--- title: Optimize data strategies with Snowflake slug: o4oh0190 canonical_url: https://docs.coveo.com/en/o4oh0190/ collection: coveo-analytics source_format: adoc --- # Optimize data strategies with Snowflake When considering data management and analysis, clients face important decisions regarding their data architecture. This article lists the advantages of leveraging a [Snowflake reader account](https://docs.coveo.com/en/l9e90297/), as well as the advantages of using your own Snowflake account in conjunction with [Coveo's Data Share](https://docs.coveo.com/en/m1rf0146/) feature. The article explores each approach in terms of customizability, scalability, cost efficiency, and ease of integration. By examining these features, organizations can better determine the most effective strategy to access and manage their Coveo [data](https://docs.coveo.com/en/259/) to drive informed decision-making. ## Comparing the reader account to a full Snowflake account There are advantages to using either a Snowflake reader account or a full Snowflake account with the Data Share feature. With a reader account, you can directly access your Coveo data in Snowflake without the need to create a Snowflake account. It's a good option if you're looking to only access your data and don't need to perform any write operations. Meanwhile, with the Data Share feature, you can share your data with other Snowflake accounts and have full control over customizability. This feature is particularly useful for large data volumes and when you share data with multiple accounts, such as multiple departments or clients. However, it's important to note that the Data Share feature is only available with a Snowflake account. **Example** As an analytics manager for _Barca Sports_, you've been using a Snowflake reader account to access your Coveo data. You've been able to leverage the account to access your data and perform read operations. You have also run queries and analyzed your data using the account. However, over the past few months, you've observed an increase in data volume and complexity because of a couple of initiatives: * The marketing department has launched a marketing campaign using different channels to attract new customers. Each channel interaction generates data related to customer engagement, conversion rates, and other metrics. * The company introduces a new product line that features a range of products with different attributes that cater to different customer segments. This initiative leads to more complex inventory data (for example, product attributes, stock levels, and sales data). With these changes, you consider getting a Snowflake account to take advantage of the Data Share feature. With a full account, you can create dedicated service accounts, offering tailored access and controls to different teams, thereby streamlining collaboration within _Barca Sports_. If you've been using a reader account and are thinking of switching to a Snowflake account, there are some advantages to consider. The following table provides a rundown of what you can achieve using a reader account and a Snowflake account. [%header,cols="4*<.^"] |=== 2+^|Supported features ^|Reader account ^|Snowflake account - Data Share [.footnote]^[[1](#feature-management)]^ .4+^|Accounts |User accounts ^|[check] ^|[check] |Service/API accounts ^|[x] ^|[check] |Read access ^|[check] ^|[check] |Write access [.footnote]^[[2](#write-access)]^ ^|[x] ^|[check] .2+^|Performance |Warehouse size customization [.footnote]^[[3](#warehouse-site)]^ ^|[x] ^|[check] |CPU processing time [.footnote]^[[4](#cpu-processing)]^ ^|[x] ^|[check] .7+^|Security |User security profiles ^|[x] ^|[check] |Network policy ^|[check] ^|[check] |=== -- 1. With the exception of read access, all features available through a full Snowflake account are managed by the client who is the Snowflake account owner, and not Coveo. 2. The reader account provides views only. 3. The warehouse size customization isn't available with a reader account. 4. The CPU processing time is limited to a fixed allocation of monthly Snowflake credits. --