By Saket Agarwal
In 2012, a leading global financial services company accrued a loss of over $440 million after a glitch in its trading software executed erroneous orders. This software glitch impacted 148 companies listed on the New York Stock Exchange, and the global financial services company almost went bankrupt. Software glitches can be an expensive and often lethal problem to deal with, especially if they extend to an entire system.
As the pandemic drove digital adoption across the world, businesses have accelerated the digital transformation of their services by increasing the online touch points for their users. From partner and/or customer on-boarding processes to verifying transactions, and performing diagnostics, companies run multiple online operations simultaneously. Resilience and efficiency in online operations has now become mission-critical for companies across the world. A minor defect in one section of the platform can affect the experience for all users. Additionally, neglecting any threat in today’s hostile cyber environment can cause a severe disruption to business operations.
Realising that there is almost zero tolerance for poor performance, tech companies are creating a roadmap to enterprise observability to mitigate such risks. But with greater dependence on complex tech infrastructure such as the hybrid cloud, businesses are also confronted with a large volume of organizational data. Achieving enterprise observability is possible only when the observability tech enables them to identify patterns in organizational data that would otherwise be hard to detect, due to the variety and veracity of logged data.
Ninety percent of organisations worldwide claim to have deployed some variation of an enterprise observability solution. In a market anticipated to reach $19 billion capitalisation, this is among the highest adoption rates for any tech solution. The market for log management itself, a critical component of observability, is forecast to reach $4.1 billion by 2026. But there is a lack of knowledge about how to deploy these tools to reap the benefits of enterprise observability.
The primary reason for this is that organizations conflate observability across their applications with having the right insights. This negative correlation can lead to a tunnel vision in analyzing organizational data, since many companies tend to look at such data in silos, instead of looking at it in a holistic manner.
Observability must be seen as a facilitator to achieving key goals, such as making data-backed decisions, shortening feedback loops, and making the company as a whole more responsible. In simpler terms, enterprise observability must offer businesses a more holistic view of end-to-end application performance. Achieving enterprise observability is about reducing complexity and building a unified view of how the tech stack is functioning.
The question now becomes how does one achieve this level of observability? What are some of the critical challenges that a consumer tech company should address in its observability tech, in order to build a safe, smooth, and suitable user experience?
Achieving performance observability
Prioritising your tech’s vulnerabilities is at the core of understanding what is urgent and what is not. A customer entering the wrong password while logging in is not a priority alert. However, a customer entering the wrong CVV while making payments is certainly an alert. Perhaps, the CVV was incorrect because the customer forgot what it was. It could be a breach or a failure to connect with the bank. Customizing your observability tech to accurately define the cause, location, and impact of such alerts will aid in faster analysis, and reveal loopholes and patterns not seen before.
Security observability
From a simple software upgrade to a comprehensive third-party integration, there is potential for security breaches at any point. The security of the infrastructure is directly linked to the performance of the tech and user experience. By closely monitoring alerts such as a sudden spike in server usage, multiple attempts to log in from the same account, an employee accessing unprivileged information, etc., you can identify patterns to limit breaches.
Adhering to compliances
Regulators today are no longer lackadaisical about compliances. As consumer tech companies aim for the public markets, regulators will be closely watching data and security compliances. Added to this is the sheer diversity of compliances. How a consumer tech company stores, processes, and monitors user data in the US will be vastly different than in India and the EU. As we have seen in India, failure to comply with data regulations and the inability to curb glitches on online platforms can lead to the abrupt closure of business operations.
Enterprise observability is like solving Rubik’s Cube – the key is identifying patterns, and gaining actionable insights from the organizational data. Similarly, building a holistic user experience depends on how well a company gains visibility over its data, understands patterns, probes the data, and resolves any potential disruption to other interconnected parts.
The author is founder and CEO, Onnivation