Three Things You Need Before Automation Makes Sense
Automation and artificial intelligence are the hot words right now.
Everyone wants to buy software or equipment that will "make things faster" and "reduce errors."
Here's the problem: Most processes aren't slow because they're manual. They're slow because they're broken.
Unclear handoffs. People waiting for information. The same question getting asked three times because nobody documented the answer.
Automation doesn't fix that. It just does the broken process faster.
Before you invest in automation, you need three things:
1) A Process Worth Automating
The real problem is that most processes have waste throughout them. Before automating, you need to walk through the process, map out the current state, identify the waste, and eliminate it.
What waste looks like:
Waiting - For approvals, information, or other people
Searching - Hunting for documents, data, or answers
Rework - Doing it over because something wasn't clear
Handoff delays - Work sitting between steps
Workarounds - People bypassing the "official" process
Example: A manufacturer wanted to speed up their quoting process. Sales was losing opportunities and deals were moving too slowly because quotes were taking too long.
The obvious answer seemed to be automation - software to speed up the process.
But when I mapped what was actually happening, here's what I found:
The waste in the process:
Quotes sitting with no prioritization system - first in didn't mean first out
Sales submitting requests without all the information engineering needed
Engineering waiting days for vendor quotes on materials or tooling
When engineering had questions, quotes sat in sales' queue waiting for answers from the customer
Once answers came back, quotes sat in engineering's queue again
Data being entered multiple times across different systems
Team capacity unbalanced - some people overloaded, others with bandwidth
No single person owned getting the quote out
The key insight: Most of this waste had nothing to do with speed. Automation wouldn't have fixed most of it.
We needed to fix the process first.
2) A Standard Everyone Follows
Once you've fixed the process, everyone needs to do it the same way. Not "their way." THE way.
If three people handle the same task three different ways, you're not ready to automate. You're about to lock in inconsistency.
What standardization actually means:
Document the improved process (not the old broken one)
Make it accessible at the point of use
Get buy-in from the people doing the work
Assign clear ownership
Actually follow it
Continuing the quoting example:
After identifying the waste, we had to standardize how quotes actually worked:
What we standardized:
Single owner for each quote - One person responsible from sales request to customer delivery
Required information checklist - Sales couldn't submit without all the inputs engineering needed
Clear prioritization criteria - Rush quotes, strategic accounts, and standard quotes had defined rules
Communication protocol - Questions went through the quote owner, not back and forth between departments
Standard data entry - Information entered once, used everywhere
Why this came after fixing the process: We didn't standardize the broken process. We fixed the handoffs, clarified ownership, and removed the waiting first. Then we documented what worked.
The test: Ask three different people how a quote flows through your system. If you get three different answers, you don't have a standard - you have individual interpretations.
3) Proof That It's Stable
Before you invest in automation, you need to know:
What your current performance is
What "better" would look like
Whether the gap justifies the investment
You can't improve what you don't measure. And you can't justify automation without a baseline.
What to measure:
Cycle time (how long does it take?)
Error rate (how often does it go wrong?)
Capacity (how many can you handle?)
Cost (what does it cost per transaction?)
Continuing the quoting example:
After fixing and standardizing the process, we measured it for several weeks to establish a baseline.
This showed us two things:
The improvements we'd already made had significant impact
Where automation would - and wouldn't - add value
The test: Can you answer: What's our current performance, what would success look like, and does the gap justify the investment?
If not, you're not ready to spend money yet.
Closing
Back to the quoting example:
After we fixed the waste, standardized the process, and measured the baseline, we could finally see where automation would actually help.
We implemented targeted automation for specific steps, like eliminating duplicate data entry and automating parts of the workflow.
Combined with the process improvements, quote turnaround time dropped 75%.
The lesson: Automation worked because we fixed the process first. If we'd automated the broken process, we would have just made expensive chaos.
The Real Sequence:
Fix it - Remove waste and unclear handoffs
Standardize it - Document what works
Measure it - Establish your baseline
Then automate - Make the good process better
Most companies skip straight to step 4.
Result: Expensive software that doesn't deliver. And teams working around automation instead of with it.
Before your next automation project, ask:
"Is this process good enough to amplify?"
If people are waiting, searching, or working around the system, the answer is no.
Fix it first. Then automate it.
