It is a modern-day alchemy. The new discipline of data science promises to transform raw data into new riches.
No wonder, according to HBR, the offering was dubbed “the sexiest job of the 21st century”. Data scientists mine data, in all its various forms, for hitherto-unseen connections, and make recommendations that can help businesses save or generate millions of pounds.
This is not your typical business intelligence. When we talk “data science”, we mean large-scale price optimisation, next-level segmentation, classification of customer bases, deep-diving on product profitability and building a forward-looking system that responds to unseen historical evidence.
In the UK, Parts Alliance Group, an auto parts distributor, recently used advanced data science to analyse 20 million transactions in order to build a dynamic pricing engine. The end result? After identifying which products should be discounted to match historical purchase patterns, the company is expected to boost its revenue by 30 per cent, or £6 million - not bad for just five weeks’ work for four junior Data Scientists.
We estimate data science is now being adopted by 30 per cent to 35 per cent of companies, and these tales of transformation are great. But, whilst larger companies find adoption easy, and have plenty to gain, stories of success amongst smaller businesses are thinner on the ground.
Small and medium-sized enterprises (SMEs) have heard the same hype as the big guns. Many are keen to harness the capabilities. But they are often stuck in a nervous state of paralysis. While corporations like banks, retailers and ad agencies can easily bolt on new disciplines at scale, smaller companies don’t necessarily boast the skillsets or budgets to follow suit and they are left continuing to hammer out rear-view reports in Microsoft Excel.
This is a real waste. SMEs are the backbone of our economy, so therefore it stands to reason that they are in fact the sector which stands to benefit most from data science. A report conducted by the London.gov has stated that SMEs make up to 99.8 per cent of the UK’s private sectors businesses, with their turnover equating to £430 billion. So, if we could get just 1 per cent adopting these tactics, we could grow our overall GDP significantly.
For SMEs to get started in data science, it is really no different than big companies. Businesses, regardless of size, need to know the problem they’re trying to solve - and there are distinct advantages a small business has in determining the outcomes of big data.
Data consultant Bernard Marr agreed with this in an article he wrote about big data for SMEs, saying, “In many ways, big data is suited to small business in ways that it never was for big business – even the most potent insights are valueless if your business is not agile enough to act on them in a timely fashion. Small businesses have the advantage of agility, making it perfectly suited to act on data-derived insights with speed and efficiency.”
It is a real shame that there are so few publicly-presented cases available to ignite and inspire business owners’ data odyssey. Those early adopters naturally see data science and analytics as a competitive advantage, and keep their cards close to their chest. So it is time we started making a real effort to spread the benefits of this new technology more widely.
Data science’s fundamental leap forward is the shift from retrospective business intelligence to forward-facing actions. Every business is likely tracking its sales and inventory, its revenue and profit performance. But, by the time it is reported, it may well be out of date.
When you bring analytics to bear on that data, however, you crunch it into forward-looking recommendations for efficiencies, like resource allocation, demand prediction and marketing optimisation.
That is what auto lease group Zenith decided to make better use of not only its own data but also publicly-available data on road traffic accidents. Over five weeks, four Data Scientists were able to predict the profile of a driver at risk of crashing one of its fleet cars with an accuracy of 98 per cent, ensuring it could advise its fleet customers to send particular drivers on advance training courses to mitigate the risk of crash.
These are the kinds of applications that can help transform operations for small and mid-sized companies. But how can bosses seize the opportunity?
1. Identify your business objective
You can’t catch a fish if you don’t know where to dip the rod. True to say, data science can answer many questions, so begin by adopting a mindset of curiosity. You need to know the problem that you are hoping to solve before you start. But aim to produce a clear and quantifiable business outcome, one that speaks to your unique pain points or growth goals.
2. Assemble a team
Can you staff the project from within, or do you need to look outside? Can you spend a six-figure salary on a specialist data ninja, or are you going to outsource the resource and work with a team? Your project may involve a mixture of partnership with on-hand specialists so start by getting the mix right.
3. Start small
Don’t overthink the outcomes, much less the process. Begin merely by examining the integrity and condition of your underlying data, then move on to speccing out a proof-of-concept.
4. Be patient
Results may not come overnight. Data work takes time - time to strategise, time to cleanse, time to analyse, time to disseminate. If your data is yet to be captured, that will add still more time. Measure and capture the outcomes and end results in order to truly ascertain the effectiveness of the project.
5. Scale the method
If it works, blow it up. If you see that data science works for your business, do it again and again. Companies that repeat and grow the process, who apply it to additional areas of operation, find the efficiency gains can be orders of magnitude greater than their prior systems. I hope yours is one of them.
Just remember, it’s never too soon for small business to start thinking big.
Jason Muller, COO and co-founder, Pivigo
Image Credit: IT Pro Portal