Data-led Business Intelligence is helping millions of businesses make better data-driven decisions. But it’s only being used by the few rather than the many. If you’re thinking about implementing Business Intelligence in future, what are the challenges you might face when choosing the ideal solution? Here’s what you need to know.
First, pin down the most relevant solution
Get it right and BI generates an ROI of $10.66 for every dollar spent. It’s brilliant. But you can’t make a good buying decision until you’ve carried out a common sense ‘internal needs analysis’, the best way to find the exact right solution for your organisation. If you forget this initial essential step you’ll probably find – as do a disturbing 70% of all BI implementations – that yours will fail, and you’ll miss out on achieving your goals.
Talking about goals, if you don’t already have them, you’ll need to develop some data-led business goals. Your first step is to talk to key stakeholders, including people in IT, finance, marketing, sales and operations. You will need their buy-in early. And that means identifying the key problems they want to resolve.
It’s helpful to create three categories to help set your priorities: must haves, want to haves, and things it’d be nice to have. These will guide the research process, showing vendors of the BI systems you’re interested in exactly what you need. The more work you put in at this stage, the more useful your chosen BI app or software will ultimately be.
Second, justify the expenditure
Companies use BI to improve efficiency and save money. But you probably won’t get approval to invest in it unless you can show there’s a good return on investment to be had. Your Board of Directors or boss will probably want to know the evidence behind the benefits, see them measured quantitatively, so they’re able to put a reasonable timescale on the financial return. The problem is, it’s often tough to prove until you’ve implemented the system and are using it the way it should be used. And that can make a BI system a hard internal sell.
Maybe data reveals better warehouse efficiencies when the warehouse staff take staggered lunch breaks, simply because there’s always someone present. There’s no way to know that without first buying the tool and using it to analyse the data. But a data champion, someone who understands BI back to front and inside out, will be able to create the strong, positive arguments you need to get approval for the spend. A data champion will analyse the business in detail to discover where BI can help most. They’ll help build good relationships between IT experts, executives, business users, and suppliers. They’ll help you develop a powerful strategic vision. And they’ll support you in pinning down the ROI as well as predicting when it’s likely to start flowing in.
Third, employ the right people
You can implement BI systems until you’re blue in the face but if you don’t have the right people on your side, life soon gets difficult. To get the most out of a BI solution, you need the right personnel. People with the data-related skills required, who know how to transform information into insight and recommend the right actions to make it happen. You need someone who can scientifically experiment to solve business problems, who is a maths whiz, and who has really good data literacy, able to spot the real value in data.
Today’s BI tools are simpler and more intuitive than ever. But there remains a serious lack of technical expertise, which means data experts are in huge demand right now, commanding impressive salaries. No wonder so many BI system creators are working hard to change the nature of the partnership between IT and end users in an effort to resolve the skills gap. It’s all about simplification, which has brought about solutions like independent business user-led analytics, with no IT involvement, plus a raft of new IT-managed self-service analytics tools.
Fourth, make sure the data is worth analysing in the first place
You’ve heard the term ‘rubbish in, rubbish out’? The thing is, poor quality data will only deliver bad quality insight, no matter how ‘big’ that data happens to be.
Imagine you have a large Singapore company database to analyse. But the data is all over the place. Some of it is expressed one way, the rest another. Some of the data is input in words, some in numbers, and there’s not a lot of consistency. Some of the records are for customers, some for prospects, and some are dead records, people who haven’t bought from you in years. Before you can draw any valid conclusions from the data and use it to target marketing and sales campaigns it has to be cleaned, tidied, formatted properly and updated. Otherwise any insight you draw from it will be fatally flawed.
When you merge multiple data sources, the end result can be a very poor quality database. If you don’t adhere to strict data hygiene practices, you can fast end up with bad data, data so misleading it can send the business in completely the wrong direction. And when a business makes a bad data-led decision, they can lose everything.
Many businesses get their data from a variety of sources, in different formats. One database might stash phone numbers without gaps, another might include gaps, and something that seemingly simple can be problematic. Then there’s incorrect personal information, inputted wrongly with no checks and balances. Get your prospects’ and customers’ personal details wrong and you can’t fault them for deserting your brand.
Treat big data right and it’ll reward you. Get it wrong and you’ll fail. Make sure you get it right first time!