Common questions

No. Start at whichever stage fits. If you already have a warehouse, metric definitions or dashboards, we build on what works instead of starting over.

We do. Seth leads every engagement from the first call to the last build, and we take on only a few clients at a time so that stays true. Everything is built in your accounts and documented, so your team can run it after we’re done.

Not always. Many start lightweight and grow into one. The Assessment tells you which you need.

Yes. Focused Projects answer one specific question, like a lead list from your CRM or tying deals back to the campaigns that brought them in. Fixed scope and fee, done in your tools, and you keep the work. Most of our work is larger builds, and if a project shows where one would help, we’ll tell you.

Yes. We start with the customers you already have and find what the best ones have in common. Then we build a list of companies that look like them, from your CRM and outside data, and load it into your CRM with a score so reps start with the best fits. This often works well as a Focused Project.

Yes, and we’ve done a lot of them. Our Data Migration work maps every field from the old system to the new one, cleans up duplicates before anything moves, and tests the loads until counts and totals match. Your history goes into a warehouse too, so reports don’t start over on go-live. We work alongside whoever is setting up the new system.

Whatever fits. We don’t resell software. We work with dbt and Claude or ChatGPT, on the warehouse you already have, like Snowflake, BigQuery, Databricks or Postgres. No warehouse yet? We help you pick one sized to your data.

No. Your business intelligence tool reads the same governed metrics. Need one? We often suggest open source like Metabase or Superset, with no per-seat fees.

Yes. We call it the context layer, and it’s newer work for us. It makes your notes, call transcripts and threads searchable, tagged to a customer and date. We link notes to the metrics they explain so answers can point to them, and we agree with you who can see which notes before anything is connected.

No. AI handles the legwork: pulling numbers, drafting reports, finding context. Judgment still takes people who know your business.

Turn them on. They’re good engines, but they can only be as right as the data and definitions underneath. On a messy foundation they answer faster and more often, and wrong. We build that foundation and test the questions your team actually asks, so whichever AI you use starts from the right data.

When a forecast or score would change a decision and you have the history to back it. Seasonal forecasts need two or more years of clean data, and scoring needs a few hundred past outcomes, like won and lost deals. We compare every model against a simple baseline, like last year’s numbers or your team’s judgment, and we agree a monitoring and retraining plan with you before we hand it over.

It can be, when access is set up with care. The AI works through an account your team creates and can switch off. Before it goes live, we agree what it can reach and set it up so it can only read, never change, your data. Each team gets its own access, so people only see what they’re allowed to (row-level security), and every question it runs is logged where your warehouse supports it. Your data stays inside the accounts you approve: your warehouse and the AI tools you’ve signed off on. More on our security page.

The Assessment usually takes one to two weeks, depending on your company’s size and data maturity. A full build usually takes 6–14 weeks, with something usable in the first few.

Before we build, we list the questions your leadership asks every week. Then we work toward answering them in the tools your team already uses.

Yes. We work with companies nationwide, mostly remotely.

You own the models, metrics and dashboards we build for you, and it all runs in your accounts. Code that comes from our own templates stays ours. You can keep using it in your own business for good, but you can’t sell it or pass it on to another company. If you want us to stay on, the Data Team on Retainer keeps it running, and a Fractional Head of Data can set direction.

A Fractional Head of Data is senior data leadership, part time, focused on strategy and execution: they set the data and AI roadmap, help set priorities, guide your team or vendors, and advise on hiring and tools. The Data Team on Retainer handles upkeep: monitoring pipelines and fixing issues, keeping the warehouse, models and tests current, adding new sources and reports, and retraining models. Pick either one, both, or neither.