By Jason Foster, CEO & Founder, Cynozure
We think data strategy is awesome (if you’ve met any of the Cynozure crew, this won’t be a surprise)… If this is new information, read on and I can explain. We’ve seen data strategies that have completely revolutionised the way a business uses its data.
It’s no longer good enough to simply set up a few data projects and hope that they create huge results for your business. In today’s world, a well-defined and executed data strategy is what separates the trailblazers from those trailing behind.
Strategies lead to super results
To really unlock the potential of any data project, it needs to be part of a wider strategy – and you might want to have a few data projects on the go. The best, most innovative, most switched on businesses are data-driven. Think of Google, Amazon, Alibaba, Facebook – they’re all built on a massive pile of data. You can bet that they all have solid data strategies too.
Using data as a way to get an edge is still a pretty new concept, so not everyone is clued-up on the many benefits it brings. First off you’ll have to gain buy-in from senior stakeholders, and also the people that are going to be a part of your data projects (the whole company, in other words). The quickest and most efficient way of doing this is through a data strategy.
Convincing people is probably going to be easier if, say, you’re the CEO and can greatly influence others in your organisation, but you’re still likely to hit some roadblocks along the way (it’s almost expected with any type of change). For example, some members of staff may not recognise that dealing with data is an integral part of their role. The same goes for any of the last remaining technophobes (a bit of a dying race these days).
We all create data, so everyone should be involved
The reality is that everyone has to deal with data. Once they realise it, everyone will begin to rave about data. It holds huge potential to make everyone’s working day run much more smoothly. Plus, we’re all creating it, constantly. From the moment we wake up and check Facebook, to when we ride the tube, brew a cuppa using our smart kettle, and tell Alexa to wake us up the next morning.
Everything and anything your team does has the potential to create data, and from that, to create some really interesting insights. Your whole organisation should be part of your data journey too.
The board need to get clued up on data
Generally speaking, one of the biggest hurdles to an effective data strategy is a lack of nitty-gritty data knowledge at board level (a good example of this is the confusion betweenand business intelligence). If your organisation is already getting ahead and you have a Chief Data Officer (CDO) or Data Leader, that’s brilliant! A large part of the data strategy can fall to them.
If you don’t have a CDO you can look at the best way of getting this level of strategy and leadership knowledge into your business – that might involve hiring in a third party or external expertise. If you don’t know what a CDO even does, then by the end of this, you’re going to want one. A CDO is the person who will lead on your business’ ENTIRE use of data. From collecting and cleaning it, through to analysing and interpreting the results. They are, quite frankly, the person you want on your side when using your data as an asset and not a by-product (more on that later).
What exactly is a data strategy anyway?
A data strategy is essentially a vision of how an organisation intends to use its data, and a framework for how they will go about it. Every business has a different data strategy – there’s no set rulebook on what to do exactly. Your data strategy can be as all-singing, all-dancing, or as simple as you need. But there are some principles and a certain framework you should work around.
A data strategy is, in some ways, similar to a business strategy. Just as you set out your vision for your business, your data strategy will outline how you use your data to give your business a competitive edge.
It will then go into detail on how you intend to achieve that, including:
- The business practices that will be adopted across your organisation,
- What technology you will use,
- What your data management and governance will look like,
- And how you will bring your entire team on board.
In the past, many businesses created data as a by-product of everyday activities, and it was a little bit like the fat on bacon – not everybody took a bite and some wasted it. Now data is integral to every business process, and it definitely deserves to be a part of senior decision making. If you want to bring home the bacon, you’re going to have to tap into your data.
Silos are your data’s worst nightmare
Because of this, many businesses set up projects to use and analyse data. However, these projects are set up in silos, without any master oversight and nobody really taking the lead over them all. Without a strategy, data projects are left to their own devices, all producing information, but without the structure of a common goal. Even worse, these projects can also overlap, meaning more work, duplicate data, wasted resources and probably a few frustrated data scientists working behind the scenes!
A data strategy is also almost a legal requirement, at least in terms of how some data is stored, transferred and governed. Because of the General Data Protection Regulation (GDPR), any business that uses personal data needs to prove that it is compliant with the Regulation. The only way you’re going to be able to prove this, if the Information Commissioner’s Office (ICO) comes knocking, is by having a clear and coherent data strategy.
Stuff your data strategy with these key things…
So, we’ve covered the many reasons why a data strategy is a must-have for all businesses. But not all strategies are created equal. There are a few crucial components that an effective data strategy must have.
1. What you get out of it
This one seems like a bit of a no-brainer for most results and revenue driven organisations, but you’ll be surprised by just how many companies fail to set out what value they wish to get from data.
The potential benefits of using data include:
- being able to gain competitive advantage,
- drive new business opportunities,
- streamline operations,
- and solve business problems.
That’s all pretty inspiring stuff, and it’ll catch people’s attention. So, be sure to include it in your data strategy.
2. What the data is and how it is stored
To know how to use something, you first need to identify what it is. Therefore, the first part of your data strategy should outline the data your company currently has and what it is intending to collect. Making sure everyone names data sets in the same way means that everyone is on the same page. It means you’ll get way more bang for your buck – and you’ll be able to use your data multiple times.
Oh and don’t forget about your metadata too. This is like data’s trusty sidekick, and it’s going to save your tail if GDPR enforcement comes knocking. Under GDPR you’ll have to record consent for personal data use with the data that it relates to. Otherwise you’ll be in breach and liable for a hefty €20 million (or 4% of global turnover) fine. Ouch. So record your metadata now and keep that €20 million squirreled away for other important stuff. Like hiring a kick-ass data consultancy.
Once you know what your data is, then you’ll discover the best way to store it. All data needs to be kept somewhere, but not all data storage is made equal.
If you’ve ever sat down with a data engineer then you’ll know just how complicated data storage is. Suddenly you’re five minutes into a coffee with your head data engineer and they’re reeling off frameworks you can use, like Apache Hadoop, or perhaps Spark? And, because you’ve got a ton of unstructured data you’re going to need a NoSQL database – as your data engineer begins gesticulating wildly, your mind wanders to who even names these things?! Hadoop and Spark sound like a terrible boyband. Sound familiar?
The complexity of data storage shouldn’t put you off, and you should be aware of (at least) a little bit about how it works. Otherwise, with no senior storage knowledge, you could face a data breach of Yahoo-level proportions.
In today’s world, it’s not enough to just build storage around each new data set that your business creates. Data storage solutions should be designed with collaboration (and privacy) in mind.
It’s pretty common for data strategies miss the mark when it comes to efficiently managing data storage. They can also make it a bit of a Mission Impossible for different teams to access and share data if not developed and implemented effectively right from the get-go. Data portability is also a requirement under – you guessed it – GDPR, so being able to swiftly and securely move data between systems is now a legal requirement, rather than just good practice.
In short, you need an effective central sharing system that’s going to bring an end to separate teams and lines of business duplicating their efforts, and data, across silos.
But that’s not to be confused with a single location. All your data doesn’t have to be kept off the shores of Iceland. It just needs to be able to be accessed quickly by the people who need it, when they need it.
3. Your data dream team
To become a data-driven organisation, specific skills and specialists are required, such as data scientists, engineers, and perhaps a few analysts… possibly a CDO… maybe a data protection officer, oh, and don’t forget students who are itching to get their feet wet with real commercial data experience! This is where outlining your team and organisational structure really comes into its own. Everyone knows who reports to who, and who is responsible for what.
It will also help you identify gaps in your business where you’ll need to hire additional help. The industry is suffering a real skills gap. So it’s pretty likely that you’re missing someone critical. Better to find that out at the start of your data initiatives, than at the moment that you really need a Hadoop hero.
It also helps you avoid hiring a person who, on paper has all the right skills, but who is lacking in critical experience. If you’re using your data to take a long, hard look at your marketing effectiveness, your best bet is a data scientist who has worked on similar stuff – not one who has spent the majority of their career undertaking academic research measuring the health of cows through Fitbits.
4. What tech you’re using
Data isn’t really much without some science, and a cool analytical model that makes sense of it. For which, you’re going to need some tools. Incidentally, this also plugs into your dream team. Without knowing what tech you’re using, you won’t know whether to hire a Python master or R aficionado.
Your technology stack should extend far beyond just the data analysis. It should also include the tech you intend to use to collect your data, store and manage it, and finally how it should be visualised.
If you already have some technology that you’re happy using, your strategy should outline how you intend to mix the old with the new (or whether that’s even possible).
There are many other aspects that should be included in a data strategy, including more information on data governance and tools that can be used – we’re only scratching the surface. Don’t be tempted to make it a copy and paste job either, as every data strategy is unique to the business that uses it. Plus, it should be updated regularly.
Data is constantly evolving, and your strategy needs to keep pace and evolve with it.
If you learn one thing from our efforts here, then it should be this: there’s no point in investing your time and resources into data projects if you cut corners when defining and executing your data strategy.
A data strategy done right will drive your business forward and bring in new revenue streams, actionable insights and much more. It’s worth investing as much time and energy in your data strategy as you do with a business strategy. It will pay dividends in the long run.