Looker alternative for small teams

Who each one is for

Looker is a platform for organisations that want one governed model of the business. Analysts write LookML, and everyone else explores that model without writing SQL. It is a Google Cloud product, quote-based: a platform fee, per-user licences and an annual commitment.

arcodash is for a small team that already writes SQL. You connect a database, write queries, turn them into charts and dashboards, and put checks and alerts on the same queries. It is hosted only and fixed at $59 or $300 per month for the whole org depending on plan, with unlimited viewers.

If you are a five-person data team at a startup and Looker feels like a platform you would have to staff, that is the gap arcodash is aimed at.

Where Looker is the better choice

Be honest about this part, because it decides most evaluations.

• A governed semantic layer. LookML lets a team define “active customer” once and have every dashboard use it. arcodash has no semantic layer. If metric definitions drifting between teams is your real problem, Looker solves it and arcodash does not.
• Self-serve exploration for non-SQL people. Explores let a marketing manager slice data that analysts have modelled. arcodash is SQL-first. A person who cannot or will not write SQL depends on someone else to write the query.
• Embedded analytics. Looker has a dedicated Embed edition. arcodash has iframe embeds for a dashboard or a single visualization, and an embed can be created as anonymous so signed-out visitors can view it; otherwise the viewer must be a signed-in member of your organization.
• Google Cloud and BigQuery. If your stack is BigQuery and you buy through Google Cloud, Looker is the native choice.
• Scale and permissions. Large organisations with many teams and content-governance requirements are Looker’s home turf.

If two or more of those describe you, stay with Looker or choose it.

Where arcodash fits

• SQL-first. The unit of work is a query in the SQL editor. No modelling layer to learn or maintain before the first chart.
• Checks and alerts built in. A data check is a scheduled SQL query. You write the freshness, volume or validity question as SQL, and a monitor alerts when the result crosses a condition. See data quality monitors. arcodash can also suggest checks for a source.
• Fixed price, unlimited viewers. $59 per month on Team or $300 on Org, for the whole org, however many people open dashboards. You don’t count licence types.
• Small-team setup. 14 days free with no card, then you email to convert.

Alerts, compared concretely

Looker alerts attach to a dashboard tile. Per Google’s docs, they work on query-based or Look-linked tiles, compare a numeric value (greater than, less than, equal to, greater than or equal to, less than or equal to), and add “increases by”, “decreases by” and “changes by” conditions when the query has a date or time field. Notifications go to email or Slack. Check frequency ranges from every 15 minutes to monthly, with a daily default of 05:00. Alerts can’t be set on a Look page directly, only on a dashboard tile. Optional Gemini key driver analysis is off by default.

arcodash alerts attach to a saved query rather than a tile. A monitor reads one column from the first row of the query’s latest result and compares it with a threshold using >, <, >=, <=, == or !=. Or it uses a baseline: it learns the column’s normal range from recent history and flags outliers without a number from you. When the query’s data refreshes, the monitor re-checks and notifies on a state change. Destinations include email, Slack, Teams, webhook, PagerDuty, Mattermost, Discord, Google Chat and Datadog. Alerts can carry a “Why did this fire?” explanation. Details are in the alerts overview and notification destinations.

The practical difference: a Looker alert watches a number someone put on a dashboard. An arcodash check can watch a table that has no dashboard at all, such as “no rows landed in orders for 3 hours”.

Moving over

There is no import tool. LookML models, Explores, Looks and dashboards are not converted. Plan on rewriting each dashboard you still use as SQL, and skip the ones nobody opened last quarter.

A workable order:

1. List the five to ten dashboards people actually use.
2. For each tile, write the query against the warehouse directly. If LookML held a business rule, copy its logic into the SQL or a view.
3. Rebuild the dashboards from those queries.
4. Add freshness and volume checks on the tables behind them.
5. Run both for a couple of weeks, then switch off Looker licences when your contract allows.

If most of your Looker value is in LookML, expect this to be a rewrite of that logic, not a move.

Pricing, plainly

Looker is variable and quote-based, so getting a number means talking to sales. When you do, ask what the platform fee covers, how many Developer, Standard and Viewer licences you need, and what the annual commitment is. arcodash is fixed: $59 or $300 a month for the whole org depending on plan, unlimited seats, no usage billing. The price on the page is the bill. For how per-seat pricing behaves as teams grow, see flat-price BI.

For the other enterprise BI option, see the Tableau comparison.