thing.

In its most basic form, an automation is just a mechanism to ensure that the same task gets done in the same way every time. So instead of focusing on an all-bells-and-whistles technical solution, think about how you can try some lower-stakes automation to solve your problem first. That way, you can start to get an idea whether your automation will actually make a difference before you invest a load of time and money.

For example, you could put together a checklist, with step-by-step instructions, for following up on every donation you receive, such as:

1. Send a thank you email
2. Check to see if the person made a Gift Aid declaration (and follow up if not)
3. Dedupe them in your database
4. Tag them with relevant categories in your CRM, and so on.

This might not save you much time in the short-term, but that doesn’t mean it doesn’t have value as an automation. Maybe you’ll start getting more Gift Aid declarations. Maybe, when you’re doing your year-end reporting, you won’t have to spend ages untangling which customers are missing which tags, because the checklist means you tagged everyone correctly at the time. If nothing else, think how much easier a checklist like this will make onboarding new hires.

If a checklist feels too lofi, you could also look at ways to use technology you already have for some simple automations, rather than building out a new custom solution of your own. A spreadsheet can help you get through complex calculations you do frequently; ChatGPT can help you turn meeting notes into an executive report. Or your device’s OS might already have tools for creating your own custom automations of repetitive tasks. As a society, I’m sure we’ve collectively saved years of time since Siri was launched, simply by asking “how old is [celebrity]?” instead of going to Google to type in the same question.

This means defining up front what you think your automation will accomplish, working out a way to measure success, and establishing a baseline. It’s important to establish a clear hypothesis upfront, otherwise you’ll find yourself lost in the data six months from now, unsure what to look at or what any of it means.

So whatever matters most to your organisation — whether that’s revenue, sales, customer satisfaction scores, etc. — figure out which metrics you think your automation will affect, and then set about measuring them for success.

To go back to the above example, maybe you think your checklist reminder to follow up on missing Gift Aid declarations will cause the number of declarations to go up.

In that case, before you start using the checklist, look at a year of historical data and calculate the average number of Gift Aid declarations per month. Then, in the three months after you start using your checklist, see if that number is consistently higher than the old average.

If so, that’s a sign your checklist is making a positive difference. Which brings us to…

If the GiftAid follow-up on your checklist seems to have increased declarations – and hence income – you now have a great case for investing money in that automation. Can you make it even better with technology?

For example, can you set up your website or CRM to send your follow-up emails automatically? Or better yet, can you update the UX of your donation pathway to follow up during the transaction itself?

This still doesn’t need to mean a big budget development project. Maybe you just need to spend half an hour adjusting the wording on your donation form, or pay your web developers for an hour or two of time to move some elements around or add some conditional logic. If that effort translates to a significant increase in fundraising income, the time will pay for itself.

And if, after that, you think you can make an even bigger impact by building a custom AI chatbot to take donations — well, great! Just remember, you don’t need to start there.

So don’t be afraid to get your feet wet with something smaller. And once you’re ready to scale up, give us a call.