Explore Snowplow data – Modeled Data
An example of modeled customer data
Below, you’ll find an example of a modeled data table derived from a large ‘atomic’ table – or single source of truth for all the different models. This data can then be used for BI and AI purposes and refreshed as often as necessary.
Data modeling from raw customer data
Once we’ve collected the raw atomic data from a digital source, we can begin to create data models which describe a sample of users’ interactions.
This example is from a cohort of users across the Snowplow marketing website, documentation site and user interface (Snowplow BDP Console).
| session | timestamps | user | app | engagement | first page | referer | marketing | last page | location | ip | useragent | |||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| session_id | session_index | session_start_date | session_start_time | session_end_time | user_id | domain_userid | network_userid | app_id | page_views | conversions | session_length | time_engaged_in_s | first_page_title | first_page_url | first_page_urlpath | first_page_urlquery | page_referrer | refr_urlhost | refr_urlpath | refr_urlquery | refr_medium | refr_source | mkt_medium | mkt_source | mkt_campaign | mkt_clickid | channel | last_page_title | last_page_url | last_page_urlpath | geo_country | geo_region | geo_city | geo_zipcode | geo_timezone | user_ipaddress | br_name | br_version | br_lang | os_name | os_version | device_family | device_class | device_name |
| 289d6851-77ca-45ad-a611-… | 4 | 2020-04-01 | 2020-04-01 10:00:03 | 2020-04-01 10:00:18 | be5016ad-c6ef-442d-aa18-… | 28172b1a-96aa-4fe2-a4e3-… | website | 2 | 0 | 15 | 10 | Snowplow | The Data Coll… |
https://snowplowanalytic… | / | https://www.google.com/ | google.com | / | search | organic search | Snowplow BDP | A Da… |
https://snowplowanalytic… | /snowplow-bdp/ | US | CA | San Francisco | 94102 | America/Los_Angeles | 192.171.81.XX | Chrome | 80 | en-US | Mac OS X | 10 | Other | Desktop | Apple Macintosh | ||||||||
| 01e67be0-600b-4e48-a57e-… | 22 | 2020-04-01 | 2020-04-01 10:00:15 | 2020-04-01 10:01:42 | 292a1aaa-89f9-4441-b001-… | 93252393-6d33-4dab-90b7-… | website | 2 | 0 | 87 | 20 | Snowplow | The Data Coll… |
https://snowplowanalytic… | / | direct | direct | Snowplow Docs – Document… |
https://docs.snowplowana… | /docs/setup-snowplow-on-… | GB | MA | Manchester | M1 | Europe/London | 185.242.7.XX | Chrome | 80 | en-GB | Windows 10 | Other | Desktop | Desktop | ||||||||||||
| 501d77d3-6e2a-4a6e-b66e-… | 1 | 2020-04-01 | 2020-04-01 10:04:33 | 2020-04-01 10:04:38 | 5f1c3cd1-046d-4af3-a8f6-… | 7128ff2f-781f-416c-99e2-… | website | 1 | 0 | 5 | 0 | Snowplow BDP Produc… |
https://snowplowanalytic… | /terms-of-service/snowpl… | https://www.google.com/ | google.com | / | search | organic search | Snowplow BDP Produc… |
https://snowplowanalytic… | /terms-of-service/snowpl… | GB | EN | London | EC3 | Europe/London | 196.245.163.XX | Chrome | 80 | en-GB | Mac OS X | 10 | Other | Desktop | Apple Macintosh | ||||||||
| 7f0a23a5-e392-48d2-b37b-… | 67 | 2020-04-01 | 2020-04-01 10:12:07 | 2020-04-01 10:30:00 | [email protected] | e39fd884-8992-480f-8e37-… | 4b82d683-813b-4969-a4ae-… | console | 2 | 0 | 1073 | 10 | Snowplow BDP | https://console.snowplow… | / | direct | direct | Snowplow BDP | https://console.snowplow… | /4d9aac25-a9f7-46d2-8b4f… | US | NY | New York | 10021 | America/New_York | 138.68.41.XX | Chrome | 76 | en-US | Linux | Other | Desktop | Linux Desktop | |||||||||||
| b23a6994-9520-4347-b1c7-… | 3 | 2020-04-01 | 2020-04-01 10:14:55 | 2020-04-01 10:15:26 | eb8c2f9b-8357-4e01-84e9-… | f4605d71-c3cd-4dc3-962c-… | website | 2 | 0 | 31 | 20 | Building a data quality … |
https://snowplowanalytic… | /blog/2020/02/04/buildin… | utm_campaign=data-qualit… | https://www.google.com/u… | google.com | /url | sa=t&rct=j&q=&esrc=s&sou… | search | paid-search | data-quality | EAIaIQobChMIwPu5t4qs1AIV… | paid search | Snowplow BDP | A Da… |
https://snowplowanalytic… | /snowplow-bdp/ | DE | BE | Berlin | 10365 | Europe/Berlin | 2.205.35.XX | Chrome | 80 | de-DE | Mac OS X | 10 | Other | Desktop | Apple Macintosh | |||
| 694baa7b-67d8-470a-9dc7-… | 14 | 2020-04-01 | 2020-04-01 10:18:21 | 2020-04-01 10:24:06 | [email protected]… | 463cc6f3-c4e2-4aee-b946-… | 192f5fc0-2f41-4fcb-93d9-… | console | 4 | 0 | 345 | 110 | Snowplow BDP | https://console.snowplow… | / | https://snowplowanalytic… | snowplowanalytics.com | / | internal | internal | Snowplow BDP | https://console.snowplow… | /45d4104d-4f5d-4134-8588… | US | CA | Mountain View | 94040 | America/Los_Angeles | 74.125.151.XX | Edge | 81 | en-US | Windows 10 | Other | Desktop | Desktop | ||||||||
| c9a528c2-72ae-458a-9e05-… | 1 | 2020-04-01 | 2020-04-01 10:29:59 | 2020-04-01 10:30:10 | f729daad-549e-4a0a-ae1c-… | c27d4731-1a2f-495b-9615-… | website | 2 | 0 | 11 | 0 | Snowplow | Collect rich … |
https://snowplowanalytic… | /mobile/ | utm_campaign=mobile&utm_… | https://www.linkedin.com… | linkedin.com | / | social | paid-social | mobile | paid social | Snowplow Customers | Her… |
https://snowplowanalytic… | /customers/ | US | TX | Austin | 78702 | America/Chicago | 207.91.133.XX | Mobile Safari UI/WKWebVi… | 13 | en-US | iOS | 13 | iPhone | iPhone | Apple iPhone | |||||
| 86e9281f-8a12-4fa1-b761-… | 5 | 2020-04-01 | 2020-04-01 10:42:28 | 2020-04-01 10:44:32 | be5016ad-c6ef-442d-aa18-… | 28172b1a-96aa-4fe2-a4e3-… | website | 1 | 1 | 124 | 70 | Snowplow BDP | Get … |
https://snowplowanalytic… | /get-started/ | https://snowplowanalytic… | snowplowanalytics.com | /snowplow-bdp/ | internal | internal | Snowplow BDP | Get … |
https://snowplowanalytic… | /get-started/ | US | CA | San Francisco | 94102 | America/Los_Angeles | 192.171.81.XX | Chrome | 80 | en-US | Mac OS X | 10 | Other | Desktop | Apple Macintosh | ||||||||
| db33e52e-9dfd-4d18-b804-… | 6 | 2020-04-01 | 2020-04-02 10:22:04 | 2020-04-02 10:25:44 | 3d16fcf1-9b9d-404e-b014-… | aed54c2d-dbec-4328-86b6-… | website | 3 | 0 | 220 | 60 | Snowplow | The Data Coll… |
https://snowplowanalytic… | / | direct | direct | Introduction to marketin… |
https://snowplowanalytic… | /blog/2020/04/06/introdu… | NL | NH | Amsterdam | 1011 | Europe/Amsterdam | 192.81.220.XX | Chrome Mobile | 80 | nl-NL | Android | 10 | Samsung SM-G975W | Phone | Samsung SM-G975W | |||||||||||
| 042d0e66-f3ea-4d55-b7b4-… | 3 | 2020-04-01 | 2020-04-03 10:35:26 | 2020-04-03 10:39:49 | 91e669dd-7ab2-472b-9557-… | 535179d3-1d4b-4506-b5b6-… | website | 7 | 1 | 263 | 140 | Data modeling 101 | https://snowplowanalytic… | /lp/data-modeling/ | utm_source=google&utm_me… | https://www.google.com/ | google.com | / | search | search | data-modeling | paid search | Snowplow BDP Pricin… |
https://snowplowanalytic… | /pricing/ | US | MA | Boston | 02129 | America/New_York | 73.238.158.XX | Safari | 13 | en-US | Mac OS X | 10 | Other | Desktop | Apple Macintosh | |||||
The advantages of modeling data in a data warehouse
Similar to our out-of-the-box web model, packaged tools such as Google Analytics model your data for you behind the scenes, but the packaged approach has many inherent disadvantages.
Despite the apparent convenience of packaged tools, they often make it challenging to customize data. All Snowplow’s models are fully customizable to make them perfectly relevant for your business model. You might, for example, want to redefine what counts as a session, or how time on page is calculated.
Snowplow also gives you access to the raw data you can run custom analyses for multiple scenarios and use cases. Below we share an example of some custom modeling.
Custom modeling raw customer data
Imagine our marketing team have requested a report about content performance. From the raw data collected through Snowplow BDP, we can build them a custom data set with interesting dimensions around conversion and engagement.
| content | total engagement | avg monthly engagement | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| title | page_urlpath | author | date_published | month_published | year_published | unique_users | sessions | page_views | avg_time_engaged_in_s | avg_scroll_depth | subsequent_demo_requests | conversion_rate | avg_unique_users | avg_sessions | avg_page_views | avg_subsequent_demo_requests |
| Introducing Snowplow Mic… |
/blog/2019/07/17/introdu… | Joshua Brady | 2019-07-17 | 07-2019 | 2019 | 543 | 727 | 6758 | 148 | 77% | 26 | 3.6% | 60 | 81 | 751 | 2.9 |
| Time spent is the most i… |
blog/2019/08/07/time-spe… | Lewis Newman | 2019-08-07 | 08-2019 | 2019 | 389 | 430 | 2552 | 162 | 74% | 28 | 6.5% | 49 | 54 | 319 | 3.5 |
| Re-thinking the structur… | /blog/2020/01/24/re-thin… | Marie Hertzog | 2020-01-24 | 01-2020 | 2020 | 383 | 461 | 2610 | 121 | 82% | 23 | 5.0% | 128 | 154 | 870 | 7.7 |
| Building a data quality … |
/blog/2020/02/04/buildin… | Anna Soininen | 2020-02-04 | 02-2020 | 2020 | 147 | 183 | 649 | 157 | 45% | 8 | 4.4% | 74 | 92 | 325 | 4.0 |
| Understanding the SameSi… | /blog/2020/02/17/underst… | Sean Hutchinson | 2020-02-17 | 02-2020 | 2020 | 276 | 302 | 1466 | 189 | 63% | 7 | 2.3% | 138 | 151 | 733 | 3.5 |
Use cases which are possible with data modeling
This is a simple example of how you can harness the raw, granular data provided by Snowplow. Imagine some other scenarios you may want to model from the same data:
- The Product team want to understand feature engagement – so we build them a model which provides insight into granular use of specific features.
- The Lead Generation team want to understand which of their channels is performing best – so we build them a marketing attribution model that tracks conversions across time per channel.
- The Customer service team wants to create customer service analytics or helpdesk enablement to streamline this pivotal touchpoint.
Improve time to value with out-of-the-box data models
Snowplow is the third most prolific publisher of dbt models in the world. Alongside this large collection of data models for different contexts and industries, we’ve created Data Product Accelerators, step-by-step guides which help you build advanced data products in a fraction of the time.
From the modeled data table above, you can create a user level view of your data across touchpoints – aka a customer 360.
Snowplow has an unparalleled number of fields which come out of the box, as well as prevalidation and a testing sandbox environment for ultimate reliability.
Finally, your Snowplow pipeline can be deployed entirely in your own cloud environment, for a highly-compliant and private solution.