AI is Only as Good as Your Process

Why glazing firms need to fix the flow before they automate

If there’s one lesson I’ve learnt from helping glazing businesses adopt AI, it’s this: AI doesn’t fix chaos. In fact, when you apply AI to a broken process, you just make the chaos faster. And more expensive.

That might sound blunt, but it’s an important truth for fabricators, installers, and even suppliers who are trying to “use more AI” without really knowing where it fits. The good news? If you already have a reasonably structured workflow, AI can save hours, even days, of manual effort.

Let’s take a practical example: quote checking. It’s a task almost every business in our sector does, whether you’re reviewing incoming orders, pricing up a customer job, or checking supplier quotes. It’s also an area where AI can add serious value if the groundwork is in place.

The Everyday Reality: Manual Quote Checks

Here’s what quote checking looks like for most glazing firms today.

A typical request arrives by email, often with an attached PDF or Word document. Something like:

“Please quote 6 x white uPVC casement windows, 1200x1200, A-rated glass, chrome handles.”

From there, someone in the office (often with years of product knowledge in their head) manually reviews the spec. They cross-reference the details against price lists, product availability, known configurations, and recent jobs. They’ll likely spot if the spec doesn’t allow for trickle vents or if the customer forgot to ask for cills. Maybe they even check whether the U-value meets the latest requirements.

The job gets priced, questions go back and forth, and eventually the quote gets sent. It’s slow, error-prone, and depends heavily on who’s doing it.

The Same Workflow — Supercharged by AI

Now let’s imagine that same business has taken the time to document their quoting process. They’ve got templates for common product types, consistent terminology, and standardised options for handles, finishes, and glazing.

In this environment, AI becomes a powerful assistant.

When the quote request comes in, AI can:

• Read and extract key details from the document (product types, dimensions, finishes, etc.)
• Match those details to internal product codes or SKUs
• Check for errors or omissions (e.g. no cill specified, incompatible spec)
• Verify that the proposed window spec meets current Building Regulations
• Draft a polite response or quote summary with suggested changes

Instead of spending 30 minutes on a manual check, the human now spends five minutes reviewing what the AI has prepared. It’s not about replacing people, it’s about giving them a running start.

The Critical Factor: A Structured Process

But here’s the catch: this only works if the business has a clear, repeatable process to begin with.

If every salesperson writes quotes in their own style, with no shared product names or standard specs, AI won’t know what to do. One job might say “white casement window, shiny silver handles”, another says “double opener, 1fixed 1200x1200”, and a third says “two side hung sashes with chrome hardware”. Same product; three different descriptions. AI can’t fix that inconsistency.

This is why throwing AI at a messy process just speeds up the confusion. You still miss things. You still redo work. You just do it faster.

Fix the Flow First

If you’re serious about using AI in your glazing business, don’t start with the tech. Start with the flow.

• Map the process Whether it’s quote checking, lead handling, or job scheduling, write down how it works, step by step.
• Standardise your inputs Use consistent templates for incoming requests. Define product names. Set clear defaults for handles, finishes, or hardware.
• Only then, add AI Once the process is stable, tools like ChatGPT or automation platforms can be trained to follow the steps, fill the gaps, and check the logic.

At thinkivity, we’ve helped dozens of businesses walk this path and the difference is night and day. When your process is sorted, AI becomes a genuine timesaver. Without it, AI just adds another layer of complexity.

What About Tools?

This is where people often ask, “So what’s the best tool for that?” The truth is, there isn’t one. In fact, most of what you need is already in front of you.

AI platforms like ChatGPT can read documents and make comparisons. Automation tools like Make.com or Zapier can trigger alerts when a quote looks suspicious. Google Sheets can be your audit log. But none of that matters if the business hasn’t agreed on how a good quote should look.

That’s why we built GlazePro AI. Not as a tool, but as a membership for glazing professionals to get this stuff right. Through weekly sessions, live support, and shared prompt packs, we help firms across the sector map their processes and learn how to use AI in practical, real-world ways. No jargon. No gimmicks. Just sector-smart support that moves the needle.