How to run analytics without a data team
You don't need to hire a data person to get answers from your data. Here's how Tilt runs their entire analytics with zero data hires.
You don't have a data team. You have a Postgres database, maybe a warehouse, and one engineer who everyone asks when they need a number. That engineer has better things to do.
The “can you pull this” problem
At most startups under 50 people, analytics works like this: someone needs a number, they message an engineer, the engineer writes a SQL query, pastes the result in Slack, and goes back to what they were doing. This happens 5-10 times a day.
The engineer loses context on their actual work. The person who asked waits hours or days. And the answer dies in a Slack thread that nobody will ever find again.
The obvious solution is “hire a data person.” But at Series A, you're competing with every other startup for the same small pool of data engineers. The hire takes 3-6 months. And even after you hire them, they'll spend their first quarter building pipelines and dashboards before anyone gets faster answers.
What Tilt did instead
Tilt is a livestream e-commerce marketplace. Balderton-backed, Series A, growing fast. They had the same problem: a small team, no data hire, and everyone bottlenecked on the same engineer for every data question.
Instead of hiring, they connected their Postgres database to Contextflo. It took about 10 minutes. No data model to build, no dashboards to configure, no training for the team. People just started asking questions through Claude.
Founder asks:
“What's our gross merchandise value this week compared to last week?”
Ops lead asks:
“Which sellers had the most returns in the last 30 days?”
Marketing asks:
“How many new buyers came from the Instagram campaign last month?”
Each question gets answered in 2-5 minutes. No engineer involved. No ticket filed. No waiting.
The results
In the first month after connecting:
4,000+
queries from the team
2-5 min
average time to insight
50%
of the team self-serving
0
data hires needed
Tilt scaled through 3 product launches without hiring a data person. The engineer who used to spend half their day pulling numbers went back to building product.
How to do this at your company
You need three things:
1. A database
Postgres, BigQuery, Snowflake, ClickHouse, Redshift, or Databricks. If your product has users and stores data, you already have this.
2. Ten minutes to connect it
Contextflo auto-generates the business context (table relationships, column descriptions, metric definitions) so the AI understands your data without someone manually mapping everything.
3. A team that has questions
No training needed. If someone can type a question in plain English, they can get answers from your data. The founder asking about revenue. The ops lead checking supplier performance. The marketer measuring campaign ROI.
When to actually hire a data person
This doesn't mean you'll never need a data team. When you're processing millions of events, building ML models, or need real-time streaming pipelines, you'll want dedicated data engineers.
But if you're at Series A and your main data challenge is “people can't get answers without asking the engineer,” that's not a hiring problem. That's an access problem. Solve the access problem first. Hire the data team when you actually need data engineering, not when you need someone to run SELECT queries.
Get started
Connect your database and let your team start asking questions. No BI tool to configure. No data model to build. No hire to wait for.
Related reading
- You don't need a BI tool: why dashboards fail small teams
- You don't need a data warehouse for analytics: query Postgres directly with Claude
- What it takes to maintain an agentic analytics stack: the hard parts nobody talks about