If you’ve been working in the professional world long enough, you know how valuable data is for every aspect of the business. At the same time, you also probably know how difficult it can be to get the answers you need from a data team with competing priorities. The truth is that while ticketing systems, triage processes, and roadmaps all exist for good reason, these elements can also slow you down when you’re looking for quick answers. With that in mind, we wanted to share five things to keep in mind regardless of your role that will have you getting the answers you need faster than ever!

1) If a number looks wrong, try to tick and tie

Let’s say you’re looking at a dashboard the data team just sent your way. You’re running the marketing campaigns every day, so you’re pretty sure the CAC number isn’t right. Before you send a “this looks wrong” message to your analyst, it will go a long way if you can prove it does not reconcile with a known source and share a little more information.

How to do this in practice:

  1. First, go to the system of record for the metric and find the value you’re more used to seeing
  2. Next, make sure you’re looking at the same window of time as the report in question and that there aren’t any differences in applied filters
  3. If the numbers still definitely do not tie it’s helpful to take a few screenshots but even better if you can export (download) a small sample of the data to share with your team

Once you can pass this info along, not only will you save your data team some time but you’ll show that you’re engaged in solving the problem and willing to get your hands dirty. Trust me, this goes a long way!

2) Write a real data brief, not a vague request

On the theme of being a good partner, lack of specificity when sending a request is the easiest way to get your ticket moved to the back of the queue. A vague “can you pull this” message doesn’t really help anyone. Think about when someone from another team messages you “hey” without any further context … in all likelihood you’re sitting there thinking “what exactly am I supposed to do with this?”

Don’t be that person! Instead, try using this template:

  • Decision: what is the problem we are trying to solve, or target we are trying to hit?
  • Metric: precise definition in one sentence, try to get as specific as possible
  • Grain: What is the granularity you’re looking for here - should the results be summarized by day? By week?
  • Time window: is there an exact date range you’re looking for, or a preset like last 8 complete weeks?
  • Required filters: What should these results be limited to? Any particular channels, regions, product lines, cohorts, etc.?
  • Output shape: Should this be a table or a different type of graphic? How do you want it sorted? Should it include totals?
  • Deadline and Priority: When is this needed by, and how critical is it to operations?

3) Get dangerous in spreadsheets

It’s not sexy, but you can get very far by relying on clean exports and a few moves in Excel or Sheets. If you know your data team doesn’t have the capacity to build out a full dashboard, you can likely unblock yourself by asking for a query with the key data columns you need and doing the rest yourself.

The following are all the ‘advanced’ functions you really need to be dangerous:

  • SUMIFS: Sum a numeric column only where multiple conditions are true
  • COUNTIFS: Count rows only where multiple conditions are true
  • INDEX + MATCH: Flexible lookup that can pull from any column and supports two-way lookups when you nest MATCH for row and column
  • FILTER: Returns only the rows that meet a condition into their own table
  • UNIQUE: De-duplicates a column or whole rows. Handy for quick distinct lists like campaign names or SKUs
  • IFERROR: Wraps a formula to return a blank or message instead of an error. Keeps working sheets clean

A pro tip: Though they’re quite popular, be careful about an over reliance on pivot tables. They tend to obfuscate the data a bit and are not extremely easy to audit or to quickly re-use. Where possible, try building your “pivot” tables from scratch using INDEX MATCH. It will be more difficult at first, but once you get the hang of it you will be thrilled by the flexibility and dynamic nature this adds to your spreadsheets.

4) Learn just enough SQL to verify and explore

SQL seems scary if you’ve never used it before but when you break it down, it’s really just using instructions and parameters to manipulate tables. The goal here isn’t for you to become an analytics engineer, but just to be able to pull and verify things that you need without having to submit a ticket.

Here are a few simple sample queries that you can repurpose for your needs (note, this is in Snowflake SQL syntax):

1. Pulling an aggregate metric over a recent time window

SELECT

  COUNT(*) AS orders_last_7D,

  SUM(TOTAL_SPEND) AS revenue_last_7D

FROM orders

WHERE order_date >= DATEADD('DAY', -7, CURRENT_DATE())

AND order_date <  CURRENT_DATE();

2. Analyzing a simple daily trend

SELECT

  order_date,

  COUNT(*) AS orders

FROM orders

WHERE order_date >= DATEADD('DAY', -14, CURRENT_DATE())

AND order_date <  CURRENT_DATE()

GROUP BY order_date

ORDER BY order_date DESC;

3. Count distinct people / things

SELECT

  COUNT(DISTINCT customer_id) AS unique_customers

FROM orders

WHERE order_date >= DATEADD('DAY', -30, CURRENT_DATE())

AND order_date <  CURRENT_DATE();

5) Own a lightweight metric dictionary

As leaders and team members come and go, definitions drift. When they do, teams argue. A single sheet that memorializes how you think about all the metrics you care about fixes that.

Create one tab with these columns: Metric name, definition in one sentence, grain, system of record (where this metric comes from), and known caveats. Keep it short and readable. Then, when someone asks “how exactly should we be defining an active customer”, you have an answer that makes sense.

Conclusion

Remember - you do not need a course certificate or a new title to work smarter with data. If you can write a clear brief, clean an export, run a couple of checks in Sheets, read a simple query, and point to shared definitions, you’ll get answers faster and make better calls. Start with one active project and try the playbook this week. Share what you find with your analyst as a partner, not a passenger. Do that a few times and you won’t just move more quickly - you’ll raise the bar for everyone around you!

P.S. If you want a one-pager with the request template, the spreadsheet functions, and the SQL snippets in Snowflake syntax, say the word and we will send it over.

Thanks for reading! Want more? Check out our blog and our YouTube channel for deeper dives and walkthroughs. We'll be back each week with more content - subscribe to stay in the loop.

#DataTips #SQL #SouthShoreAnalytics