3. Don’t let leadership get distracted by correlations. Many new technologies outperform the ones they’re replacing. But a new product isn’t better because it’s new. The relationship between when a product was launched and how well it works is simply a correlation. Yet in data analytics and other fields, a CEO or senior manager may assume that newer is better and tell you to throw the latest technology at whatever problem you’re trying to solve.
As an IT professional, you can remind leadership that newer technology is not always better in every way (if it was, nobody would ever write with a pencil). In fact, for IT, bugs and compatibility issues often make the latest software a poor choice for many companies. When it comes to seeking improved performance in new technology, for example, you can point out that there are often multiple causal factors — a more powerful chip, faster software, etc. — that may each play a role. Perhaps most importantly, you can guide managers past the “newer equals better” correlation and focus on which factors will impact the problem you’re trying to solve. Benchmark various solutions if time allows, then examine the data to determine which works best. Get your best people to weigh in on the technology in question. In the world of data analytics, for example, an experienced engineer’s approach to processing data can dramatically reduce processing time on existing hardware and software, which may make the big data solution the vendor is pitching you less attractive.
4. Beware of enthusiasm surrounding the next big thing. A senior vice president forwards you a research report claiming that “The Internet of Things Will Thrive by 2025” and asks for your suggestions regarding investments in new IoT technology. But before you start working on your report, consider the emotions that may drive people to put faith in predictions. People don’t like uncertainty. We want a crystal ball — or, at the very least, an article written by experts — telling us what to expect. Predictions, however, typically involve at least some level of uncertainty. If a sunspot fries the electric grid, for example, it may be another 50 years before the IoT is thriving. Of course, as with most data, it’s often important to understand where it comes from. In the case of the IoT study, for example, the claims are from an opt-in survey — which raises a number of questions: Who is being surveyed? What motivations might the respondents have in giving (or not giving) answers? Were the questions structured to allow for the types of nuanced answers that a complex topic such as IoT might spur? The conclusions may very well be valid, but these types of questions should at least give a manager pause and help you temper the initial enthusiasm regarding these types of predictions.
Sign up for CIO Asia eNewsletters.