Article

Address Validation vs. Address Autocomplete: What's the Difference and What Your Online Store Needs

Address autocomplete helps shoppers while they type; validation checks the finished address. Find out when one is enough, when you need both and how to test them.

Address Validation vs. Address Autocomplete: What's the Difference and What Your Online Store Needs

Published on: 19.11.2026 • Written by: Team Launify Reading Time: 5 min

Address Validation vs. Address Autocomplete: What's the Difference and What Your Online Store Needs

A customer types “12 High Street, Leeds” and places the order. But they actually live at number 21 – and two days later the courier is standing outside a stranger's door. This isn't rare: in one round of checkout testing by the Baymard Institute, 9% of manually entered shipping addresses contained a typo or error.

There are two tools for fighting bad addresses, and online stores often confuse them: address autocomplete and address validation. Both rely on an address database, but they step in at different moments and catch different mistakes. This article covers how they differ, where each one has blind spots, ready-to-use form messages and a cheat sheet to help you decide.

Autocomplete helps while typing, validation checks the result

The simplest way to tell them apart is timing:

  1. Address autocomplete works while the customer types. They enter “High S” and see a list of real addresses below the field. They pick one, and street, number, city and postcode fill in at once.
  2. Address validation works after the form is filled in – usually when the customer leaves the field or clicks “Continue”. It compares the address with a database and answers: this is fine, you probably meant this, or we don't recognise this address.
Autocomplete prevents mistakes. Validation catches the ones that slip through anyway. Neither replaces the other.

Where autocomplete has blind spots

Autocomplete speeds up the form and removes most typos in street and city names. But it has three gaps you need to plan for:

  1. The customer ignores the suggestion. According to Baymard, 69% of test participants entered their address manually when it didn't show up in the suggestions. Nobody checks what they type.
  2. They edit the selected address. They choose the right street, then manually change the house number or postcode.
  3. Extra details are missing. Autocomplete doesn't know the flat number, floor or “leave with neighbour” note – and those are exactly the details couriers often lack.

The takeaway: autocomplete is the first layer. If the form stops checking once an address is selected, manually typed and edited addresses go through unnoticed.

What validation actually does and how to set it up

Validation compares the finished address with a reference database built on official address and postal registers. The result has three possible states, and each needs a different response from the form:

  1. Address matches – show nothing and let the customer continue.
  2. Address is close to an existing one – suggest a correction and show what the customer typed next to what you suggest.
  3. Address not found – warn, but let the customer continue. New builds and new streets may not be in the database yet.

Baymard reports that 47% of the sites it benchmarks have no address validator, and recommends showing the original entry next to the suggested correction. Make the warning prominent: participants overlooked subtle inline messages.

Bad vs. good: three real-world situations

Example scenarios comparing a form without checks and a form with them:

  1. Swapped digits. Bad: the order ships to “12 High Street”. Good: autocomplete suggests “21 High Street, Leeds” with the full postcode, the customer selects it and the error never happens.
  2. Postcode typo. Bad: the customer enters a postcode one character off and the parcel goes to the wrong delivery office. Good: validation spots that the postcode doesn't match the street and suggests the right one.
  3. New street. Bad: the form rejects the address and the customer leaves. Good: a message saying “We couldn't find this address” with a “Yes, my address is correct” button – the order goes through and your team just double-checks it.

Form messages you can copy right away

The wording decides whether the customer accepts a correction or gets spooked. Keep it short, blame-free and give a clear choice:

  1. Suggested correction: “Did you mean 21 High Street, Leeds? Use suggested address / Keep the address I entered.”
  2. Address not found: “We can't find this address in our database. If it's correct (for example, a new build), feel free to continue.”
  3. Missing house number: “Please add your house number so the courier finds the right door.”
  4. Missing flat or floor (optional field): “Live in a block of flats? A flat number or the name on the buzzer helps the courier.”

Avoid messages like “Invalid address” with no explanation. The customer doesn't know what's wrong and often just closes the form.

What your store needs: a decision cheat sheet

  1. Most parcels go to pickup points and lockers: the address is mainly for the invoice. Autocomplete is enough; validation is a bonus.
  2. Lots of home delivery by courier: autocomplete and validation. Every bad address means a return or a phone call.
  3. Cash on delivery or expensive, bulky goods: both, plus a mandatory house number. This is where an undelivered parcel costs the most.
  4. Cross-border sales: both, with a database for each target country. Customers easily make mistakes in unfamiliar address formats, and your support team struggles to spot them.
  5. Lots of mobile orders: prioritise autocomplete – it saves typing on small keyboards, where Baymard observed more typos.

Work out how big your problem is. Example calculation: a store ships 2,000 courier parcels a month and 1% are returned or delayed because of the address. That's 20 parcels. If each costs €6 outbound, €6 back and about 15 minutes of support and warehouse time (say €6), that's €18 per parcel and roughly €360 a month – before counting lost customers. Plug in your own numbers from an order export.

How to test both tools in an hour

Before you roll anything out, or when auditing what you already have, run these test addresses on mobile and desktop:

  1. A street name with a common misspelling or missing accents (if you sell abroad).
  2. An address with a building name and flat number.
  3. A rural address with no street name, only a house name or number.
  4. A correct street with the wrong postcode.
  5. An address selected from suggestions, then a manually changed house number.
  6. A completely made-up address – does the order go through, and if so, is it flagged for review?
  7. A foreign address for every country you ship to.

Note what the form did in each test. The most common finding: autocomplete works, but nothing gets checked after a manual edit.

Summary

  1. Autocomplete speeds up the form and prevents typos; validation checks the finished address.
  2. If you deliver to homes, offer cash on delivery or ship abroad, you need both.
  3. Validation should suggest a fix, not block the order.
  4. Write specific messages and always offer “keep the address I entered”.
  5. Test your form with the seven addresses above and calculate what bad addresses cost you.

If you're looking for a ready-made solution that both autocompletes and checks addresses in the checkout form, take a look at Launify – it works as an add-on for common e-commerce platforms.

Related Reading

Delivery 17.11.2026

Czech House Numbers Explained: Why Customers Get Them Wrong and How Your Checkout Form Can Prevent It

A Czech building can have two house numbers, and customers often enter only one. Here is what each number means, the most common mistakes and how to set up your form, with ready-made messages.

Read More
Expansion 12.11.2026

Address Formats in Czechia, Slovakia, Poland and Hungary: What to Watch for in Your Checkout Form

A five-digit postcode rule fails in Hungary, and a Polish parcel without a flat number comes back. Here are the address differences across four countries and validation rules ready for your developer.

Read More
Checkout 10.11.2026

Mobile Checkout: 8 Form Fixes That Lift Conversion on Phones

On a phone the keyboard covers half the screen, so every extra field and every error costs orders. Here are 8 concrete form fixes, with code, error message templates and a 20-minute test.

Read More