Build vs buy: in-house Snowflake DORA pipeline

The honest TCO comparison. In-house has hidden costs vendor pricing makes explicit.

Oliver Wakefield-Smith

Founder, Digital Signet. Built siemcost / monitoringcost / itbudgetcalculator.

Updated 21 Jun 2026

Direct answer

A self-built Snowflake + dbt DORA pipeline typically costs $120K-$220K per year all-in (one analytics engineer + warehouse credits + ongoing schema work) Snowflake Pricing2026. That crosses Sleuth around 400-500 devs, LinearB around 350-400, and remains cheaper than Jellyfish at almost any scale.

At 100 devs

Sleuth Team ~$30K/year. LinearB Essentials ~$35K/year. Faros CE ~$80K-120K/year. In-house Snowflake ~$120K-180K/year. Buy unless you have a strong data-residency reason.

At 250 devs

Sleuth ~$75K. LinearB ~$87K. In-house ~$150K-200K. Buy still wins unless multi-source requirements break the vendor connector matrix.

At 500 devs

Sleuth ~$150K. LinearB ~$175K. Jellyfish ~$200K-$400K (quote). In-house ~$180K-250K. The decision becomes genuinely close; data control and roadmap autonomy start to matter more than per-dev price.

The hidden costs of in-house

Schema drift across source systems, vendor-API breakage, analyst onboarding, dashboard maintenance, on-call rotation for the pipeline. Budget at least 25% above the base estimate for these. Most build-vs-buy calculations understate this.