One of the biggest questions I hear from small and medium sized businesses is a simple one. Where do we start with AI and automation?
The temptation is to dive straight into tools. Try a bit of everything. See what sticks. The problem is that without a clear starting point, you end up with a few half finished experiments and no real time saved.
This is where a scalability audit comes in. It is a short, practical exercise that shows you exactly which tasks are slowing you down and which ones are ready for AI or simple automation.
No heavyweight consulting. No complex frameworks. Just a structured way of looking at how you spend your time.
What is a scalability audit?
A scalability audit is a simple review of what you actually do during the week, rather than what you think you do.
The aim is to answer three questions:
- Where does your time really go
- Which tasks create value for the business
- Which tasks are repeatable and suitable for automation or AI support
Once you can see that clearly, choosing what to automate first becomes much easier.
Step 1 - Track your work for 3 to 5 days
For three to five working days, track what you do. Nothing fancy. A notebook, a basic spreadsheet or a simple notes app is enough.
For each block of work, capture:
- Task name - a short description such as "write proposal", "answer support emails"
- Time spent - roughly how long you were on it
- Frequency - how often this type of task comes up in a typical week or month
The important bit here is mindset. Do not try to tidy or optimise as you go. Just record what is really happening. You can clean it up later.
Step 2 - Mark value creating versus admin
Once you have a few days of notes, go back through the list and mark each task as either:
- Core value - activities that directly move the business forward, such as sales conversations, delivering work for clients, strategic planning
- Admin tax - necessary but non strategic tasks, such as formatting documents, copying information between systems, chasing signatures, retyping notes
Both categories matter. The point is not to eliminate admin entirely. The point is to see how much of your week is spent on work that could be streamlined, delegated or automated.
Step 3 - Look for patterns and repetition
Next, scan through the admin tax items and ask:
- Which tasks appear every day or several times a week
- Which tasks follow the same pattern each time you do them
- Which tasks make you think "this really should not still be manual"
These are your first candidates for automation or AI support. The more repetitive the task and the more often it appears, the higher it should move up your list.
Step 4 - Score tasks for impact and effort
To decide what to tackle first, give each candidate task a very simple score from 1 to 5 in two areas.
- Impact - how much time or frustration you would save if this task were streamlined
- Effort - how difficult it would be to improve or automate this task
High impact and low effort tasks are your sweet spot. These are often things like:
- consistent follow up emails after meetings
- turning meeting notes into action lists
- creating first draft social posts from blogs
- standardising parts of a proposal or quote
If something is high impact but also high effort, it may still be worth looking at, but it is rarely the best starting point.
Step 5 - Decide the right solution type
Not every repetitive task needs an AI assistant. Some can be solved with simpler tools.
For each high impact, low effort task, ask:
- Is this mainly about moving data between tools - a job for basic automation with something like Zapier or Make
- Is this mainly about drafting, improving or restructuring text - a job for an AI assistant in ChatGPT, Claude or Gemini
- Is this mainly about team habits or process - a job for clearer instructions and ownership first
The right answer might be a mix. A simple automation moves data from your form tool into your CRM, then an AI assistant drafts the follow up email, and you give it a quick check before sending.
Examples of what a scalability audit reveals
When SMEs carry out this kind of audit, a few familiar themes tend to appear.
- Proposal heavy weeks where the owner spends hours copying, pasting and adjusting the same sections
- Follow up gaps where leads go cold because no one has time to send consistent messages
- Reporting overhead where client reports take far longer than they should because data and commentary are scattered
- Content bottlenecks where ideas exist, but nobody has time to turn them into posts or emails
These are the kinds of patterns that make great starting points for AI assistants and small automations.
Keep it honest, not perfect
The biggest risk with a scalability audit is overthinking it. This is not a performance review. It is not a test. It is simply a snapshot of how work really gets done.
Messy notes are fine. Imperfect time estimates are fine. What matters is that you see, clearly, where your energy goes and which tasks are quietly taxing the business every week.
What to do with your audit results
Once you have your list of candidate tasks, you are ready for the next stage of this series.
In the next article we will take one of those high impact, low effort tasks and walk through how to turn it into your first AI assistant. You will see how to turn a rough process into something clear, repeatable and supported by AI.
For now, your job is simple. Track your work for a few days, be honest about where your time goes and notice which tasks you never want to have to do manually again.
That is where AI can start to make a real difference.