Measuring cross-product adoption using dbt_set_similarity | by Matthew Senick | December 2024

For the sake of this example, we will read this csv file as a “seed” model called seed_example within the dbt project.

Now, let's say we want to calculate the jaccard index (cross-adoption) between our document and photo storage products. First, we need to create an array (list) of the users who have the document storage product, along with an array of the users who have the photo storage product. In the second cte we apply the jaccard_coef function of the dbt_set_similarity package to help us easily calculate the jaccard coefficient between the two user ID matrices.

with product_users as (
select
array_agg(user_id) filter (where doc_storage = true)
as doc_storage_users,
array_agg(user_id) filter (where photo_storage = true)
as photo_storage_users
from {{ ref('seed_example') }}
)
select
doc_storage_users,
photo_storage_users,
{{
dbt_set_similarity.jaccard_coef(
'doc_storage_users',
'photo_storage_users'
)
}} as cross_product_jaccard_coef
from product_users

As we can interpret, It appears that just over half (60%) of users who have adopted either product have adopted both. We can graphically verify our result by placing the sets of user IDs in a Venn diagram, where we see that three users have adopted both products, out of five users in total: 3/5 = 0.6.

Using the dbt_set_similarity package, creating segmented jaccard indices for our different user categories should be quite natural. We will follow the same pattern as before, however we will simply group our aggregations based on the user category a user belongs to.

with product_users as (
select
user_category,
array_agg(user_id) filter (where doc_storage = true)
as doc_storage_users,
array_agg(user_id) filter (where photo_storage = true)
as photo_storage_users
from {{ ref('seed_example') }}
group by user_category
)
select
user_category,
doc_storage_users,
photo_storage_users,
{{
dbt_set_similarity.jaccard_coef(
'doc_storage_users',
'photo_storage_users'
)
}} as cross_product_jaccard_coef
from product_users

We can see from the data that among homeowners, cross-product adoption is higher when considering jaccard indices. As shown in the result, all owners who adopted one of the products adopted both. Meanwhile, only a third of tech enthusiasts who have adopted one product have adopted both products. Therefore, in our small data set, cross-product adoption is higher among homeowners than among technology enthusiasts.

We can graphically verify the result by creating the Venn diagram again:

dbt_set_similarity provides a simple and efficient way to calculate cross-product adoption metrics, such as the Jaccard Index, directly within a dbt workflow. By applying this method, Multi-product companies can gain valuable insights into user behavior and adoption patterns across their entire product portfolio.. In our example, we demonstrate the calculation of overall adoption across products, as well as segmented adoption for different user categories.

Using the package for cross-product adoption is just a simple application. In reality, there are countless other potential applications of this technique, for example some areas are:

• Feature Usage Analysis
• Impact analysis of marketing campaigns.
• Support analysis

Besides, This style of analysis is certainly not limited to just SaaS.but it can be applied to virtually any industry. Happy Jaccard-ing!

References

(1) Sean Ellis and Morgan Brown, <a target="_blank" class="af rp" href="https://www.amazon.ca/Hacking-Growth-Fastest-Growing-Companies-Breakout/dp/045149721X” rel=”noopener ugc nofollow” target=”_blank”>Hack growth (2017)

Resources

dbt package center