Could Better Business Visibility Have Mitigated Tech Industry Layoffs?
Another day, another round of tech layoffs. From giants like Alphabet, Amazon, Meta, and Salesforce to startups and scaleups, tens of thousands of workers have recently been dismissed from their positions. Tech layoffs in just the first half of 2023 have already eclipsed the total number of industry job cuts in 2022.
Why is this happening? Today, companies often fall victim to the “tyranny of averages,” relying on past norms to make critical business decisions and plan ahead. In this instance, many layoffs have directly resulted from overhiring based on anticipated growth.
Examining the past and forecasting the future can be useful, but businesses also need a way to stay tuned into the present so they can identify signs of trouble and quickly correct course. Fortunately, recent technological advances make it easier than ever to respond quickly to real-time events — a vital capability to have when the unpredictable occurs.
Operating blind in turbulent times
Layoffs are just one instance of unexpected business disruptions that have become increasingly common. Geopolitical events impacting markets, cyberattacks grinding operations to a halt, and bank/crypto fund shutdowns exposing firms to financial risk are just a few examples of the sudden external forces that can cause distress.
A McKinsey survey on the need for speed in the post-COVID-19 era found that building faster decision-making mechanisms is a top priority for company leaders. “Fast organizations outperform others by a wide margin on a range of outcomes, including profitability, operational resilience, organizational health, and growth.”
Fast and effective decision-making requires complete business visibility or the ability to see and understand the real-time state of the organization. Businesses that don’t prioritize complete visibility are unable to spot outliers and anomalies that spell trouble because they are using the past, old data as their main reference point. To avoid getting blindsided, they need a way to identify the warning signs of a crisis and successfully navigate the crisis once it’s underway.
Rolling the dice on AI as a remedy
Businesses today love pairing their data with AI to predict and, ideally, completely avoid problems. This is great when it works out — but there are some things AI simply cannot predict or cannot predict well. Significant transparency and accountability concerns arise when AI makes autonomous decisions (or influences major decisions) and only a select few people can explain the underlying reasoning.
There is far less interest and investment in identifying and investigating the issues that are happening right now. The romanticization of long-term predictions and underappreciation for real-time visibility and response not only delays the inevitable, it makes things worse. Small unchecked problems can easily snowball into a major disaster when no one is looking.
Predictability and visibility should be used as complementary tactics. Look to the future and consider what might go wrong, but stay tuned into the present and pay attention to the things that are definitely going wrong today and tomorrow. It is a liability for businesses only to be able to react to problems they see coming a quarter in advance.
Thinking and acting in the present
Many companies have good insight into what happened in the past and limited insight into what might happen in the future. But they also have very little awareness of what’s actually happening across the business at any moment.
Streaming data offers a way for companies to better observe the present state of their business so they can spot signs of trouble and adapt to unpredictable events in a more agile way. Streaming data refers to data that is continuously generated by a wide variety of sources across different types and formats, such as telemetry data, geolocation information, IoT device exchanges, and much more.
Traditional data pipelines have a data layer, an analytics layer to provide high-level insights and an application layer that is designed for specific use cases. Because most application services have not changed significantly in several decades, they are unable to handle streaming data.
Fortunately, a new technology — the real-time streaming application — has emerged to extend the streaming data pipeline into the application layer. As a result, businesses can now combine multiple streaming and static data sources into stateful application models that interpret data based on its contextual meaning and run business logic (rules and domain knowledge) against the data.
Using streaming apps to collect, organize, and analyze data provides real-time visibility while enabling immediate action in response to real-world events. For example, one large telecom provider uses real-time streaming applications to create digital twins of high-value entities for a live view of their entire network. This allows them to quickly detect and remediate anomalies — in minutes, not days — before they cause major problems.
Adapting to the unpredictable
Organizations must be able to respond rapidly to issues to thrive during uncertain times. This entails seeing the current state of the business, understanding what it means, and taking action accordingly. Gleaning insights from the past and trying to predict the future can be useful, but they should not be prioritized over remaining alert to important changes in the present.
Streaming data and real-time streaming applications, in particular, provide a way to achieve continuous, correlated analysis of constantly changing data to inform the right corrective actions.
View our traffic demo in action for more information on building streaming data applications.
Recommended ReadingView All Resources
Brokers Aren’t Databases
The rise of event streaming as a new class of enterprise data that demands continuous analysis is uncontroversial. What’s puzzling...
Forget Data – It’s State That Matters
Streaming data contains events that are updates to the states of applications, devices, or infrastructure. When choosing an architecture to...
A Quick Note on Continuous Analysis
Analyzing data on the fly is tricky: Data sets are unbounded and real-time responses demand fast analysis. Incremental algorithms can...