@Sarah - Thanks for including us! Also another corrections regarding Amplitude - all our integrations allow exporting data directly to customer's Redshift/BigQuery/Snowflake WH. We can also ingest with BigQuery and Snowflake. You can read more about how we support both ingesting and exporting data from data warehouse and other cloud storage here - https://www.docs.developers.amplitude.com/data/index.html
Thanks for mentioning, I wasn't sure how supported this feature is because it's in beta. Should this be concerned the status quo for Amplitude moving forward?
The connections to the Datawarehouse themselves are GA eg Snowflake import/Export. There are parts of the UI (where you access the catalog if you have migrated to the new Data UI) that are in Beta, hence the label.
Great article. Do you know any good reading material for understanding GDPR compliances. What data fields can be tracked and what cannot under the compliance rule?
I'm no GDPR expert, but anything that's PII is not GDPR compliant as far as I understand. Additionally, websites using cookies and tracking tools usually must ask the user's consent to use them in a very up-front way. This isn't specific to one tool, but more general to the strategy. If you find any good outlines of GDPR, feel free to follow up and comment!
This is something a few responders have said, so I'm curious as to your perspective. Have you seen companies use 2 or more of these tools in conjunction? What other companies (besides Posthog, Heap) would you wanted to see in the list?
You're not really comparing Snowplow fairly here - Snowplow's job entire job is to collect data, what you do with it after is entirely up to you - print it, model it, bop it, delete it. The other tools all have interfaces, modelling, user management etc etc that is not Snowplow's realm. This is a major distinction that not a lot of people understand (and most businesses aren't kitted out to do).
Saying Snowplow's analysis is limited to errors and counts of events is also misleading - it doesn't do any 'analysis' out of the box, because that's not it's job. (And in reality their error logging is pretty crap - you need to write your own)
Snowplow has a managed service.
Pricing for Snowplow is somewhat externalised - you're paying for engineers, analysts etc to model your data. The actual service is quite cheap when compared to other providers, but not when you factor in the totality of effort needed to model it. That comes with major pros and cons which many companies don't fully appreciate.
Integrations, again, not its job, integrates with most cloud providers or anything that can accept a request.
It's a good call out and something I didn't dive into but completely agree. The one thing I do disagree on is the fair comparison—none of these other tools (namely calling out the larger companies here Amplitude, Segment) actually intake Snowplow events directly. So, if you go with Snowplow, you are signing up for piping those events into a warehouse and can't easily expose the data to product managers without getting a data team involved. If it were me, I would discourage any organization with a minimally staffed data team to go with Snowplow. However, I do think it's a fantastic alternative if cost needs to be distributed through a team as opposed to through a SaaS line item.
This is a great summary, and I really appreciate the insights. I have worked with a couple of these, and I am investigating Snowplow now (mainly due to cost concerns). Am I missing something? From their
EDIT: The below corrections have been made to the original article.
A few corrections that have been pointed out (which is great, because I haven't personally used all the tools on this list!):
- Segment has functions, which can transform data in flight
- Segment has a free tier (1000 visitors/mo)
- Snowplow supports a pixel tracker
- mParticle now owns Indicative and has analytics capabilities
- Posthog, Heap both tools to investigate
- GA4 will be less anonymized
- Amplitude warehouse destinations in beta
I will keep this comment up to date with other additions, and then update the post.
@Sarah - Thanks for including us! Also another corrections regarding Amplitude - all our integrations allow exporting data directly to customer's Redshift/BigQuery/Snowflake WH. We can also ingest with BigQuery and Snowflake. You can read more about how we support both ingesting and exporting data from data warehouse and other cloud storage here - https://www.docs.developers.amplitude.com/data/index.html
Thanks for mentioning, I wasn't sure how supported this feature is because it's in beta. Should this be concerned the status quo for Amplitude moving forward?
The connections to the Datawarehouse themselves are GA eg Snowflake import/Export. There are parts of the UI (where you access the catalog if you have migrated to the new Data UI) that are in Beta, hence the label.
Great article. Do you know any good reading material for understanding GDPR compliances. What data fields can be tracked and what cannot under the compliance rule?
I'm no GDPR expert, but anything that's PII is not GDPR compliant as far as I understand. Additionally, websites using cookies and tracking tools usually must ask the user's consent to use them in a very up-front way. This isn't specific to one tool, but more general to the strategy. If you find any good outlines of GDPR, feel free to follow up and comment!
Thanks for the rough overview. I will share resources if I find any good ones.
Your mix of companies makes the comparison incomplete and not apples to apples
This is something a few responders have said, so I'm curious as to your perspective. Have you seen companies use 2 or more of these tools in conjunction? What other companies (besides Posthog, Heap) would you wanted to see in the list?
You're not really comparing Snowplow fairly here - Snowplow's job entire job is to collect data, what you do with it after is entirely up to you - print it, model it, bop it, delete it. The other tools all have interfaces, modelling, user management etc etc that is not Snowplow's realm. This is a major distinction that not a lot of people understand (and most businesses aren't kitted out to do).
Saying Snowplow's analysis is limited to errors and counts of events is also misleading - it doesn't do any 'analysis' out of the box, because that's not it's job. (And in reality their error logging is pretty crap - you need to write your own)
Snowplow has a managed service.
Pricing for Snowplow is somewhat externalised - you're paying for engineers, analysts etc to model your data. The actual service is quite cheap when compared to other providers, but not when you factor in the totality of effort needed to model it. That comes with major pros and cons which many companies don't fully appreciate.
Integrations, again, not its job, integrates with most cloud providers or anything that can accept a request.
It's a good call out and something I didn't dive into but completely agree. The one thing I do disagree on is the fair comparison—none of these other tools (namely calling out the larger companies here Amplitude, Segment) actually intake Snowplow events directly. So, if you go with Snowplow, you are signing up for piping those events into a warehouse and can't easily expose the data to product managers without getting a data team involved. If it were me, I would discourage any organization with a minimally staffed data team to go with Snowplow. However, I do think it's a fantastic alternative if cost needs to be distributed through a team as opposed to through a SaaS line item.
This is a great summary, and I really appreciate the insights. I have worked with a couple of these, and I am investigating Snowplow now (mainly due to cost concerns). Am I missing something? From their
docs it appears they have robust support for web hooks as well as a pixel tracker. https://docs.snowplow.io/docs/collecting-data/collecting-from-own-applications/pixel-tracker/
https://docs.snowplow.io/docs/collecting-data/collecting-data-from-third-parties/
Thank you for calling this out, I will publish a comment with a few corrections!