Amazon S3 - How to format and send data to MadKudu via a bucket

Paul Cothenet
  • Updated
Follow

✅  Source  ❌ Destination

 

Pre-requisite

  • You have access to an AWS account to create / manage an S3 bucket
You want to use product usage or CRM data from a source MadKudu does not currently have an integration with? 
No worries, we can easily set up a transfer using Amazon S3 from Redshift or flat files (JSON or CSV). 
 
MadKudu's preferred way is to pull data from your S3 bucket where the data is formatted as described below, and from which MadKudu has access through an IAM role.  Please refer to this documentation to give MadKudu access to your bucket.
You will find below the instructions on:
- how to format the data for MadKudu to be able to ingest it.
- how to format the files
- how to store the file in the bucket
 
For transfer from Redshift, please refer to this documentation.
 
Depending on the volume of data to transfer, it may take from a few hours to a few weeks (>500M records) given the transfer rate limit between Amazon S3 to MadKudu rate limit. We recommend to only send the events necessary to configure MadKudu otherwise your implementation will be delayed until we get the full history of data. Please refer to this documentation or consult our implementation team at success@madkudu.com to understand what are relevant versus irrelevant events to configure your scoring. 

1. How to format your data 

MadKudu works with 3 types of objects:

  • Event: what are users doing?
  • Contact: who is the user? 
  • Account: what accounts my users belong to?
  • Opportunity: what deals do I open? (when you are not using Salesforce Opps or HubSpot Deals)

Depending on what type of segmentation you will want to build (based on behavioral activity, based on lead/account attributes or based on both), you may send us data of 1 or more objects.

  • If you track behavioral activity in a system that does not integrate with MadKudu and would like to build a behavioral segmentation, you would need to send at least Events. If you plan on having account scoring, please send us your Accounts as well. 
  • If you have a homemade CRM or CRM which does not integrate with MadKudu, you would need to send at least Contacts and Opportunities for MadKudu to understand who are your contacts and who converts. If you plan on having account scoring, please send us your Accounts as well. 

 

1.1. Event

To send behavioral data (product usage, web activity, marketing activity...), create a file named event with the following attributes:

Attribute   Format Example Description 
event_key
 required String "abc123" A unique key identifying the event. If you do not have one, we suggest creating a combination of event_text + contact_key + event_timestamp 
event_text
 required String “signup”, “login”, “invited a friend” The action taken by the user. 
event_timestamp
 required Unix time “1436172703” The time at which the event happened 
contact_key
 required String "abc123" or "paul@madkudu.com" The unique identifier of the user who performed the action. This needs to be the same as the contact_key field in the identify file. 
event_*
 optional String or Numeric   properties describing the event (e.g. event_url for the url of visited page, event_form_title for the title of form submitted...)
 

 

Example in JSON format

{"event_key": "abcd1234", "event_text":"signed up", "event_timestamp":1234567890, "contact_key":"abc1234"}
{"event_key": "abcd2345", "event_text":"visit web page", "event_timestamp":1234567890, "contact_key":"paul@madkudu.com", "event_url":"http://www.domain.com/pricing"}

 

1.2. Contact

To send enrichment or CRM data on the contact level (CRM, demographic, firmographic traits ...), create a file named contact with the following attributes:

Attribute   Format Example Description 
contact_key
 required String “abc123”, “paul@madkudu.com” Unique identifier for the user in your database. It can be the email. It must be the same as used in the Event file if Event file sent as well.  
email
 required String "paul@madkudu.com" Email of the user. Pass it even if the contact_key already contains the email 
created_date
 required Unix time 1436172703 Creation date of the contact (required if MadKudu does not have an integration with your CRM) 
contact_*
 optional String or Numeric   Enrichment traits you know about the user (examples: contact_title, contact_country, contact_subscription_plan...)
 

 

Example in JSON format

{"contact_key":"abc1234", "email":"paul@madkudu.com"}
{"contact_key":"432535", "email":"paul@madkudu.com", "contact_title":"cto"}

 

1.3. Account

To send enrichment or CRM data on the account level (CRM, demographic, firmographic traits ...), create a file named account with the following attributes:

Attribute   Format Example Description 
account_key
 required String "def456", “madkudu.com” a unique identifier for the account the user belonged to. It can be the domain of the account 
contact_key
 required if sending Contact data String "abc123", "paul@madkudu.com" the unique identifier of the user who performed the action. This needs to be the same as the contact_key field in the Contact and Event files. 
domain
 required String "madkudu.com" Web domain of the account 
created_date
 required  Unix time 1436172703 Creation date of the account (required if MadKudu does not have an integration with your CRM) 
conversion_date
 optional Unix time 1436172703 Conversion of the account into paying customer. 
ARR
 optional (highly recommended) Numeric $20,000 Annual Recurring Revenue of the account 
account_*
 optional String or Numeric   Enrichment attributes you know about the account (examples: account_industry, account_ARR, account_subscription_plan...)
 

 

Important: Our system supports one account per contact. If there are several, we’ll use the latest one. If you have a use case for having a user belonging to several accounts, we’d love to hear about it. Please let us know at support@madkudu.com.

 

Example in JSON

{"contact_key":"abc1234", "account_key":"madkudu.com", "domain": "madkudu.com", "name": "madkudu"}
{"contact_key":"abc4983", "account_key":"ibm.com", "domain": "ibm.com", "account_ARR":"3000"}

 

1.4. Opportunities

If you send us who are your conversions (it can be the list of your customers), you'll be able to train segmentations based on different types of conversions and have access to Insights reports and Performance report of you segmentations in the MadKudu app. If you don't track opportunities, send us the accounts flagging the ones converted as described above in 1.3 Account.
Otherwise, create a file named opportunity with the following attributes:

Attribute   Format Example Description 
opp_key
 required String "iuh374", “madkudu.com - new biz” a unique identifier for the account the user belonged to. It can be the name of the opportunity 
contact_key
 required if sending Contact data String "abc123", "paul@madkudu.com" the unique identifier of the user who performed the action. This needs to be the same as the contact_key field in the Contact and Event files. 
account_key
 required if sending account data String "def456", “madkudu.com” a unique identifier for the account the user belonged to. It can be the domain of the account 
created_date or close_date
 required  Unix time 1436172703 Creation date of the opportunity or closed date when the opportunity was won (prospect becomes a customer) 
opp_amount
 required Numeric $2,100 the amount of the opportunity 
opp_probability
 optional Numeric  10% the probability of closing the opportunity. This can be replaced by a boolean "has_converted".  
opp_type
 optional String "New business" type of the opp 
opp_*
 optional String or Numeric   Enrichment attributes you know about the opp (examples: opp_subtype, opp_ARR, opp_subscription_plan...)
 
 

 

 

1.5 Specific format for Customer Fit training

If you are not able to send your Opportunities as described above, MadKudu still needs to understand who converts from your historical data to configure a customer fit model. You can send us a unique flat file extracted from your CRM that tells us among your leads who has converted with the following fields:

Attribute   Format Example Description 
email
 required String “paul@madkudu.com” the unique identifier of the user who performed the action 
target
 required Boolean 1 indicate if the lead converted with your conversion definition (Opp created, Opp stage 2…) 
amount
 required Numeric 2,300 if target =1, amount of the opportunity converted (as defined by the conversion definition). 0 otherwise.  
target_closed_won
 required Boolean 1 indicate if the lead converted into a Closed Won opp (paying customer) 
amount_closed_won
 required Numeric 2300 amount generated from the first closed won opp. 
created_date
 optional Unix time   created date of the email 
properties
 optional String or Numeric   Numeric self-input information at time of lead creation or
any other field that you’ve augmented your leads with and want MadKudu to evaluate (example: team_size, industry...) 

 

Example in JSON

{"email":"elon@tesla.com", "target":"1", "amount": "2499", "target_closed_won":"0", "amount_closed_won":"0", "created_date": "1234567890" }
{"email":"paul@madkudu.com", "target":"0", "amount": "0", "target_closed_won":"0", "amount_closed_won":"0", "created_date": "1234567810", "team_size":"5"}

 

2. How to format the files

MadKudu currently supports two file formats:

  • Newline-delimited JSON (preferred)
  • CSV

 

2.1 Newline-delimited JSON

Our preferred format for upload is newline-delimited JSON, which is more standardized and less error-prone than CSV.

In this format, the different records are separated by the newline (\n) character. Each line is a valid JSON object:

{"event_text":"signed up", "event_timestamp":1234567890, "contact_key":"abc1234"}
{"event_text":"added a friend", "event_timestamp":1234567890, "contact_key":"paul@madkudu.com", "some_other_event_field":"some_value"}


{"contact_key":"abc1234", "email":"paul@madkudu.com"}
{"contact_key":"432535", "email":"paul@madkudu.com", "some_other_contact_field":"some value"}
  • escape any double quote "  in your data with a \ (e.g. replace "  with \")

Incorrect

{"event_text":"signed up", "event_timestamp":1234567890, "contact_key":"abc1234", "key": "val"ue"}

Correct

{"event_text":"signed up", "event_timestamp":1234567890, "contact_key":"abc1234", "key": "val\"ue"}

 

2.2 CSV

We also support the .csv format, with the recommended format:

  • column names in the first line
  • separator: ~  -> separate the value with ~ (abc~def~..) Please do not use ,  or -as it easily creates parsing issues
  • delimiter: "    -> this adds quotes around the values (abc -> "abc")
  • line separator: line-break (\n)
  • Escape your " characters by adding a second " character in front of it (see here for details)

Incorrect

abc,cde,efg

Correct

"abc","cde","efg"
  • Remove all line break characters (for example \n) from your fields
  • Make sure the number of fields is the same for each lineIncorrect
"abc","cd"e","efg"

Correct

"abc","cd""e","efg"

 

Using the UTF-8 encoding would be useful to avoid any issues with special characters in the files.

 

2.3 Data validation

JSON line and CSV are relatively easy to corrupt (for example with " or , characters in the data).

We will validate the data on our side and warn you of any corruption issue, but it helps a lot if you follow the format requested above.

 

2.4 Compression

To speed up the data upload part, we highly recommend that you compress your file with GZIP before uploading them to S3.

You can call your file however you want it (we recommend event, contact and account), however please make sure to add the correct extension depending on your file format:

  • .json.gz for compressed JSON (recommended)
  • .json for uncompressed JSON
  • .csv.gz for compressed CSV
  • .csv for uncompressed CSV

 

3. How to store your file

We recommend that the files you want to share with MadKudu are in a dedicated folder and that you create an IAM policy and role for MadKudu to access these files. 

 

You will also need to set up a recurring push of your data to this folder for MadKudu to score fresh data. This is done by creating distinct files, as described below. 

 

3.1 File naming

In the S3 bucket, please upload data into separate folders by date and by objects

{object}/{timestamp} where the objects are 

  • event
  • contact
  • account
  • opportunity

If you use the S3 API, simply “prefix” your destination file name. For example, uploading to "contact/20201120-11:00:00/name_of_file.csv" will add a file name name_of_file.csv to the contact folder.

Please use the recommended file naming system using the object and the date like timestamp_object_name.extension. (e.g. 20201128_account_part_01.json.gz) We’ll automatically import any new file in the folder. You just need to make sure that the right type of files go in the right folder.

 

 

3.2 Compression

To speed up file transfer, you can compress files locally before transferring them to Amazon S3. If you want to compress your files, please use the GZIP compression method and use .gz or .gzip as your file extension (we currently don’t support other methods or other extensions) . 

3.3 Frequency - Setting up a recurring push of data to MadKudu

Depending on how your data will be used in the MadKudu platform, we recommend to provide fresh data through that bucket

  • every 4 hours or at least twice a day if used in a behavioral model 
  • in one shot if used for a punctual model training or analysis

For setting up a recurring push of data, please upload a new file for each batch of new records, naming the files as described in 3.1 File naming. 

Please email support@madkudu.com if you have a doubt, we'll be happy to assist you.

 

Frequently Asked Questions

I'm having an issue with S3 / I don't know how to use S3

Please email support@madkudu.com or contact your Customer Success Manager and we will be happy to assist you.

 

Your file format doesn’t work for me. What do I do?

If you’re having any issue with the file format, please reach out to us at support@madkudu.com and we’ll be happy to help.

 

How often is the data refreshed?

As soon as you drop data to the S3 bucket, expect results to be updated in the Madkudu platform within 6 hours.

 

What would happen if I send the same event more than once - will it appear twice in MadKudu?

Our system will deduce the events based on contact_key / event_text / timestamp. If you send the same event twice, only one will be kept:

  • If sent in two separate batches, only the most recent will be kept.
  • If sent in the same data batch, first one in the file.

 

Can we add other attributes to the Contact records?

Yes, please send any attributes you have stored in your user table (except sensitive ones (password, cc number)).

In particular, it is always helpful to get the following:

  • created_date
  • lead source
  • current plan / value of the plan

 

 

More questions and answers is the F.A.Q as well.

Related articles

Comments

0 comments

Please sign in to leave a comment.

Powered by Zendesk