Meet Priya.
Priya is a senior business analyst at a mid-tier SaaS company. She is competent, career-driven, and knows her job inside out. She has a deep understanding of the business and its needs, and can extract actionable insights from data that drive her company to make smarter decisions.
She is also, more often than not, overwhelmed.
What follows is Priya's Tuesday, told in two different ways. The first is how it goes now; the second is how it would go if her company used Datapane.
Before Datapane
8:45 AM — The first request lands
Priya opens her laptop to check Slack, and sees a message from the VP of Sales, timestamped at 8:12 AM: "Can you please query us Q2 churn by segment? Leadership needs it for the 2 PM meeting."
She opens her data warehouse to find what table she needs to query. Churn data is in one table, segment definitions in another. She already knows that the segment definitions were updated in March, and if she uses the old table, the numbers will not reconcile with the ones Finance showed at the last board meeting. It takes her 20 minutes to verify which table is which.
9:10 AM — The definition problem
In the process of determining which table to use, she finds that one table uses a different definition of a "churned customer" than the other. There are, in fact, 3 different definitions hidden in 3 different tables. One is owned by the data engineering team, one is a relic from a departed analyst and one seems to be a copy of the second one, but with a different date range. She decides to DM the lead data engineer to ask him which one to use.
She'll have to wait until later in the day for his answer.
9:45 AM — Ticket filed, moving on
The data engineering lead has stand-ups until 11, and Priya has another request in the meantime, so she tickets the question to the data engineering team and the finance analyst responsible for churn definitions, asking to clarify the definition of a "churned customer" for the purposes of segmenting. She multitasks while she waits for the ticket response.
11:20 AM — The definition arrives, partially
The data engineering lead gets back to her, saying to use table fct_churn_v2, which he notes does not include the enterprise segment as it has been migrated to a different schema. She'll have to do a union of the enterprise segment table and fct_churn_v2 to get the correct segment breakdown.
After a few minutes of thinking, she opens her SQL editor to write the query.
12:15 PM — The numbers don't reconcile
She has the query result, but it shows that her definition of a "churned customer" leads to a 4% YoY increase in churn for Q2, as compared to the number Finance shared in the deck for the board meeting last week. She is not sure if it is because she uses a different definition, or joined the tables incorrectly, or Finance used a different period for comparison, as they sometimes do. She Googles her question and decides to ask her most trusted finance analyst instead.
Her sandwich is half-eaten when her phone buzzes.
1:30 PM — Close enough
After some back-and-forth with Finance, she establishes that the difference is due to Finance not counting trial conversions as churn, whereas she did. She updates her query accordingly to exclude trial conversions from her analysis, and provides the final number. The discrepancy is now within reason for a special analysis, and she can footnote the methodology difference for reference.
She has 30 minutes to put together a slide with the analysis for the 2 PM meeting.
2:00 PM — The meeting
The VP of Sales asks her a follow-up question at the meeting that she cannot answer on the spot: "Can you please show me the numbers by contract value tiers?" She says she'll follow up by EOD. It is 4:45 PM before she gets back to the VP, by which point the 2 PM meeting has already concluded, and the crucial business decision informed by her research has already been made.
After Datapane
8:45 AM — The same request lands
Priya gets the same message from the VP of Sales as before. She opens Datapane on her laptop and searches for "Q2 churn by segment for leadership review". She adds a note saying it is for the 2 PM meeting.
8:47 AM — The answer
By the time she finishes writing the query, Datapane has retrieved the results. It chose the correct table, fct_churn_v2, and the enterprise segment table in the separate schema, which she assumes it did by joining on customer IDs as explained by the data engineering lead. It reconciled the definitions of a "churned customer" between Finance and Sales, excluding trial conversions, as Finance typically does for board reports. The number it produced has a segment breakdown, and does not differ from the one Finance would have used.
She spends 10 minutes reviewing the trends and adding her commentary, which seem to align with her expectations.
9:00 AM — The slide is built
She has an hour before her 9 AM meeting, and decides to add trend data for the last 4 quarters, to put the Q2 results in context. It is usually the sort of nuance that gets cut when there is not enough time to prepare.
2:00 PM — The meeting
She is asked the same question about contract value tiers as before, but instead of waiting until EOD to follow up, she pulls up Datapane on her laptop to run a new query. It takes 40 seconds to get the result, which she shares with the room via screen-sharing. The business decision is made in real time, on the day it needs to be made.
What Actually Changed
The data in the warehouse was the same. The analyst was the same. The question was the same. The only thing that changed was the layer of automation that understood the business vocabulary, knew which table to pick, recognized the reconciliation logic and contextualized the results for downstream users.
That layer is sorely missing in modern data stacks, with the cost being measured in wasted analyst hours.
Datapane is that layer. It is not a replacement for your warehouse, or your semantic layer, or your analysts. It gives you something that the rest do not: organizational context. It empowers analysts, BI tools, data engineers and end-users to extract maximum value from their data, at the speed of business.
Priya still has a job. It is just finally the job she was hired to do.
