Industrial IoT

A Guide to Finding the Right Analytics Tool

6 min readThere’s no such thing as a one-size-fits-all solution when it comes to analytics. [...]

6 min read

There’s no such thing as a one-size-fits-all solution when it comes to analytics.

In the 20th century, the key business intelligence challenge we faced was a shortage of useful empirical data. In the 21st, the issue is having so much it’s hard to know where to start looking. Companies with adequate resources and institutional willpower can throw considerable sums towards establishing advanced-analytics programs. The issue becomes what to do with all of the information being gathered.

In recent years, McKinsey Analytics has conducted extensive research on companies making big investments in advanced-analytics tools, and found that as little as 10% of the potential value these approaches unlock has been realized by the lion’s share of investors. One recurring issue is a consistent failure to tie these technologies to clear, actionable business goals.

In this guide, we’ll explore what kinds of data are most relevant to common business problems, and look at key features offered by various analytics tools that can help you solve them.

Business Intelligence

In order to find the right analytics tool, you need to identify a clear business objective. A bank, for example, might seek better modelling for credit risk assessments, or insights into customer usage patterns. A manufacturer, on the other hand, would prefer BI analytics that help with inventory optimization.

Analytics tool for inventory management

Here are some key considerations:

  • Relevant Metrics: If you try to measure everything, you often end up measuring nothing. Your tool should simplify customizing the metrics you mine and collect. Microsoft’s Power BI – for example – offers industry-specific packages. Their insurance dashboard makes it easy to segment open claims by loss type (collision, property damage, theft etc.), geography and branch performance.
  • Ease of Integration: Analytic insights mean nothing when they’re siloed. Analytics should be integrated at every level of the company so day-to-day operations can become more efficient. Tableau has become a leader in the field thanks to its intuitive collaboration tools, which make it easy to empower key employees to access and contribute to data preparation.
  • Time-to-Insight: Don’t settle for anything less than real-time engagement metrics, and seek products that reduce the time required to generate actionable insights. Domo’s slick platform is an example of how data from disparate streams can be combined in real-time to create revealing visualizations. These make it simple to grasp trends, no matter how far you drill down into the data, and to test out potential solutions.

Web Analytics

Web analytics capture the behavior of users browsing the internet. Thanks to the prevalence of Google’s free Universal Analytics platform, it’s difficult to find a business – from the Fortune 500 down to a rural “mom and pop” feed store – that doesn’t take a great interest in how its web presence is performing.

Analytics tool for Web Analytics

Unfortunately, web analytics is also probably the single biggest source of wasted time and money in the industry, thanks to confusion about how its metrics translate into actionable insights. Are you failing to convert leads because of unintuitive navigation, or because your online marketing isn’t bringing in the right audience?

Google Analytics remains an extremely comprehensive and flexible tool for those with clearly defined business goals and the skill to make use of its reports. It can be augmented with enterprise options (Analytics 360) that provide dedicated support, deeper support for AdWords and search optimization, and training.

The real x-factor in web analytics is integration, which is where third-party solutions distinguish themselves. By linking web metrics to other data streams (including business and marketing intelligence), you can see web analytics within the greater context of your business as a whole.

IBM’s Watson Customer Experience Analytics applies artificial intelligence to automatically identify points of customer “struggle.” Using sophisticated customer behavior modelling, the AI isolates problem areas, allowing you to craft specific fixes. Its insights evolve in real time, helping you to see immediate results.

IoT Analytics

Because the language and functionality of web analytics have become so familiar, there’s a temptation to assume what works for web will apply to other use cases. Beware!

Not long ago, we wrote a piece on the differences between web and IoT analytics, and as a deep-dive we think it’s well worth your time. For now though, we want to summarize a few key points:

  • The variety of use cases in IoT complicates data collection and analysis. Broadly, the types of metrics web analytics tools collect are fixed: no matter how a user interacts with your website, their behavior can be expressed through measurements like time on page, number of clicks and bounce rate. A reasonable evaluation of an IoT connected refrigerator will rely on different factors. Every IoT use case requires its own custom metrics and data modelling.
  • Integrations go far beyond the dashboard. Web analytics are primarily useful when it comes to distribution, engagement and appeal. They’re usually consumed via a dashboard, and their insights mined for marketing. IoT devices – by comparison – are functional, and often those functions must be integrated with many other systems.
    Consider this piece by HIT Infrastructure on the complexities of IoT integration in health care. The real-time information IoT devices offer could be a game-changer for patients. However, their integration with legacy data management systems actually risks slowing down delivery of care.

The initial IoT analytics offerings by big guns like IBM and AWS betrayed a “jack-of-all-trades” mentality. It only had generic dashboards with a handful of preset metrics and reports. If you wanted something customized, you had to look elsewhere.

Those heavy-hitters eventually began to recognize they were losing marketshare to more agile firms like Mnubo, C3, or PTC which provided purpose-built IoT analytics solutions. Over the past year or so, they’ve begun to shift approach. But for companies that have made a serious top-to-bottom commitment to analytics, there are significant time and cost benefits to using platforms designed for IoT analytics rather than imperfect in-house setups.

  • Life-Cycle Analytics: It sounds obvious, but IoT products exist in the world and as such are subject to infinite variables. The best analytics platforms take you from beta testing through obsolescence.
    Mnubo’s SmartObjects system homes in on the type of data you’ll need to collect during beta. It also assists with pre-rollout troubleshooting. During deployment, the focus shifts to offer insights into actual use cases. Over time, you’ll identify usage trends and performance expectations that will inspire the development of future products.
  • Data Enrichment: If your product functions outdoors, you need to factor in weather data. If it’s mobile, geo-location is a must. Seek out platforms that make it easy to integrate these additional data sources with your core metrics. That is the secret weapon to a fully intelligent IoT solution.
  • Full-Stack SaaS Solutions: Take every opportunity to simplify your data management and reduce the burden on in-house IT resources. The ideal analytics tool plays nice with modern Big Data storage architecture. It also offers flexible APIs and allows for custom dashboard creation.

What type of analytics tool do I need?

  1. Determine your business objectives
  2. Look at the type of data you’re gathering
  3. Map out your ecosystem and decide which integrations you need
  4. Choose the analytics tool that provides you with the fastest time-to-insight


In the end, there’s nothing more important in analytics than knowing what you need to know. Generic solutions provide generic answers — because they only allow you to ask generic questions.

The right analytics tool will help guide your discovery process by adapting to the information you want it to measure. And that’s how you arrive at questions worth answering.

Need to talk to one of our data experts to help you make a decision? Click here.