8 tips for data analytics success from Data Strategy Symposium
- 02 December, 2014 16:52
This year’s Data Strategy Symposium speaker line-up offered a diverse range of business insight, advice and expertise around data and analytics. Here, we share 8 key pieces of advice we took away from the event that could prove invaluable to your organisation’s future.
1. Have a data strategy
Mia Dand, principal at Lighthouse3, said the way to business success through data comes down to four strategic steps: Organise, analyse, act, and optimise. Organising data can be done using a succinct list of useful questions, she said:
- How will you use it?
- How will you secure it?
- What people and resources do you have, and what do you need to find?
- What processes do you need to follow to protect the integrity of the data?
- Which technology do you need?
Analysis, meanwhile, is not only about gathering a vast array of data to draw insights from, but also requires reporting and integration of assets to drive intelligence. Dand added the ‘holy grail’ of analysis is to go from attribution through to contribution and then prediction.
But data gathering and analysis are not end goals, nor are either useful unless you then act and optimise your findings by constantly testing and measuring the effectiveness of every action, she said.
“Achieving the desired business outcome is the ultimate objective,” Dand said.
2. Close the customer feedback loop
According to chief of Data@Ogilvy, Kuba Tymula, the best way of achieving the ‘3Rs’ of customer-led marketing – recognise, relate and retain – is by commercialising the data in your organisation. Doing this requires not only a 360-degree customer view, but also engaging in an ongoing dialogue with customers using data.
“Brands need to start acting across the customer journey and pick up all the crumbs of data around that, in order to take people across that journey,” he said. “Every interaction should improve the next interaction.”
4. Be a data analytics connoisseur
With disruption the rule rather than the exception in business today, Dr Eugene Dubossarsky of data analytics consultancy, Presciient, labelled data analytics the tool every organisation needs to rely on to remain competitive.
“We’re going to see increasing sophistication of in-house analytics,” he told attendees at Data Strategy Symposium. “Competitors are hitting your saturated market and they’re using analytics to do it. You need to invest in analytics not to win, but to simply not die or lose your customer base.”
But it’s not enough to launch a big data initiative, put in some predictive modelling and tick the completion box, nor is setting up a specialist team and hoping for the best, Dr Dubossarsky said. You have to continually innovate right across the business. And that means knowing what you’re doing with data and analytics.
“You can’t play politics with analytics or pretend to know what you’re doing if you don’t,” he warned. “If your analytics efforts aren’t working well, you’re dead; if they’re working, you might just be OK.
“Analytics is central to your business and it affects everybody. You have to have analytics literacy. You don’t necessarily have to be able to do the maths, but you have to understand what a predictive model is for, and what makes one work versus not.
“Be connoisseurs, if not chefs.”
4. Don’t use data as a crutch
Financial Times head of analytics from the UK, Robin Goad, said too many organisations only look at data as a way of reporting on existing decisions and actions. Instead, they should see it as an opportunity to drive innovation.
“We need to use data for illumination, rather than just reporting or supporting decisions made at a senior level,” he said.
While there is always going to be business-as-usual work to be done by analytics teams, Goad advised organisations to allocate a percentage of time to focus on new things.
“Try and carve out time to build out stuff that is innovating, empowering your stakeholders or improving the reporting you do in that business,” he said.
5. Prioritise ‘foundational analytics’
Many organisations assume – wrongly – that improved business performance will result from simply having more data and the right tools to mine it, Analytics Partners senior director, Dr Shawna Thayer, said.
“So why is it organisations aren’t able to activate on their data strategy?” she asked. “A lot is because there’s an assumption that once you have the data in-house, and you have the right technology, it will drive better decisions. But that’s not all we need.
“Organisational leaders may not be ready to consider new ways of making [fact-based] decisions. And when that happens, you can have the best data and tools in the world but nothing will come of it.”
Drawing on her work at Kraft and SuperValu within analytics teams, along with her experience at Analytics Partners, Thayer said the key is to prioritise foundational analytics.
“A lot of people will say, ‘wait a second, you can’t do analysis unless the data is fantastic’. The reality is the data sources are changing all the time, and they’re constantly improving. Before you know where you want to go with your data strategy, and what is the next great implementation you’re going to do, you’d better have an understanding of your business first,” she said. “Foundational analytics will help you drive that data strategy.”
It’s a process – creating your data strategy and activating it is a process
Using the example of a retail client, Thayer explained her team first worked to understand the business objectives, which was to improve its marketing activities. The first step was an analysis of the retailer’s existing media mix, quantifying what was and wasn’t working and showcasing the opportunities in digital marketing channels. Analytics Partners then evaluated the upcoming marketing plan to assess budget optimisation using ROI figures as well as industry benchmarks on newer media options.
Only once those foundational analytics were in place, could the retailer improve digital results, Dr Thayer said. And once those first objectives were met and as foundation of analytics in place, the retailer could assess enhanced data analysis and other business opportunities.
“It’s a process – creating your data strategy and activating it is a process,” Dr Thayer advised. “Just start and keep going. Ground that data strategy in business objectives with activation at the forefront. You need to know that going in.
“Start with foundational analysis. It’s easy to say you have objectives, and say that’s the data to get where I need to go. There are things you can do with the data you have currently, and ways you can get creative. Give yourself the foundation you need so you ensure whatever tool you invest in, is the right one for you. And just keep improving with a constant feedback loop.”
6.Unite data with experience
According to AMEX vice-president of consumer acquisition and customer loyalty, Dean Chadwick, brands should be combining data with creativity to become part of a customer’s real-time context. He explained the financial services company is bringing together art and science in order to interact better with customers in real-time, and putting data at the heart of commerce.
“There’s no point having a great set of data that drives a recommendation if there isn’t an awesome experience that’s connected with an audience on top of that,” Chadwick said. “We need to move away from random acts of marketing, and have data drive those contextual experiences… delivering real-time value to people on an ongoing basis.”
One of the ways AMEX is achieving this is by investing in data-driven resources and bringing together data scientists, engineers and marketers. It’s also using data to inform a number of new products and programs. To do this, Chadwick said it’s important to look at nodes of data, as well as the edges connecting these nodes together.
These can be defined in three ways: Stated nodes, such as where a customer lives; triggered nodes, or the event or action that has occurred; and inferred nodes, which are based on data marketers have on customers, such as an interest in fashion based on purchase history.
“What we’re aiming for are new concepts that are customer-centred, but with an overlay of good product design and good data science, that delivers an awesome experience,” Chadwick added. “We try to get products out to the market place quickly. We don’t always get it right, but we always learn from it.”
7. Create a social data map
Lighthouse3’s Dand also advised organisations to have a holistic approach to social data. One of the biggest problems is that companies still retain highly siloed business functions. Yet social is not only marketing’s responsibility, it’s also about customer care, IT, product development, innovation and more. In fact, Dand claimed 13 different departments on average will be engaging in social media in one organisation.
This also means marketers need to ensure buy-in for their social strategy from across the organisation.
Dand suggested using Altimeter Group’s social data maturity map as a way of finetuning social success. The maturity map based on several key components that add up to the desired strategic outcome: Scope, strategy, context, governance, metrics and data.
A holistic and ideal strategy will stretch across the enterprise (scope), with all social metrics mapped to business outcomes (strategy), Dand explained. Benchmarks established and shared (context), governance will be entirely programmatic and integrated into established workflows, and metrics will be clearly defined with shared KPIs across the business.
Most importantly, social data will be just one part of an integrated data strategy where all customer-oriented data sets sit together.
8. Test, test and test
Throughout all presentations over the three-day conference, the importance of testing and learning using data for insight and measurement was emphasised repeatedly. In fact, Data@Ogilvy’s Kuba Tymula claimed that if 70 per cent of your tests aren’t failing, you’re not doing enough of them.
As Reckitt Benckiser head of digital, Andrew Wong, put it, be prepared to fail, but do it quickly. The FMCG company is looking at a timeframe of 90 days or less to develop an idea, because any longer means “you’re overthinking it”, he claimed.
“Our mantra is to fail fast – you need to be brave, innovate quickly and fail fast, but ensure you learn from your mistakes,” he said.
Follow CMO on Twitter: @CMOAustralia, take part in the CMO Australia conversation on LinkedIn: CMO Australia, join us on Facebook: https://www.facebook.com/CMOAustralia, or check us out on Google+: google.com/+CmoAu