Ad Hoc Reporting Requests: How to Cut the Queue So Business Teams Answer Their Own Questions
Why ad hoc reporting requests pile up, and how to cut them: a request form, one monthly report, slicers business users filter, and a rule for the analyst.
- Reporting Workflows
- Self-Service
- Excel

On this page
- Why do ad hoc reporting requests keep piling up?
- What should an ad hoc report request form ask?
- How does one standard monthly report remove repeat requests?
- How can business users filter a report without asking for a new one?
- What should an analyst still own?
- Which tools let business teams answer their own questions?
- Where does Glow Reports fit?
- Common questions
Ad hoc reporting requests pile up for the same reasons in most teams. Sales and finance need a number, the data only lives in exports, and one analyst knows how to clean them. So every question becomes a ticket, and the queue grows faster than anyone can answer it. You cut it in four steps, in this order. A short request form that asks for the decision behind the question. One standard monthly report refreshed from the export. A pivot table with slicers that business users filter themselves. And a written rule for what the analyst still owns. Then pick the tool that fits your team. Everything below works in Excel or Google Sheets.
Why do ad hoc reporting requests keep piling up?
The queue grows because three kinds of work look the same from the outside: one-off questions, the same monthly report rebuilt by hand, and data that only one person can pull from the exports. Each arrives as a new request, so the analyst rebuilds what already exists.
Forum threads describe the pattern well. On Hacker News, a member of a startup's data team described business and product people asking for data in chat all day. The data team ran a quick query, exported a CSV and emailed it, and the person then had to open it in Excel to read a small table Hacker News. In another thread, a commenter described a finance lead calling half an hour before a meeting for a cost figure the BI model had never included, because nobody had asked for it before. The analyst exported the data to Excel and answered with a lookup and a filter, since a change request to IT would take weeks Hacker News. A third commenter said that teaching SQL to product managers didn't work because even a couple of days was too much time for something outside their job, so they kept sending ad hoc requests to the data team Hacker News.
Three causes come back in all of them:
- The question arrives without context. "Can I get revenue by region?" says nothing about the period, the definition of revenue or the decision it feeds.
- The same report gets rebuilt. A request that comes back every month is a report, not an ad hoc question, but nobody turned it into one.
- The data sits in exports. The CRM, the billing tool and the shop each export a file, and only the analyst knows how to clean and join them.
What should an ad hoc report request form ask?
Ask for the decision first, then the figure. A form with five fields makes many requests clearer or unnecessary, because a person who has to name the decision often finds the answer in a report that already exists.
Keep it short enough to fill in under two minutes. Any shared online form works, and so does a table in a shared sheet.
| Field | What to write | Why it helps |
|---|---|---|
| Decision | What you will decide with the answer | Filters out the "nice to know" requests |
| Figure | The number in one sentence, with its definition | Stops "revenue" meaning three different things |
| Period and filters | Dates, region, product, segment | Saves the follow-up email |
| Deadline | When you need it, and what happens if it is late | Lets the analyst sort by real urgency |
| Will you ask again? | Once, every month, every week | Flags the request that should become a report |
Then add two rules. A request with no decision waits behind the ones that have one. And any request marked "every month" goes on the list for the standard report in the next step, instead of being answered by hand again.
How does one standard monthly report remove repeat requests?
Most repeat requests are cuts of the same few figures: revenue, orders, customers, margin, by month, region or product. Build one monthly report that holds those figures, refreshed from the raw export, and those questions stop reaching the queue.
Build it so that next month takes minutes:
- Paste the raw export, untouched, into a sheet called Data and turn it into a table with . New rows then extend the table.
- Build every pivot table and chart from that table, never from a copy.
- Add a sheet called Definitions with one line per figure: what counts as revenue, which date a sale belongs to, what makes a customer active.
- Next month, paste the new export over the old rows and run DataRefresh All.
The detailed layout of such a report is in the monthly sales report in Excel. If the refresh misses new rows, refreshing a pivot table covers the usual causes. And when the report grows to several exports joined by hand, replacing manual Excel reports shows how to automate the steps.
How can business users filter a report without asking for a new one?
Give them slicers. A slicer is a set of buttons on top of a pivot table: click a region or a month and the table filters. Business users answer "the same report, but for France" themselves.
In Excel, click inside the pivot table and select InsertSlicer, then tick the fields to filter by. One slicer can drive several pivot tables that share the same data source through SlicerReport Connections, and the slicer shows which filter is on, so nobody misreads the table Microsoft Support.
In Google Sheets, click the chart or pivot table and select DataAdd a slicer. Sheets allows one slicer per column, and a slicer applies to every chart and pivot table on the sheet that uses the same data range Google Docs Editors Help. The pivot table itself is built with InsertPivot table Google Docs Editors Help.
Pick four or five slicers at most: month, region, product line, sales rep, customer type. More than that and the page turns into a puzzle. To group dates by month inside the pivot, see pivot table by month.
What should an analyst still own?
The analyst owns the definitions, the data preparation and any question that needs something new: a new source, a new field or a new definition. Questions the standard report and its slicers can answer belong to the business team.
Write the rule down and share it:
- Self-serve: any figure already in the standard report, cut by any slicer it offers.
- Ticket: a figure that needs a new export, a join between files, or a definition nobody has written yet.
- Analyst only: changes to the Definitions sheet, and any figure that goes to the board or to a client.
Microsoft describes the same split for Power BI. Its "managed self-service BI" scenario keeps the data architecture with a central team of BI experts, while reporting belongs to creators in departments, an approach it calls "discipline at the core and flexibility at the edge" Microsoft Learn. The central team builds a shared model once, and report creators reuse it instead of building their own. The scale differs from a small team with one workbook. The rule is the same.
Which tools let business teams answer their own questions?
Spreadsheet templates cost nothing and suit a team of a few people. Power BI suits a larger organization with someone to maintain a shared model. AI reporting on the export suits questions the standard report doesn't cover. Most teams end up using two of them.
| Tool | The business user | The analyst | Watch out for |
|---|---|---|---|
| Excel or Sheets template | Filters the standard report with slicers | Pastes the export and refreshes each month | Copies of the file drifting apart |
| Power BI self-service | Builds reports on a shared semantic model | Builds and maintains the model | Licenses and setup time |
| AI reporting on the export | Uploads the export and asks in plain language | Writes the definitions and checks the figures | An AI guessing a definition nobody wrote |
On Power BI licenses, Microsoft says users with a free license can create reports and dashboards for their own use, while a Pro license lets users publish to other workspaces and share content with others Microsoft Learn. So a self-service setup where business users share what they build means paid licenses for them. For a longer comparison of the two approaches, read Power BI vs AI reporting.
AI reporting answers the one-off question no template covers, but only as well as the definitions it gets. Paste the Definitions sheet into the request, and check one total against the export before the answer goes anywhere. The wider list of analytics problems and their fixes is in analytics problems and solutions.
Where does Glow Reports fit?
Glow Reports covers the requests that fall outside the standard report. A business user uploads the CSV or Excel export, or several of them together, and asks the question in plain language. The AI writes the SQL, builds the charts and drafts the text, with no ticket and no wait for the analyst.
The SQL sits under each figure and can be edited. That keeps the analyst's role from the rule above: they read how a number was counted, fix a definition when it is wrong, and the business user asks again. Reports export to PDF and stay in the workspace, so the next person with the same question finds the answer there first. The free plan gives 300 credits once, and a report costs 100 credits.
Common questions
What is ad hoc reporting?
Ad hoc reporting answers one question at a time, outside the regular reports. It covers the "can you pull this for me" requests that come by email or chat. When the same ad hoc request comes back every month, it should become part of a standard report.
Should analysts say no to ad hoc requests?
No. They should sort them. A request with a clear decision and a deadline deserves an answer. A request the standard report already answers gets a link to it, and a request that keeps coming back becomes a new figure in that report.
Do we need Power BI for self-service reporting?
Not to start. A standard report in Excel or Google Sheets with slicers covers most repeat questions for a small team. Power BI makes sense once several departments need the same governed model and someone owns it.
How do you stop the same request from coming back?
Ask "will you need this again?" on the request form. Every request marked monthly or weekly goes into the standard report at the next refresh, with its definition written on the Definitions sheet.
I am a product engineer and the founder of Glow Reports. I spent more than 6 years building products in tech, data and cybersecurity, for large groups, startups and SMEs. In most organizations I saw data treated as a technical topic, when it drives business decisions. I built the reporting tool I was looking for, so the people who run small and mid-sized organizations can read their own numbers without a data team.
More from El Mehdi BenaneUpload your export to Glow Reports, ask the question in plain language and get the report with the SQL behind each figure.


