Someone in finance asks a question: "What was our revenue in March?"
The number exists. It sits inside a database, somewhere between a sales system, an ERP, and a reporting tool. The company owns the data. The question is clear, and the answer is a single figure. Easy, right?
In most organisations, this question enters a queue as it waits behind other requests. It lands on the desk of someone in the admin, finance or data team who must locate the right source, write a query or export the data, clean the output, and send it back in a format the original person can use.
That process can take a day, sometimes three days. Sometimes even longer, depending on how busy the relevant team is and how many other questions arrived first.
Questions about data are asked constantly in important meetings. A sales director wants to know how pipeline changed this quarter. A marketing lead needs to show campaign return on investment by channel. An operations manager asks why delivery times increased the previous week.
These are not complicated questions but answering them requires technical or analytical skills that most business users do not have, primarily the ability to write database queries or manipulate data effectively and navigate multiple disconnected systems. As a result, the problem is not the people, rather the whole system design. Most organisations structure data access in a way that requires specialists for even the simplest question.
Research from Wren AI found that data analysts in competitive industries such as finance, retail, and technology spend between 50-70% of their time handling ad hoc requests. In quieter environments, that figure still sits around 30-50%. The pattern repeats across sectors: a small group of data “gatekeepers” or specialists fielding a constant stream of one-off queries from every department.
The data or admin team becomes a bottleneck because demand exceeds capacity. Finance waits. Marketing waits. Sales waits. Questions sit in the queue because someone else asked first.
When a question takes three days to answer, decision proceed without educated answers and information.
When data access is slow, people adapt. They estimate and predict. They use outdated reports and make decisions with whatever information they have. Some might say these employees are careless, but it is a reasonable response to a system that cannot deliver answers quickly enough.
Harvard Business School notes that while more than half of people rely on intuition to decide what to believe, decisions based on data allow organisations to verify assumptions before committing resources. Without timely access, those assumptions go untested.
Research from BARC, cited by DigitalOcean, found that businesses making effective use of their data reported an 8% increase in profit and a 10% reduction in costs. Among those companies, 69% noted improved strategic decision-making, 54% reported better operational control, and 52% said they understood their customers more clearly.
McKinsey research indicates that small and medium-sized businesses with strong data practices achieve 6% higher profits than competitors. A European Commission survey found that data-driven strategies could improve performance by up to 15%.
These figures reflect what happens when more people in an organisation can answer their own questions without filing a request and waiting for someone else to process it.
Most organisations have dashboards, reporting platforms, and business intelligence software. The data exists; therefore, storage and cloud space is not an issue.
Dashboards answer the questions someone anticipated when designing them. They show what was deemed important six months ago. But business questions arise in response to what is happening now, and dashboards cannot predict every question in advance.
Ad hoc analysis fills the gap between scheduled reports and unexpected questions. But each request often requires data sets that are not readily available. Gathering, cleaning, and preparing this data takes time, especially when information sits in separate systems with different formats and access permissions.
The analyst becomes a translator. They convert plain-language questions into queries, pull data from multiple sources, clean it, and return it in a usable format. That work is valuable. It also creates a dependency that does not scale when every department has questions and only a few people can answer them. Dashboards describe the past; what organisations need is a way to answer the present.
At the core, every department asks the same thing: “Can I get the number I need, right now?”
And too often, the answer is silence.
A business owner knows their company has data, at least most of them do. And most companies have a lot unused and untouched data. The challenge arises when people who need data-related answers can get them without joining a queue.
The speed at which questions get answered affects the speed at which decisions get made. In most organisations, that speed is limited by how many analysts are available and how many questions are already waiting. This gets compounded as most answers lead to more questions, before acting, slowing the process down further.
For example,
“What was my gross profit last month?”
33%
“What caused it to drop?”
Less billable hours than previous month.
“What caused this?”
Seasonality.
Most dashboards can only answer the gross margin question directly. The other questions require a different approach: you ask, you get a number, then you must think about what that number means and ask again. The real answer comes from combining several numbers with your own reasoning, not from a single data point.
This is the first article in a five-part series on closing the gap between having data and being able to use it. The second article explores what changes when that gap disappears, when anyone in a company can ask a question in plain language and receive an accurate answer quickly.
Xcelerate developed the Xcelerate Business Analyst, a system that connects to your existing data and allows anyone in the company to ask questions in plain English. Instead of submitting a request and waiting days, users type a question and receive an accurate, sourced answer in seconds. The technology reduces ad hoc query requests by 60-80% and cuts decision time from days to a few seconds.
The remaining articles in this series explain how the system works, who uses it, and what it takes to build one.