Six things to look for when choosing an analytics platform
In today’s big data economy, choosing the right analytics platform to make sense of the flood of intelligence flowing through businesses is key to success.
An analytics platform can boost the effective running of any business, not least, by giving some insight into the behaviour of its customers. It’s an invaluable tool, enabling a more data-driven, evidence-based approach to the development of new products and the improvement of existing ones.
Enterprises are increasingly turning to cloud-based versions of these platforms to manage and interpret the mountains of data they collect, as they offer a tantalising duality of pre-configured and flexible options. For many enterprises the flexibility and scalability of cloud platforms opens up new capabilities incomparable to “on-premise” alternatives. Combined with the ability to self-manage or out-source security and maintenance, these solutions can meet the needs of the smallest players through to large conglomerates.
But with dozens of cloud-based analytics platforms on the market, what should data chiefs look out for? The trustworthiness and reliability of the supplier are, of course, predominant factors, but there are a host of other considerations that Chief Data Officers should consider - especially as these platforms tend to have life-spans of over ten years.
The ability to handle large volumes of data is unquestionably paramount to a modern analytics platform. Data as a commodity continues to grow and the ability to ingest today’s plethora of data, as well as plan for ten years’ time is fundamental. This will allow companies to ingest new data, store full data sets and maintain accurate records of data over time, thus making it possible to track the historic trends that make predictive analytics possible with increasingly popular machine learning techniques.
A common practice we hear about across the industry is of analysts running code overnight and over weekends because the volume of data they need to use is too much for the platform they’re using, and so hours, or days, are needed for scripts to run. As data volumes increase, the processing power needs to increase too, and cloud computing facilitates this expansion of capability in a seamless, and seemingly endless, capacity - budget notwithstanding. Accordingly, most cloud-based platforms allow for the purchase of flexible levels of computing power, thus making it possible to find a balance between cost and speed that is more bespoke to the user and a more efficient use of budget compared to other non-cloud-based platforms.
In a business where multiple analysts might be working on large datasets at the same time, processing power can be in high demand and so the ability to process multiple jobs at once is crucial. Many platforms don’t have this ability, but a solution to this is an Hadoop system. This type of system allows for jobs to be shared across a network of cores (computer processing units) at once, allowing them to run in parallel. This considerably reduces the waiting time for analysts.
Ease of use
It is important that an analytics platform is intuitive to use and the data is easy to access. Some vendors will include access to a range of industry-standard tools and processes, allowing users to adopt them much more fluidly, reducing both disruption to the business and the time and cost of training.
While it is important that a platform has the right analytical tools, another essential component is that of documentation software. Enabling processes to be documented as they are developed, within the platform, is key and self-documenting tools can ensure accuracy while reducing the time spent on these tasks by users.
Cloud-based platforms are more flexible compared to their onsite alternatives, they don’t have a large sunk setup cost, and maintenance can be included in the service charge. It is easier, cheaper and faster to modify the processing power, and storage capacity of a cloud-based platform, with price points for the exact needs of a business. This flexibility also allows companies to pay for processing power used, moving away from building platforms that cater for the maximum expected load, but to an elastic usage and pricing model.
The market for analytics platforms is competitive, but when differentiating between technical capabilities becomes challenging, take note of the support and services that vendors offer. As high availability platforms are the new standard, so Service Level Agreements and technical support should have adapted to ensure better technical capability and a faster way of working isn’t impeded by incongruently slow support.
Whilst it might be overlooked - it’s easy to narrow your focus on a checklist of features and functions - the industry know-how of a vendor can also prove to be incredibly valuable. One that understands your business, your customers and the way you work can help you plan and implement data and analytics strategies aligned with your goals, as well as provide tools that meet your functional needs. In addition, vendors that offer services over and above the core platform may be able to provide a more holistic analytics solution, for example, additional data sources or industry expertise that can complement and enhance an organisation’s internal capabilities.
Security and Governance
With Europe adopting GDPR, security and data governance is under more scrutiny than ever before; and nowhere is this more prevalent than in a big data store like your analytics platform. A good platform provider will offer services to manage firewalls and encryption, ensuring that data stored is protected; with governance tools that allow control and monitoring of what data is available, and who has access to it.