Your 10-Step Guide to Creating the Ultimate BI Roadmap

Your business needs an effective data strategy - and this fact should not be news to anyone. In the last few years, technologies such as business intelligence (BI) and big data have risen to the top of every firm's agenda as they look to determine the best way to transform the wealth of information they possess - on their customers, on their competitors and on wider industry trends - into useable insight.

But there's a big difference between knowing you need to do something to take advantage of data, and knowing how to actually go about it. Analyst firm Gartner has proclaimed that as many as 70-80% of BI projects are destined for failure, and this will often be because organizations approach this task without a clearly formulated plan for what they want to achieve, and how they will get to this point.

Therefore, before you embark on any BI strategy, its vital to have a clear roadmap in place to ensure you stay on the right track. By following the steps below, you should be able to develop a coherent plan that can be followed by every individual within your IT team and the wider organization, ensuring you generate useful insights that can be applied directly to the business' decision-making.

1. Understand the scope of the task

The first step to creating a BI plan is to ensure that everyone involved fully understands what you’re looking to achieve and why you’re doing it. This might not be as straightforward as it seems, so it's vital everyone involved is going into the project with the right mindset.

Good BI strategies take a lot or work, will require investment in both people and technology, and will be highly complex operations. It can't be done half-heartedly, or with a 'let's see what happens' attitude, so it's vital everyone in your team is on board and fully committed.

2. Document your current position

The next stage is to establish where you're currently at, what resources you have available, and where the biggest potential areas for improvement lie. This should begin with a thorough audit of the data sources you have available to work with - how it is generated, where it is stored, what format it is in, and how it is processed.

Understanding this will be vital in ensuring BI tools are able to collate, store and analyze data sets effectively, as well as highlighting any gaps where information is missing and questions that can't currently be answered.

3. Identify your key objectives - and how you'll measure them

Initial goals for BI strategies are likely to be fairly vague, perhaps no more than a general idea of being able to better spot emerging trends and opportunities as early as possible. But once you know what data you have, and what you still need, now's the time to start firming up your goals.

The first thing to do is set clear, realistic and practical goals for your BI strategy. As part of this, it's also important to identify several key performance indicators (KPIs) that can be measured to determine whether you're actually meeting these goals. For example, this might be sales revenue, or number of leads generated. There are a wide range of KPIs you can look at, but don't get carried away trying to report on everything at once - focus on a couple of core areas and it will be much easier to see what effect your BI recommendations are having.

4. Get support and stakeholder buy-in

Ensuring other parts of the businesses are bought in to your new BI strategy is key to its success. While your IT team may have faith in the power of your BI tools, if business stakeholders are skeptical of what you're telling them, they'll be unlikely to put recommendations into practice.

It will be hugely helpful in this regard to have a supporter or sponsor from the board level to back your strategy. C-suite executives set the tone for the business, and if you can get them on board, this will help ensure other departments have confidence in following through on your conclusions.

5. Put business needs ahead of data

You don't want to be coming to business stakeholders with recommendations that feel irrelevant to them, or don’t reflect the practical realities of how they operate. This will undermine trust and give the impression that you're not connected with what the company actually needs.

Therefore, talk to business stakeholders to establish what questions they would like to answer and where there are gaps in their knowledge. BI strategies that focus solely on the data itself and don’t consider how findings will actually help business stakeholders will be far more likely to fail, so ensure you're always putting the needs of the wider organization ahead of any interesting but irrelevant data analytics.

6. Define the language you'll use

When presenting your findings at the end of a BI project, it will be important to ensure everyone is on the same page (more on that later). Therefore, it's important to ensure everyone is working off the same 'data dictionary' that defines exactly what each data point is and how it should be interpreted.

For example, if finance and sales teams define terms like 'gross margin' differently, their numbers won’t match. Make sure this doesn't happen by putting together a comprehensive business glossary that includes a description of every term you'll be using, to ensure you're all speaking the same language.

7. Select the right solution

Once you're ready to actually start analyzing your data, you'll need the right technology tools to do so. It may seem odd to wait until so far down the road to determine this, but by taking the time to do the right prep work first, you'll be able to select a solution that best meets your needs, rather than deciding on your technology path first, then having to reshape your processes.

Factors to consider include what skills you have within your organization, what scalability demands will be required, and whether cloud or on-premise tools will be better suited to your way of working. Put together a checklist of what you'll need before evaluating options and remember, just because a tool is the biggest or offers the most features, it won't necessarily be the right fit for your organization.

8. Make sure you're using the right data

Next, you need to look at the data itself to ensure it's fit for your needs. Poor-quality data can lead to any results you get from your analytics being skewed, rendering them worthless. Therefore, it's essential to cleanse your data to ensure its as accurate as possible.

Duplicated data, inconsistent formatting, missing fields and details that are just plain wrong will all mean your results can't be trusted, so getting this sorted before you start processing will ensure you're achieving the best possible results. Use automated data quality tools to fasttrack these tedious manual processes.

9. Communicate effectively

Once you've run your analytics and reached key conclusions, you'll need to pass any recommendations on to business stakeholders for implementation, and this can often be where BI processes break down. If other departments don't understand your results, or can't see the logic behind your suggestions, they won't be willing to implement them.

That's why good communicators are a vital part of your BI team. They need to be able to break down technical concepts for non-technical people and present findings in a clear, easy-to-understand format that also spells out the rationale for decisions. If the data reveals a spike in orders at certain times, illustrate this visually and offer supporting details to explain why this is the case. If you do, stakeholders will have greater confidence in the reasoning behind what they are being asked to do.

10. Continuously review your results

Finally, no BI project is complete unless its results are reviewed, areas for improvement identified, and recommendations acted on. This is where the KPIs you set earlier will be important, as they should give a clear indication of your success. If they're positive, great, but if they fall short of your expectations, this doesn't mean you should abandon your efforts.

Reporting and reviewing should be a process that takes place continually throughout the project, rather than something that's left only to the end. This ensures any lessons you learn can be applied quickly before you go down the wrong path. It should also give you an opportunity to review traditional thinking. For instance, are old KPIs still relevant to the business today?