505

Get industry leading matching in 3 easy steps

484
Upload your data and tell matchIT whether you want your matching job to match within one file or across two.
485
Check matchIT’s analysis of your data and answer a couple of simple questions about the matching you want to do.
486
Start your job and depending on how many records you uploaded to the queue, download your results in no time at all – a million records can take as little as 5 minutes!

"61% of marketers struggle with duplicate records"

- Royal Mail Research [2017]

So how does matchIT On Demand make matching so simple?

Conventional wisdom is that matching and deduplication is more of an art than a science: good results are more dependent on the skill and experience of the user than the tools that they use.

487
488
matchIT On Demand will prove to you in minutes that effective matching is a science and even a user who hasn’t run any matching before can get the best results!
504
490

Where other Matching Platforms Fail

Most conventional matching software is not designed for contact data but for generic data such as product names/codes, so it allows for keying and reading errors, and mishearing of individual letters – but not phonetic (sounds like) equivalence such as Naughton and Norton or typical inconsistencies in data supply and entry.

Some solutions do allow for obvious inconsistencies – Tony or Anthony, Ltd. or Limited etc. but the example below shows the kinds of error and inconsistency that they can’t handle.

493
Fundamentally, conventional matching solutions require “analytics-ready” data: restructured then standardised using Royal Mail and other reference datasets. Typically, they rely on extended match keys which must be largely the same to find matches.

 

This approach is prone to missing matches and delivering “false positives” – records that are incorrectly identified as matches. Unlike product codes, order numbers etc., contact names and addresses often have several acceptable forms so that standardisation to a single form is not always possible or desirable. What’s more, most people have more than one phone number and many have multiple email addresses, with different numbers/emails being used for differing purposes.

 

This is why solutions that need standardised data and rely on lengthy match keys miss a lot of good matches.

497

How to Enjoy Next-Level Matching with matchIT On Demand

First, select the file to upload – matchIT checks to see if it can figure out the layout automatically for you:

494
In the example above some of the data is in the wrong fields, but matchIT understands that and labels each column according to the most common type of data in it – where needed, it will automatically move data into the correct fields for matching, using a copy of your data.
Next, when you have checked the data has been understood, set the level(s) you want to match at such as person, family, address or company, and how tight you want the matching i.e. how much tolerance do you want to allow for data differences, but still treat the records as matches?
495
Competitors to matchIT will only match at one level at a time, leaving you to spend time rerunning jobs if you want to look for (say) duplicate contacts as well as duplicate companies. matchIT On Demand enables you to select multiple levels of matching in the same job!
It’s ahead of the rest… by a big margin.
matchIT does not rely on long match keys. Instead, matchIT compares all the records that have just one or two things broadly the same, giving each pair of such records a match score: the higher the score, the greater the similarity. All matches scoring above a threshold (higher for tight matching, lower for loose matching) are reported.
matchIT was designed to utilise all the information held about the person or organisation: this includes obvious contact data such as phone number and email address, which it includes in the matching automatically.
496
Above all, matchIT was designed to grade matches effectively so that you can confidently accept the vast majority of matches, leaving very few in a grey area. You can then elect to review these matches manually, ring fence them for later review, or simply accept or reject all in the grey area, depending on the use case – for example:

Use Case 1:
Notification of an interest rate change

Set match threshold to tight matching

Use Case 2:
Suppress customers from a mailing list

Loose matching ensures the offer is not sent to existing customers

Use Case 3:
Create an accurate Single Customer View

Minimise Manual Review of the grey area which can be done as a background task after the higher scoring (tight) matches have been linked

521

Other tools either don't grade matches at all, or the grading is so crude it allows the most obvious matches to be handled automatically – leaving you to choose between reviewing every match and deciding whether to accept or reject each one, or simply to go with just the obvious matches.

This often means letting a very large proportion of the matches go uncorrected or undetected.

And last, when it's all done, download the results…

500
As an example, when the job’s run you can get a file with a match reference appended to each record – sets of matching records have the same reference whereas records that aren’t duplicated will have a unique reference.
499
After finding all the pairs of records that get a high enough match score to be reported, matchIT On Demand groups records together in matching sets – so in the example above, it brings together the 1st and 3rd records (which matched on name and email) with the other records in the set (which matched on name and address), using intelligent bridging logic. matchIT selects one record in each matching set as the “Master” record, which is the record it keeps if you want a deduped file output – typically this is the most complete record.

Read enough? Why not try it for yourself with a free trial account.

Get started