Management wants to know in 20 minutes how the margin in a specific product segment has performed over the last three quarters. Not next week. Right now. What happens in most controlling teams? Either a frantic search in Excel, a hastily cobbled-together export from the ERP system – or the honest answer: “I won’t be able to tell you until tomorrow.”
Ad hoc analysis is the ability to answer exactly these kinds of questions immediately – not because someone types particularly fast, but because the data structure allows for these questions without anyone having to program a query or build a report beforehand.
This article explains what ad hoc analysis means in controlling, how it differs from standard reporting, and what conditions are necessary for it to actually work.
Want to set up ad hoc analyses in your controlling system?
Whether IBM Planning Analytics is the right fit for your needs can be figured out in a quick call.
What is an ad hoc analysis – and how does it differ from standard reporting?
An ad hoc analysis (Latin for “for this purpose”) is an unplanned, situation-specific data analysis. It arises from a specific question that cannot be answered by existing reports – and it must be answered immediately, not in the next reporting cycle.
The difference from standard reporting is structural:
The key point: Ad hoc analyses are not a special case that the controlling team prepares manually. In modern finance teams, they are self-service – business units ask their own questions. This requires the right data structure, such as that provided by business intelligence systems.
What types of ad hoc requests typically arise in Controlling?
The most common triggers for ad hoc requests in the finance department:
Variance Analyses
“Why is the margin in the Northern Region 3 percentage points below target?” – This question comes up while reviewing the monthly report. It requires an immediate answer, not a new report.
Scenario Questions from Management
“What will happen to our cash flow if we postpone the launch by six weeks?” – A classic ad hoc question that requires a dynamic data structure. It is closely related to rolling forecasts.
Individual Sales Reports
“Show me all customers who haven’t placed an order in more than 90 days and had sales of over 50,000 euros last year.” – A combination of multiple filter levels that a rigid reporting system cannot handle.
Regulatory Ad Hoc Inquiries
Auditors, banks, or internal audit departments sometimes submit detailed requests on short notice that fall outside the normal reporting cycle. Anyone who cannot respond to these within hours risks losing trust.
According to a Gartner study, finance teams spend an average of 40 percent of their working time searching for, preparing, and consolidating data for unplanned requests – rather than on the actual analysis.
What are the technical requirements for true ad hoc capability?
Ad hoc analysis is not a feature you can simply “turn on.” It requires a specific data architecture. Four components are crucial:
- Multidimensional data structure (OLAP cube): Data must be analyzable from any perspective – by region, product, time period, or cost center – without each combination having to be created in advance as a report.
- In-memory technology: Queries across large data sets must run in seconds. Only then is true self-service possible, where the user can iteratively query the data until they have an answer.
- Self-Service Front End: Business units must be able to build queries themselves without any knowledge of SQL. This can be an Excel front end that accesses the database directly, or a web-based interface.
- Controlled Data Access: Ad hoc flexibility and governance are not mutually exclusive. Users see only the data areas for which they have authorization.
How does ad hoc analysis work with IBM Planning Analytics (TM1)?
IBM Planning Analytics is built on OLAP-based ad hoc analysis. At its core – the TM1 in-memory database – is designed to answer multidimensional queries in real time.
In practice, this means:
Controllers open Planning Analytics for Excel and drag data slices directly into their workbook – from the live model, not from an export.
- You can freely combine dimensions: margin by product group, filtered by a specific sales channel, for the last 6 months, compared to the previous year.
- Drill-down to the individual transaction level is possible if the model is structured accordingly.
- Changes in the plan take effect immediately on all dependent reports – no outdated data, no manual refresh.
This is what fundamentally distinguishes Planning Analytics from a purely analytical BI tool: it is bidirectional. Users can not only read data but also write it back – adjusting plan figures and modeling scenarios directly within the same system.
How does that work in practice?
Hays Germany: 25,000 to 35,000 projects – without any IT bottlenecks
Hays Germany previously used a complex, Excel-based planning process in which sales managers were unable to adjust their forecasts on their own. Since implementing IBM Planning Analytics, sales managers have direct, self-service access to their own planning data. The team now handles 25,000 to 35,000 planning projects—and can independently respond to ad hoc requests from senior management.
Swisscom: Real-time calculations for hundreds of concurrent users
Swisscom faced a specific ad hoc problem: driver-based calculations were running too slowly when several hundred users accessed the system at the same time. After migrating to IBM Planning Analytics, calculations now run in real time. As a result, end-of-month reports require 50 percent less preparation time.
How strong is your ad hoc capability today?
In a quick call, we can assess whether your current data structure supports real self-service analysis.
What are some common mistakes made when building ad hoc capabilities?
Mistake 1: Too many predefined reports instead of flexible structures
Teams that respond to every request with a new report end up accumulating a “report graveyard” over the years – dozens of reports, 80 percent of which no one uses anymore. The real problem – a lack of flexibility in the data structure – remains unresolved.
Mistake 2: Running ad hoc queries directly in the ERP system
ERP systems are designed for transaction processing, not for ad hoc analytical queries. A dedicated planning and analysis system is the better solution here.
Mistake 3: Self-Service Without Governance
Ad hoc access without a clear authorization structure leads to users accessing data that is not intended for them. Governance and flexibility must be considered together from the very beginning.
Fact Sheet: Ad Hoc Analysis by the Numbers
- Finance teams spend 40% of their time searching for and preparing unplanned data requests (Gartner)
- Companies with self-service BI reduce the processing time for ad hoc requests by up to 70% (BARC BI Survey)
- Swisscom: 50% less preparation time for monthly reports after implementing Planning Analytics
- IBM Planning Analytics is built on TM1 – an in-memory OLAP database dating back to the 1980s
- Hays Germany: 25,000–35,000 projects in planning, with sales managers having full self-service access
Why BI2run – and what we’ll discuss during the initial consultation
BI2run has completed over 200 TM1 and Planning Analytics projects in the DACH region. In many of these, the catalyst was precisely this problem: a controlling team that spends too much time responding to ad hoc requests because the data structure isn’t designed for that purpose.
What we’ll clarify during our initial consultation: What does your current process for unplanned data requests look like? How long does a typical ad hoc analysis take today? If it becomes clear that a different approach would be better suited, we’ll be upfront about it.
Ad Hoc Analysis · BI Managed Services
Setting up ad hoc analysis – where do we start?
In a 30-minute call, we’ll assess how ready your current data structure is for ad hoc analysis.
Book a call →Glossary: Key Terms
FAQ: Frequently Asked Questions About Ad Hoc Analysis
What is the difference between an ad hoc analysis and a standard report?
A standard report is predefined and always answers the same questions at fixed intervals. An ad hoc analysis answers new, unplanned questions immediately – without anyone having to create a report beforehand.
What tools are suitable for ad hoc analyses in controlling?
That depends on the use case. Power BI or Tableau are suitable for purely analytical evaluations. For controlling teams that want to plan and analyze simultaneously and update planned figures, IBM Planning Analytics is the better solution.
Is Excel sufficient for ad hoc analyses?
For small teams with manageable amounts of data, yes. But as soon as data from multiple sources needs to be consolidated, Excel quickly reaches its limits. Planning Analytics for Excel combines the Excel interface with a powerful database structure.
How can I ensure that ad hoc queries do not result in data breaches?
Through granular permission structures at the dimension and element levels. In IBM Planning Analytics, you can precisely control which users are allowed to access which data areas – down to the level of individual cells.
How does ad hoc analysis differ from AI-driven forecasting?
Ad hoc analysis answers questions about past and current data. AI-driven forecasting predicts future trends. The two complement each other: a well-organized ad hoc structure is the foundation for reliable AI forecasts.
