Cloud BI vs. SaaS
Continuing with our series on frequently asked questions by small and medium-sized businesses about business intelligence, today I’ll address cloud BI and SaaS BI. They’re not actually the same thing, though some of their characteristics overlap.
3) Should we choose an on-premise BI solution or one hosted in the cloud?
This can be an important distinction when choosing your shortlist of vendors; however more and more are offering both on-premise and cloud deployment models, including arcplan. If you’re unsure, shortlisting vendors that offer both is a good plan. That way, you can choose the deployment model after deciding that the product’s features match your needs.
Choosing whether or not to deploy BI in the cloud brings up some special considerations – how concerned about data security you are, where your data will be stored, and whether you have the resources in-house to manage the hardware necessary for your BI deployment.
If data security is a major concern, there are ways to store your corporate data securely behind a firewall and broadcast queries to your cloud BI system. As we often say at arcplan, you don’t necessarily have to move your data to the cloud in order for your BI to be there. That may seem confusing, so let me clarify…
In my last post, I ruminated on the problems that plague business intelligence dashboards. Traditional dashboards simply haven’t evolved as quickly as modern businesses, where decisions are made by many levels of management and need to be informed by various areas of the business – finance, marketing, sales, operations and more. If your dashboard is confined to your department’s metrics alone, it’s doing you a disservice. Dashboards of the future will require cross-functional views of KPIs to really be effective. And they’ll embed more self-service options than ever. The good news is that future-proofing your dashboards is possible. Here’s how:
1) Make your dashboards searchable.
Key performance indicators are essential for monitoring business performance. Last time, I talked about how “KPI overload” – common in so many businesses I’ve worked with – can cloud business insight. However in some cases, having hundreds of KPIs may actually be necessary for understanding and managing your performance. If you need KPIs from many parts of the business in order to make good decisions, it can become a complex task to find the information you need. So consider making your dashboards search-enabled so that users can easily find the metrics they’re looking for. Searching is how users find information; “Google” is a verb for a reason. So why should your BI system be any different?
2) Make your dashboards personalized.
Dashboards serve as a point of reference for business leaders, many times per day or at least a few times per week. Either way, your dashboard should be one of the things you look at the most in a given week (besides your husband, wife, or kids of course). From a personalization perspective, there’s something we can learn from a social medial platform like Pinterest…
Business intelligence dashboards are stuck in the past. It may be controversial to say but it’s true – not much has changed over the past few decades when it comes to BI dashboards. Sure, we’ve defined some decent design rules, some guidelines for developing good, traditional dashboards. But a lot of the dashboards I see as VP of Solutions Delivery at arcplan don’t adequately address the needs of today’s executives, who need cross-functional views of KPIs from various departments in order to make good decisions. We aren’t living in a silo’ world; decision makers need regular access to KPIs from finance, marketing, sales, operations and HR to make good decisions. Do you have a dashboard that enables this kind of view?
If not, let’s examine what’s wrong with your BI dashboards:
1. Their silo’d foundation is a hindrance.
In the early years of data warehouse development, data was stored according to functional areas or departments. Finance, Sales, and Operations each had their own data marts and corresponding dashboards for each department. In today’s business environment, dashboards that are silo’d like your data don’t accommodate your needs. With hundreds or even thousands at KPIs in use at your company – and limited time to access and evaluate them all – a silo’d approach to data access is problematic. You need to access important information at a moment’s notice, not waste time logging into individual systems or viewing separate dashboards to get the data you need to make informed decisions. You’ve probably quickly grown weary of this process and you may even be settling for whatever information is most convenient to use, which – needless to say – is usually not the best option. An ideal dashboard solution bridges multiple information sources to give you a holistic view of the organization – one that matches your role and includes only relevant KPIs.
2. There’s a “Where’s Waldo?” element to finding the right KPIs.
Recently I had a meeting with one of arcplan’s customers in the US, who explained that their arcplan system has grown to manage 4,000 KPIs in just 5 years.
Did you know that only 24% of companies surveyed by BI Scorecard in 2012 consider their BI projects “very successful?” That is pretty sad. And yet Gartner says global BI spending will rise this year to $13.8B and $17.1B by 2016. Companies are still willing to invest massive amounts of money in business intelligence without being able to call these projects successful most of the time.
As a BI vendor, we’d be very concerned if our customers weren’t thinking of their arcplan projects as successful. So we dug through our customers’ most impactful BI projects and identified 5 key factors for long-term BI project success. These factors go beyond the solution-deployment advice you typically see from vendors – steps like “define what success is,” “find an executive sponsor,” and “go with an incremental approach.” While these are critical as well, the factors laid out below embrace 5 different segments of the project: the IT infrastructure, the data, the BI solution itself, the users, and corporate governance.
1. Link your various (clean) source systems
It’s normal for companies to have many information systems in place, from inventory management and ordering programs to ERP and CRM solutions, which may all run on different hardware platforms. The result is a heterogeneous and fragmented IT landscape, and it is often unavoidable. If you cannot reduce the number of systems in place, then ensure that all primary data sources have accurate data whose definition has been agreed upon organization-wide. In addition, be sure your business intelligence solution can serve as an umbrella over all of the systems that need to be linked. BI applications that connect to numerous systems are crucial to the success of projects in heterogeneous environments.
2. Choose your data wisely and compile it intelligently