Page tree

Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.


Maintaining a contact base of high quality requires time and being strategic with the information that we choose to collect. One of the most common issues is the unintentional creation of duplicates. Even if Secutix has some strategies in place to avoid this, such as forcing a contact search before enabling the creation of a new contact, sometimes it's not possible to avoid it from happening.

This is why our software has a batch aimed to the identification of potential duplicates: You will only have to check the list of potential duplicates and select the ones you want to merge in order to maintain a clean contact database.

...

You can check here how to set it up.


In

First, we are going to check the "Batch of duplicates detection":

...

the Organization context, within the

...

Contact module, you can find the

...

  1. In order to launch it you will have to click on "New" and add a batch of "Duplicates detection" (aka. Détection des doublons).
  2. Once the batch is saved, you can set it up at your convenience. It is not recommended to run it more than once every 24h. Click here to know more about its set up.
  3. Every time the batch is run, an algorithm will check the contact base and identify potential duplicates: these contacts will then appear in orange in the contact search screen.

Indicators taken account by the batch for the identification of duplicates

(lightbulb) Contacts with a zip code

Those contacts that have the following identical elements are considered to be duplicates:

    • The role of the contact.
    • The SOUNDEX evaluation (phonetization feature of ORACLE) of the names and surnames of the individual contacts.
    • The first 35 characters of the standardised address (street number + type of road + street name).
    • The zip code with capital letters.
    • The country.
    • The name of the structure if the contact is associated with a structure (this is an optional element for the definition of the identifier for the duplicate).

(lightbulb) Contacts without a zip code but with a mail address

Those contacts that have the following identical elements are considered to be duplicates:

    • The role of the contact.
    • The SOUNDEX evaluation (phonetization feature of ORACLE) of the names and surnames of the individual contacts.
    • The main mail address (“Mail 1” field).
    • The name of the structure if the contact is associated with a structure (this is an optional element for the definition of the identificatory for the duplicate).

Duplicates control menu: Here you'll be able to search for potential duplicates identified by the system (Duplicate state* = Potential).

Image Added

Among others, you can filter them by probability: You can safely merge any duplicate with a probability of being a duplicate over 80%.


If you click on the number of a line of potential duplicates (if you click elsewhere it's going to open their contact file), this screen will be displayed:Image Added
Here you can

...

choose the information that you wish to

...

keep after the merge

...

or declare a "false positive" (aka. two contacts that are not real duplicates): The buttons "Merge" and "Flag false duplicates" are on the top of the screen.

After the merge, the contact selected as "

...

Top of

...

duplicate" is the one that will remain active

...

. Every element linked to the

...

contact that will be merged will then be attached to the

...

"Top of

...

Duplicates not detected by the batch

The batch might not identify some duplicates because the criterion for the detection does not store any data. However, you as a human can notice a duplicates pair. What can you do?

...

duplicate": sales, cultural, shipment, financial histories; contact connections of all types (structures, individuals, relays...).


Note on the deduplication of structures and relays:

  • If the Duplicates detection batch has been set up to detect structure contact files, the button "Including structure contact" will appear in the Check for duplicates menu.
  • You can merge structure and relays according to the same principle as individuals. However, it is important to take into consideration that relay duplicates are only identified when they are attached to the same structure.
  • When two structures are merged, you will have to wait for the batch to be run again to be able to see the potential relay duplicates resulting from the merger of those structures.


Finally, sometimes you might be able to identify a duplicate that hasn't been detected by the batch. Here's what you can do:

  1. First, in the "Check for duplicates" menu, introduce any criteria to make a search: the button “New” will appear.

      ...

        1. Remark: eventhough no results will come up, go ahead to click "New" to manually find your duplicate contacts.
      1. Click on "New": A pop up screen will be displayed, and you'll be able to:
      2. Search for the "Top of the duplicate" (the one you wish to keep)

      ...

      1. and

      ...

      You could detect the duplicates of the structure type in the same way. You can find more information on this website!

      Old page - English

      Old page - French

      ...

      Eléments pris en compte par le batch pour l'identification des doublons

      (lightbulb) Contacts avec adresse postale

      Sont considérés comme doublons les contacts dont les éléments suivants sont identiques :

      • Le rôle du contact
      • L'évaluation SOUNDEX (fonction de phonétisation Oracle) des noms et prénoms du contact individuel
      • Les 35 premiers caractères de rue (numéro de rue + type de rue + nom de la rue) normalisée
      • Le code postal en majuscules
      • Le pays
      • Le nom de la structure dans le cas où le contact est associé à une structure. (Cet élément est facultatif pour la définition de l'identifiant pour le dédoublonnage.)

      (lightbulb) Contacts sans adresse postale mais avec adresse email

      Sont considérés comme doublons les contacts dont les éléments suivants sont identiques :

      • Le rôle du contact
      • L'évaluation SOUNDEX (fonction de phonétisation Oracle) des noms et prénoms du contact individuel
      • L'adresse email principale (champ email 1)
      • Le nom de la structure dans le cas où une structure est associée au contact. (Cet élément est facultatif pour la définition de l'identifiant pour le dédoublonnage.)

      ...

      1. Search for the "Duplicate" (the one you want to merge).
      2. Click on the "Merge" button at the top of the screen.

      Image Added


      If a contact has been previously marked as a false positive, you will not be able to find it while setting up a new duplicate since the contact is linked already to a previous contact:

      1. Search either under "Potential" or "Incorrect Duplicate" duplicate state:

      Image Added

      2. The previous link will show up as a result, click on "delete" to remove it:

      Image Added

      3. Once this process is completed, you will be able to setup your manual link as explained on the paragraph above from this article.



      To continue with this training unit, click here to learn about the contact file.



      Old page - English

      Old page - French

      ...

      Doublons non-détectés par le batch

      Il arrive que le batch ne détecte pas certains doublons car les critères de détection ne sont pas remplis. Ceci dit, en votre qualité d'humain, vous détectez qu'il s'agit bien d'une paire de doublons. Que faire?

      ...