How it works

The simple steps to your redirects

Register for Fuzzy SEO

Why we built FuzzySEO

Manually redirecting a websites’ URLs takes time. A lot of time. For an e-commerce website with 5,000 links, this can take the average marketeer 40+ hours to complete.
That's 40 hours of monotonous work which FuzzySEO can do for you in less than a minute.

Simplifying The Redirect Process

Our developers noticed just how long it was taking our marketeers to create a full list of redirects, so they stepped in to help by creating a tool to simplify the redirect process. After witnessing the success of this tool, we have decided to make it available to the public.

FuzzySEO favours URLs which have a lot of unique words, so e-commerce sites with various products are ideal, as are service provides with multiple unique services.

We currently see around 80% URL match accuracy, however we are constantly tweaking things and are striving for 100% accuracy in the near future. That being said, 80% accuracy still saves hours and hours of manual work.

As an example, this FuzzySEO match CSV contains the results for a large e-commerce site which needed to redirect all of its existing product URLs to new, optimised URLs.

Fetch links

You will need data sets for your old URLs and your new URLs, as well as any custom data you want included

Old URLs

First, you must find all of the ‘Old URLs’ you wish to redirect, so this will probably be your existing live website. There are several ways to access these URLs:

  • Through your sitemap.xml file
  • From your CMS database
  • By crawling your website using a web crawler (FuzzySEO will include this feature soon)

New URLs

The second set of URLs you will need are the ‘New URLs’. These are the URLs you would like the old URLs redirected to. All links need the domain included, however this will be removed when matching.

How to structure your data

Now that you have both your Old URLs and your New URLs, you need to structure this data appropriately before running it through FuzzySEO. Both your old URL and new URL datasets should be saved separately in their own CSV file with the following structure:

  1. URL
  2. Custom Data
  3. Metadata

The Custom Data column allows you to add words or data that accompany the URL. For example, internal ID numbers/SKU codes, meta titles or database IDs. This helps to create a more accurate match

The Metadata column is for any additional data you would like passed through to the final results which shouldn’t be searched on to create a match.

Choose settings

Once you have uploaded your CSV files in the appropriate places (there is a place to upload the 'new URLs' and 'old URLs' separately) you need to hit the save button. Both CSV files will then be validated to ensure the layout is correct, so you will be notified if there is a problem.

Once both files are uploaded and validated, you can hit the ‘play’ button on the Match Set page to choose your ‘matching algorithm settings’. We currently offer two options:

Best Match

Our 'Best Match' setting uses a keyword algorithm to match words within the URL which are spelt the same.

The less frequently a word appears within the new URL dataset the more accurate the matches against them will be. For example, the URL /products/bikes could match with the URL /bikes if your new URLs have been simplified to remove the word 'products'.

This is the fastest match available. and once started normally completes in less than 10 seconds for 1,000 urls.

Fuzzy Match

Our 'Fuzzy Match' settings allows for misspelled words to be matched. We allow for a certain percentage of spelling difference between words, so the longer the words are the more of a difference we allow.

For example: Biko would match Bike as it is one change between both words, but Coke wouldn't match as it is two changes.

Canondate would match Cannondale because the word is longer so we allow more changes to occur.

Once started, Fuzzy Match normally completes in less than 2 minutes for 1,000 URLs.

When you have chosen your matching method, click 'Match Links' and your dataset will be queued for processing. We will give you an estimated cost based on the number of links which may have a potential match. You will receive an email once the match set has been completed. This email will also contain the total cost as we will only charge for links that find a match.

your results

When you receive an email telling you that your matching is complete, you are ready to download your results.

Your download will contain all of the 'old' URLs alongside the best matched 'new' URL. If we couldn't find a suitable URL to match, the old URL will appear at the bottom of the list with no new URL counterpart. Any specified metadata accompanying the new/old URL will appear in the additional columns.

Your links are ordered by how accurate the match was, with the most accurate matches appearing at the top of the list and the least accurate at the bottom of the list. The higher your matching score the better the match and we typically find that when the score gets to around 1 or 1.5 the relevance of the results starts to fade. At this point you will need to manually check these matches.