How poor data is costing companies millions

DAR ES SALAAM: EVERY month, Tanzanian companies lose money they never see. Not through theft. Not through fraud. But through poor data. A wrong price in a system. A duplicate customer record.

A loan approved using outdated income data. A warehouse stocked with the wrong product mix. These small errors compound. They distort decisions. They eat into margins. And for many businesses, this silent leak is now one of the biggest threats to profit.

Across Tanzania, firms are collecting more data than ever before. Banks track transactions. Telecom companies record billions of call and usage events. Retailers log sales daily. Government agencies collect tax, identity and licensing records. But collecting data is not the same as using it well.

The real problem is not a lack of data. It is poor data quality and weak analytics. Take the banking sector. In Tanzania, lenders rely heavily on credit history to assess borrowers.

Data from institutions like Creditinfo Tanzania has improved visibility into borrower behaviour. But many banks still struggle with internal data quality.

Customer income records are often outdated. Employment status is not updated on time. Some systems still rely on manual entries from branches. The result is predictable. Good borrowers are rejected because their profiles look risky on paper. At the same time, risky borrowers slip through because their real exposure is hidden across multiple institutions.

This is not just a credit risk problem. It is a revenue problem. When a bank rejects a good borrower, it loses interest income. When it approves a bad one, it risks default. Both outcomes reduce profit. The telecom sector shows a different version of the same problem. Telecom Companies generate massive amounts of customer data every day. Call detail records, mobile money usage, bundle purchases, location patterns.

This data is gold. It can predict churn. It can identify high-value customers. It can optimise pricing. Yet in many cases, the data sits in silos. Marketing teams cannot fully access network data. Finance teams work with separate revenue systems. Customer care systems are not integrated with usage analytics.

A telecom company that fails to identify customers about to churn will spend more on acquiring new customers. Acquiring a new customer is far more expensive than retaining one. A small drop in churn rates can translate into billions of shillings in saved revenue over time. Retail and distribution businesses face another data problem: Inventory mismatch.

Visit many wholesale warehouses in Dar es Salaam and you will find a familiar pattern. Some products sit for months. Others run out too quickly. Managers rely on rough estimates and past habits instead of real-time demand data. This leads to overstocking of slow-moving goods and stockouts of fast-moving ones.

ALSO READ: How PPP dialogue series will mobilise trillions for development plan

The financial impact is direct. Overstock ties up working capital. Stockouts lead to lost sales. Poor demand forecasting also increases transport and storage costs.

In the fuel sector, data errors also carry a cost. Fuel distributors rely on global oil prices, exchange rates and local tax structures to price their products. Small delays in updating pricing models or errors in exchange rate assumptions can wipe out margins for weeks. For companies operating on thin margins, this is dangerous.

Agriculture may be the most affected sector of all. Many agribusinesses still rely on manual records for farmer production, input distribution and crop yields. Without accurate, timely data, they cannot forecast supply. They cannot plan logistics. They cannot negotiate better prices with buyers.

The result is post-harvest loss, price volatility and missed export opportunities. Across all these sectors, the pattern is the same. Poor data leads to poor decisions. Poor decisions lead to lost revenue. The tragedy is that most of these losses are preventable. Companies do not need expensive artificial intelligence to fix this problem. They need strong data foundations. First, data must be accurate.

That means cleaning customer records, standardising formats and removing duplicates. Second, data must be integrated. Sales, finance, operations and customer data should speak to each other. Third, data must be timely. Decisions made on last quarter’s data are often already outdated in a fast-moving market. Fourth, data must be used. Dashboards and reports are not enough.

Managers must be trained to ask the right questions and act on insights. In Tanzania, some companies are already moving in this direction. Banks are investing in better credit scoring models. Telecom firms are building advanced analytics teams. Large retailers are starting to use point-of-sale data to forecast demand. But the gap between leading firms and the rest of the market is still wide.

This gap is creating a new form of inequality in business performance. Companies that use data well are increasing their margins, reducing risk and scaling faster. Those that do not are leaking profit every day without realising it. If you do not know how much bad data is costing you, you are already losing money. The first step is to measure it.

Audit your data. Identify where errors occur. Quantify how those errors affect revenue, costs and risk. The second step is to fix the foundations. Invest in data quality, integration and governance. The third step is to build a culture of data-driven decision-making. Encourage managers to rely on evidence, not assumptions. Tanzania’s economy is growing. Competition is increasing. Margins in many sectors are tightening.

In this environment, companies cannot afford silent profit leaks. Data is no longer just a reporting tool. It is a core business asset. When managed well, it drives growth. When neglected, it drains profit. The companies that understand this will lead Tanzania’s next phase of economic growth. The rest will keep wondering where their money is going.