Big Data and Analytics – Data warehousing key for SMBs

Analytics and Big Data –hot topics for 2104. And still rather abstract for many small and medium sized companies. What do they actually mean for the average business? What kind of impact can they have? Is it a story about technology, or about added value? Can I apply my standard IT applications like ERP, or are we talking major new infrastructure? Here, I look at a realistic approach to this new phenomenon for companies without exceptional IT budgets.

Datawarehouses, BI- tooling, Hadoop, in-memory databases…‘what should we do about big data?’ discussions normally involve enough of these terms to get heads spinning.

Big Data manipulation is a powerful concept, but in reality its use and application, as has been publicized at many blue chip organizations, is not a realistic concept for most SMBs. Or is it?

Managing cost
Budgeting is tricky. Potentially expensive hardware and software will be necessary. Specialists who can use it need to be found and employed. And even if money’s available, many businesses don’t really know what they want to do with the vast information they’ll make available. It undoubtedly contains answers, but what exactly are the questions? It requires a major shift in strategic and operational thinking from the C-suite down.

That said, there is still ample opportunity for the average SMB business. Start with a critical evaluation of the data that’s already available. Properly consider the role of KPIs in the business. Look at accessing transactional and performance information created by IT systems, and then incorporating it in day to day processes.

Mind-set change
Businesses that hope to make use of (big) data effectively must install a culture of data backed decision making. The data already accessible within your business software and ERP can offer you all sorts of important information – from the sales pipeline and customer contact, to contracts and delivery performance. Where do the numbers point to possible improvements?

Do bear in mind, however, that you may be looking at out-of-date information. The processes around creating and initiating traditional reporting can be time intensive. Your business moves on quickly while you’re deciding what you want to look at and how you’re going to collect it.

Think real-time
The ambition should be to work with the real-time picture, a consolidated overview of the very latest information. It can have a dramatic impact on performance – helping you uncover and respond to key trends while they’re still in progress.

What’s happening with conversion this week? Is there something specific affecting deals right now that we can mitigate? You can follow up marketing activities in detail, analyzing what worked and what didn’t for the follow up. Supply chain managers can optimize logistic planning by understanding the fast and slow movers – not from last month, but from yesterday. And the list goes on. And on.

Combining data sources a challenge
However, we continually see how difficult it is to collect this information and make it uniform and accessible. It is often spread through multiple, sometimes aging systems. Sometimes there are issues with double data, or information inconsistency. If that’s the case, which data do we trust?

Because employees manage their own databases or excel sheets, it can be difficult to see the actual situation. Be clear about where people manage their information, determine which system is in the lead for what topic, and be consistent in data management processes. The info then needs to be quickly and easily accessible for whoever needs it. There needs to be a simple and intuitive way for them to get their questions answered.

Cue the data warehouse.

Feed information streams into a single source
‘Data warehousing’ involves the combination of dispersed data, often from a range of IT systems, into one main database. By breaking open silos and pouring the data into one central reserve, everyone can see what’s happening everywhere in the business.

Once companies have mapped out their individual applications and non-aligned processes, plans can be made for a clear data structure. A central repository can be established, processes can be set up to ensure employees share the information they create, and individual systems adjusted to properly feed in.

Get help from the right tools
Once the information has been collected and aligned, it’s much easier to pull out genuinely useful insights. Traditional reporting techniques can get useful answers to specific questions, but they are inflexible. You’ve answered a question, but it’s probably only created more. Answering them requires a trip back to IT to get more queries designed.

If budget allows, it’s worth exploring what analytics tooling can be applied. There are a range of affordable dashboard and data ‘slice, dice and display’ products now available – products like QlikView or Sumatra.

Use them via mobile applications, and suddenly every employee in every layer of the organization can enjoy anytime, anywhere access to real-time data. It’s going to significantly boost their decision making – be it the status of sales orders, inventory positions or financial performance information.

Know what you want to know
Businesses need to know what information they want to use to steer their decision making. Everyone needs to be aligned, from top management to the work floor. Once the data collection necessary to monitor these KPIs has been organized from the organization’s ERP and other IT systems, steps can be taken to ensure everyone who needs to can stay up to date.

Ready for the future
Over the coming months and years we’ll see many more fascinating examples of how resource-rich blue chip companies like Bol.com or Albert Heijn uncover hidden insights and dramatic improvements. We’ll be stunned by the quantity of the data they create, the investments of time and energy required to dive into it, and the results it generates. For these billion euro businesses, the investments are realistic – both in terms of the capital they have available, and the potential impact even relatively small changes can have when multiplied by the size and complexity of their operations.

However, there are also benefits for smaller SMBs. Needing to take a more conservative view, they should first get the systems they already have working for them in the right way.

There’s still much to be won for the average business in considering effective data management, application of existing reporting techniques and newer business intelligence products. If a solid strategy and firm infrastructure are in place, it’ll be much easier to make real progress with dedicated Big Data solutions once they become more accessible and applicable in the future.

5 tips for getting started with big data

1. Thoroughly evaluate your existing data. Look at how you can make better choices by using the data you already have. It doesn’t matter how much data you have if you can’t operationalize the information.
2. Build out your data capacity. What if you could add a few additional data points to the information you already have – what could you do then?
3. Make sure you know what your hypotheses and end goals are. Lot’s of experts forget to position a hypothesis and jump straight into the numbers. If you don’t have a good handle on your theory predictions, it becomes very difficult to evaluate where you’ve had success.
4. Don’t forget to look beyond what’s highly relevant to the ‘here and now’, and consider ‘what could be’ for possible scenarios in the future.
5. At a certain point, it’s worth looking if you can develop your own tools and infrastructure for the collection, analysis and visualization of data. And if you get stuck because your information is too complicated, it might be time to consider bringing in a data scientist. They can tell you exactly how the pull the maximum benefit out of a minimal data set.