Why Comparing Three Supplier Quotes Takes Two Days, and How to Fix It

Three suppliers quoted the same job. It took two days to learn which one was cheapest.

Nobody did anything wrong. The quotes simply arrived in three different shapes. One was a PDF with unit prices. One was an email with prices in the body and freight mentioned at the bottom. One was a spreadsheet with bundled items and a different unit of measure.

So a staff member opened a blank sheet and started retyping. Then the questions started. Does the second quote include installation? Is the third price before or after tax? When can the first supplier actually deliver?

Repeat that for every purchase your team handles, and you begin to see where purchasing time really goes.

Where the two days go

Here is a rough picture of one purchase. The numbers are illustrative, so use your own.

• Retyping three quotes into one comparison sheet: 1 hour

• Emails and messages to chase missing details: 1.5 hours

• Checking what each quote includes and excludes: 45 minutes

• Manager review and follow-up questions: 45 minutes

That is about 4 hours of hands-on work per purchase. If your team handles eight purchases like this a month, that is 32 hours, or four full working days, every month.

Waiting is its own cost. While the quotes sit unresolved, the purchase sits too. The branch waits for equipment, the project waits for materials, and an early-payment discount quietly expires. The four hours are visible. The two days are what the business feels.

Notice what is missing from the list. Nobody spent time on the real decision, which is which supplier to trust. Translation took the day.

The slow part is not the math. It is turning three different documents into one comparable shape.

Fix the request, not the reply

Most of the chasing happens because the original request was vague. “Please quote 20 units, delivery by the 30th” leaves every supplier to guess what you care about. Each one answers in its own style and leaves out something different.

The fix is a one-page quote request with the same fields every time: unit price, quantity, freight, installation, payment terms, delivery date, warranty, and how long the price is valid. Suppliers already know these answers. Filling in a form costs them very little.

Suppliers usually welcome this. A clear request means fewer calls back to you, a faster reply from them, and a better chance their quote is read correctly. Clarity helps both sides.

That change alone cuts the chasing. In our example, the 1.5 hours of follow-up drops to about 30 minutes, because most of the questions are already answered.

Some suppliers will ignore your form and send their usual document anyway. That is fine. You still have a standard list of what to look for in it. If a large supplier will not fill in a form, put the same eight questions in your email and ask them to reply under each one.

Where AI earns its place

Now you have a clear list of eight fields, which makes AI far more useful. Give it each quote, ask it to pull the same eight fields into one table, and have it flag anything missing or different, such as “Supplier B excludes installation.”

A person then checks the table against the original quotes. In the example, retyping falls from one hour to about 15 minutes of checking. Reviewing inclusions drops from 45 minutes to 20, because the differences are already marked. The manager’s review drops from 45 minutes to 20, because one clean table replaces three documents.

The new total is about 85 minutes per purchase, down from four hours. Across eight purchases, that is roughly 21 hours a month back, or about 250 hours a year.

The form saves about an hour of that on its own. AI handles the rest, and it works better because the questions are the same every time. A tool fed consistent input makes fewer mistakes than a tool fed a pile of surprises.

Before and after, side by side

• Before: a loose request, three formats, retyping, chasing, and about 4 hours of hands-on time per purchase.

• After: a standard request, one table, flagged differences, and a person checking the result, in about 85 minutes.

• Same decision, made by the same person, a few hours earlier.

Track one more number: the days from request to decision. Hours of work are easy to see, but days of waiting are what delay the project. A shorter chase and a ready table could bring a two-day comparison down to the same afternoon. That is an estimate, so measure it before you promise it to anyone.

And if you never use AI at all, the form still pays for itself. A standard request removes most of the chasing on its own. The tool is a bonus, not the foundation.

What stays with your team

AI can line up three quotes. It cannot know that the cheapest supplier delivered two weeks late last time, or that the second one picks up the phone when something breaks. That memory belongs to your people.

This is the right split. Let the tool handle the reading and sorting. Keep the judgment about trust, risk, and relationships with the person who owns the outcome.

Be clear about who signs off, too. When the table is wrong, the person who approved the purchase still owns the mistake. That is why the human check stays in the process, even when it starts to feel like a formality.

The same problem shows up in many places

Quotes are one example. Vendor invoices arrive in different layouts. Job applications arrive as different file types. Client onboarding forms come back with missing answers. Expense reports, insurance claims, and customer requests follow the same pattern: information arrives in different shapes, and a person spends the day translating it.

AI is a fast translator. A good form is a better one, because it makes the translation unnecessary.

Before you ask a tool to clean up the mess, ask whether you can stop the mess from arriving.

This is also why so many AI trials disappoint. The tool is asked to read whatever shows up, in whatever condition, and then gets blamed for the result. Put a little order at the front door and the same tool looks much smarter.

Try it on one purchase this month

1. Take your last purchase that needed three quotes. List every detail you had to chase or retype.

2. Turn that list into a one-page request with fixed fields.

3. Send it with your next purchase, and time each step before and after.

4. Add AI only for what is still manual, such as reading unusual documents and flagging differences.

Keep the timing simple. A notebook is enough. After two or three purchases you will have your own numbers, and they will be more convincing than any estimate in an article.

If the numbers show little time lost, good. You have confirmed that this process is fine and your attention can go somewhere else. Not every process needs fixing, and knowing which ones do is half the value.

A slow comparison is rarely a math problem. It is a format problem, and a format problem is cheaper to fix than a software problem.

What recurring task in your company still depends on turning information from one shape into another?

If you have a process like that and want to talk it through, I’d be glad to hear your thoughts.

ADDITIONAL READING

• Why Your Next AI Win Doesn’t Need a Developer — https://jordanimutan.com/2026/10/08/why-your-next-ai-win-doesnt-need-a-developer/

• Your AI Rollout Won’t Fix What’s Actually Slowing Your Managers Down. — https://jordanimutan.com/2026/09/16/your-ai-rollout-wont-fix-whats-actually-slowing-your-managers-down/

• You Trained Your Team on AI. Their Output Didn’t Change. — https://jordanimutan.com/2026/09/15/you-trained-your-team-on-ai-their-output-didnt-change/

• Not Every Task Should Use AI. Here’s the Test We Use to Tell. — https://jordanimutan.com/2026/09/25/not-every-task-should-use-ai-heres-the-test-we-use-to-tell/

• Your Company Isn’t Slow—Your Decisions Are Trapped in Manual Processes — https://jordanimutan.com/2025/12/19/your-company-isnt-slow/