Business intelligence is morphing from a big ticket, long sales cycle, ‘nice to have’ solution for large, data- and cash-rich enterprises into a pervasive technology that can be used advantageously within mid-sized Kiwi organisations by upper-level management, operational managers, production staff, suppliers and customers.
Why? First of all, organisations of all sizes, forced to work smarter, are realising that their database holdings are rich veins of information that can be tapped into quickly, easily and, surprisingly, inexpensively. Simply put, businesses have their BI inputs already in place.
Secondly, processing power and storage are cheaper than ever which means that BI solutions can crunch ridiculous amounts of data in seconds flat in an affordable manner. On top of that, popular database management solutions, such as Microsoft SQL Server, have all of the necessary tools for building multi-dimensional data warehouses, an essential component for BI analysis
Thirdly, BI dashboards are available that can link directly into multi-dimensional data warehouses. With pre-defined KPI displays, data relationships and templates, BI pros can build basic BI dashboards in a matter of hours.
And, finally, the major BI vendors are waking up to the fact that the mid-market represents a huge opportunity for growth and if they don’t offer low-cost high-return solutions they’ll miss the boat.
SAP was the pioneer in the drive towards mid-market BI when it purchased Business Objects in 2007 for US$6.78 billion. Long a by-word for industrial-strength ERP, the acquisition gave SAP an opportunity to tap into markets that were ripe for BI but had nowhere to turn. Now SAP BusinessObjects is firmly entrenched as a BI solution for the mid-market and can be sold successfully by resellers that understand where BI can fit into an organisation’s decision-making processes. Similarly, SQL Server has put the requisite tools in businesses of all sizes. And IBM has Cognos Express, a mid-market version of its enterprise Cognos 10 BI solution.
In short, cost is not an insurmountable barrier to BI implementation. Indeed, the chief reason more organisations don’t employ BI techniques is simply that they don’t know how. So for savvy resellers, there is a huge opportunity.
Defining the business drivers
“Successful BI resellers need to understand the business drivers behind an organisation’s BI initiative,” says Colleen McArthur, partner sales manager at SAP New Zealand. “Every company is different and it is essential that resellers uncover what it is exactly that the organisation wants to know. Who exactly will be using the solution, what data will be included and what type of reports will be required? The reseller has to ask these questions and listen very carefully to the answers. This process feeds directly into the data warehouse structure, the dashboard layout and the report formats and content.”
BI is not an ‘off-the-shelf’ solution that can simply be installed and turned on. The data that feeds into the BI dashboards has to be identified and, to a certain extent, normalised before it can be used successfully. “Mid-market BI still revolves around taking the raw data and moving it into a data warehouse, often multi-dimensional, that has pre-built data relationships,” says Sanjiv Bansal, BI solution manager at SAP Australia. “Resellers with strong vertical market skills, say FMCG [fast moving consumer goods] or distribution, can add value to a BI implementation by helping the end user organisation identify the source data and their inter-relationships. A certain amount can be pre-defined but the final design has to be created as per the end-user requirements. Resellers that specialise can streamline this process and implement BI solutions that deliver real benefits at a reasonable cost.”
Resellers that have particular expertise in ERP and CRM - both data-intensive solutions - have the requisite skills to be successful BI implementation partners. “BI is based on organising and presenting data,” says Scott Tunridge, APAC programme director for BI at Oracle Corporation. “Resellers that know which KPIs are important to organisations can feed them directly into a BI framework. Typically BI multi-dimensional templates have somewhere between 60 percent to 90 percent of the data relationships set and resellers can add value by ‘tweaking’ the feeds into the template based on the particular requirements of the end users.”
Understanding the value proposition
Regardless of how fast or how inexpensively an organisation can implement an viable BI programme, the organisation has to be convinced that, at the conclusion, they will be able to see a quantifiable return on investment. “There are the obvious pay-backs,” continues Turnridge, “such as reductions in the number of hours that it takes to generate reports or budget forecasts. You can add up all of those hours and get a feel for cost savings versus expenditures. The more reports, the greater the savings.”
Similarly, the more people in the organisation who have access to BI dashboards, the greater the return on investment. “Everyone, from upper-level management to end users, can benefit by having access to information generated by BI solutions,” he says. “For instance, telecommunications consumers could compare their monthly spend on various products to help them select the most cost-effective calling plans. The essence of BI is to take action based on the information provided. The more actions, the greater the return.”
Improved decision-making, then, is the real payback on a BI investment. “Every business, regardless of the size, has to make decisions,” says Nick Cater, business analytics software group territory manager for IBM NZ. “Better decisions can be made by processing more information. In fact, we now refer to this whole process as ‘Business Analytics’ (BA). Traditionally, BI has been taking historical data - sales, inventory, balance sheets - and telling decision-makers what has happened. But if we take it one step further we can help people run scenarios to see the effects of a particular decision based on the historical data. Predictive analytics can help organisations avoid making costly errors. Organisations can now tie in more and more related datasets to offer even more accurate analysis of the ramifications of their activities.”
Big BI scared people. Finance balked at the price tag, IT was reluctant to re-design their databases and analysts and users liked BAU (business as usual) processes. It was a hard sell.
However, with most of the pieces in place, resellers can help their clients establish BI in specific areas (a proof of concept) of the business and show concrete positive results. After the first ‘wins’, extending BI across the organisation is much easier. “It’s not unusual for competent BI pros to build draft dashboards in an afternoon,” says Bradley Borrows, Azure and server business group lead at Microsoft New Zealand. “BI doesn’t have to be complicated. If your clients have the data, building intelligent windows into that data isn’t complicated nor does it have to be expensive. When you can show clients the benefits - anything from faster report generation to quicker help desk responses - everyone across the organisation will want access. We’re on the cusp of ‘BI for the masses’.”
Companies want to see a concrete return on investment with BI. “TCO and ROI go hand-in-hand,” says Glen Rabie, CEO of Yellowfin, an Australian-based developer of BI solutions for the mid-market. “TCO depends on the time it takes to implement, the license fees of any software and the cost of staff time required to add the BI factor into existing business processes. The ROI comes with better decision-making.”
“It is essential that the end users can trust the data,” Rabie continues. “This is where an understanding of the underlying data is essential. When you start adding disparate data sets into the mix and combing different classes of data, there is the possibility that errors can creep into the final metrics. It is up to the resellers to ensure that the data is organised in a logical manner and that the data integrity is maintained.”
Is big data another name for BI or a new process all-together? “If it's big data now, what will it be in 2017 when there will be twice as much data?” asks Rabie. “You have to understand the data relationships and logic in order to ensure reliable outputs. If you get the definitions right, you can rely on the results.”
SAP’s HANA (high performance analytic appliance) in-memory computing engine and Oracle's Exadata data-processing machine can now analyse huge amounts of data in milliseconds. These solutions provide a platform for crunching data on-the-fly for even faster results based on more inputs. It is just a matter of time before this type of processing becomes mainstream.
So the data is there, the processing is in place and the need for better decision-making has always been with us. The BI market is primed and ready to take off. All it takes is you, a switched-on reseller.
Business Intelligence: The building blocks
Definition of business processes: What are the business decisions being made and what data is required to support those decisions. An essential step for successful BI implementations.
Data: BI requires massive amounts of data. Most companies already have more than enough up-to-the-minute data for meaningful analysis.
Multi-dimensional data structures: Data warehouses to store data with explicit inter-relationships. Also known as data marts.
Dashboards: Graphical user interfaces into the data mart. Typically displays pre-defined KPIs and supports ad hoc queries, drill-downs and report generation.