Query scheduler and progress monitor for marketing data warehouses

A media analyst opens their analytics tool at 8am before a campaign review and kicks off a 90-day attribution query. An hour later it's still spinning. They don't know if it's working, stuck, or silently failed. They retry it, which may queue a duplicate. The meeting starts and they have nothing. This is a daily occurrence for teams running large-date-range queries across any multi-source marketing data warehouse — and the frustration is almost entirely about opacity, not just speed.

The gap persists because the analytics vendors have no structural incentive to surface honest progress information. Showing users that a query will take 40 minutes is better for the user but worse for the vendor's perceived performance. Embedding a transparent progress layer also requires deep integration with the underlying query engine — something a vendor would have to do at the infrastructure level, not just the UI level, which makes it a low-priority engineering investment.

What users actually need is the ability to submit a long-running query, close the browser, get a Slack or email notification when it completes, and come back to a cached result. They also need the system to warn them before running a query that will hit a row limit or timeout threshold — not after it fails silently. None of the tools users complained about do this: they return partial data with no warning, or simply time out.

This is a business rather than a feature because the need recurs every single morning for every analyst running campaign reviews. The cost of a failed or incomplete query isn't just lost time — it's a delayed decision on ad spend. A team spending $500K/month on paid media that can't pull yesterday's attribution data by 9am is making gut-call budget decisions. That's a concrete, recurring, high-stakes cost that justifies a paid standalone tool even if the underlying analytics platform is 'good enough' otherwise.

What to build

Build a query orchestration layer that connects to marketing data warehouses via API, estimates query completion time before execution, breaks oversized requests into paginated chunks automatically, and delivers results to a shared team inbox with Slack and email notifications — so analysts never have to babysit a loading screen.

Where to start

Start with teams already using BigQuery as their warehouse who pipe in Google Ads and Meta data — this combination is common enough to get to 20 customers fast and specific enough to build accurate query time estimates before expanding to other warehouses.

The hard part

Getting the query estimation to be accurate enough to be trusted is genuinely hard — underestimate too often and users stop relying on it, which kills the core value proposition; this requires building warehouse-specific heuristics for BigQuery, Snowflake, and Redshift that behave differently under load.

How it makes money

Monthly subscription per workspace, priced by query volume tiers — starts around $200/month for small teams, scales to $800+/month for agencies or teams running hundreds of queries daily.