Why making your business analytics-driven is not as costly as you think

Big data has been the buzzword for companies as well as marketers for the past year, with many firms excited by the idea, but only a scant few being able to utilise any part of it for their businesses.

But is making a business data-driven as difficult, or as costly, as it has been depicted? Perhaps not.

Professor Ashwin Malshe, assistant professor of marketing at ESSEC Business School, thinks big data is merely a buzzword that shouldn’t be overhyped. “At the end of the day, big data is just data.

“As of now, if I take financial statements of all the American companies that are publicly listed from 1950 to today, all that data can be put into a flash drive, less than 80 MB. Think about your business, do you have that type of data? The answer is probably no. Then you don’t require complicated software.”

Cheaper and easier than you think

The views from senior executives from small to medium-sized enterprises (SMEs) such as Roomorama and ZUJI agree with Malshe.

Teo Jia-En, founder and chief operating officer of online accommodation booking site Roomorama, said a focus on key objectives was vital. For many companies, and particularly SMEs, cost for analytics is one of the main considerations when it comes to investing in the area.

Echoing the concerns of many firms, Teo said: “We are a small company by all means; we have a limited budget and resources. What we have to do is really focus (because) data is everywhere.”

For example, every action users take on the site is informative data, helping the company understand how a user goes from being on the site to eventually making a booking.

She said it was important to segment that process and understand which part needs more attention.

For that, there is a lot of cheap software a company can purchase online. “That really allows us to run tests and see what experiments work and what don’t against the base line. Whatever doesn’t, we scrap very quickly,” Teo said.

For certain parts of its analytics, Roomorama may outsource it to agency partners for expertise. But Teo also warns that as a client, businesses should also understand their own users and conversion rates well to keep partners accountable.

Grooming in-house analytics teams

Aditya Sikka, head of analytics at Zuji, owned by Webjet, agrees with Teo’s point of a company being clear about its own analytics. Sikka talks about how Zuji is now owned by Webjet, where it was once owned by Travelocity, to the effect the company now operates more like an SME.

Echoing similar concerns of cost as Roomorama’s Teo, Sikka believes the core thing is to build an analytics team in-house, and this doesn’t have to be big.

“When you build that skill set in that team – where at least two people are aware and numerical nurtured – then you can engage with agencies for higher value, higher projects,” Sikka said.

This brings up the issue of talent crunch in the analytics and data function. While data scientists are now one of the most sought after roles by companies (Read also: Why data scientists are the next big thing, 10 jobs that didn’t exist 5 years ago), the practice still lacks manpower.

“Now if you are a small business, you will find it even harder to attract talent, even if you are not considering (talent with) PhDs, just people who are good with the data, you may not be able to keep them because of your restrictions,” Malshe said.

“What I tell businesses is that you may not hire from outside, you can groom them from inside. You do not have to train them on more complicated stuff; you can train them on very simple statistical analysis.”

Teach staff how to make managerial decisions based on the data, he added. He went on to talk about how analytics workshops are now incredibly accessible, for example, at institutions such as NUS or even the free course website Coursera.

“There are people here in Singapore who have learned from that and are so good at it that they are ranked in the world,” he said.

“You would be surprised how much people can learn on their own if you give them enough incentive – some budget, some time off and encouragement. I think that is a great way for small and medium-sizes businesses to start investing.”

Keeping it in-house versus outsourcing

While the argument so far has been that companies should have a basic in-house base, the question of when to get external help can and will arise.

“If you think that getting intelligence is not a core function for your business, but it is a supporting act, it makes sense to outsource it,” Malshe said.

The merits of keeping data and analytics in-house is you can always control your data, he added. This may not be an option for businesses with more restrictions and security issues, for example, in finance or telecommunications, where this may be illegal even.

Also, in the case where data is large, outsourcing may be out of the question as well.

“You lose control. Whenever control is important you want to keep it in-house. On the other hand, whenever you are looking at analytics to get insights or not complicated predictive analytics, if you are using for strategic decision-making, I would advise you to outsource,” Malshe concluded.

Malshe, Teo and Sikka were in a panel discussion at Marketing Magazine’s Analytics Interactive, moderated by editor Rayana Pandey.