Tag Archives: business analytics

The balance between intuition and data-driven decision making

by ON
Reply

“Marketing strategy is a combination of analysis and art – know more and then be creative.”

That quote from Brook Carter, head of marketing strategy and insights at AustralianSuper, is timely given the recent interest and/or hype in the use of business analytics as a means of understanding useful patterns within ‘big data’. This article is based on conversations I’ve had and comments heard, in particular at a breakfast event I hosted last month at the Deakin University/IBM Centre of Excellence in Business Analytics, titled ‘Finding the balance between intuition and data-driven decision making.’

The interest/hype around ‘big data’ and its implications for customer and business analytics has been driven by two things:

  1. The proliferation of digital footprints (shared images, likes, comments, tweets etc) consumers are leaving in cyberspace from an array of devices (phones, tablets etc. They are doing this prior to purchase when in search mode as well as after they purchase a product. The amount of this unstructured ‘evidence of attitudes and behaviours’ is growing exponentially, and
  2. the ability of large computers to process information at unprecedented speeds.

 

Implications for strategy 

For marketers, this provides the potential to use ‘big data’ to enhance competitive advantage in a number of ways. These include: identifying unmet needs, alternate, richer segmentation criteria, more precise targeting of offers in terms of time and channel, being forewarned of PR issues about to escalate and, finally, identifying issues that might lead to customer churn.

According to Brook: “With the amount and speed of information we have at our disposal, as well as the leanness and commercial reality of business today the old frameworks of the ‘360 degree or end-to-end marketer’ is gone. We are in an era of specialists.  If your current marketing team is focussed on ‘business as usual’ or ‘thinking for incremental growth’ activities, then hire in someone to think big. Innovation is the catch cry, so be innovative in your products, services and most importantly, the type and calibre of people you bring into the organisation.”

Gautam Bose, general manager of customer analytics and research in marketing at NAB says a ‘big data strategy’ should consider the following questions:

  • How do we personalise the information for our customers to influence them?
  • What tools do we have at our disposal to do this and when?
  • What are the different tactics we should employ?

The ‘IBM 2010 CEO Study: Telecommunications Industry Perspective’ co-authored by Skev Ioannou, market segment manager for the telecommunications sector at IBM, says a CEO’s most important priorities were:

  • 93% – ‘getting to know their customers better,’ and
  • 69% – ‘insight and intelligence’.

In a recent IBM case study it was found “good, big data is essential, but unless there is corporate alignment and clarity, then strategies will not be executed through the business and there will be no improvement.”

Implications for skills and recruitment

Bose says, “Marketing departments need to be skilled in the following areas: ability to produce and publish content, ability to share knowledge, managing the communication overflow, balancing openness with confidentiality, and finally, staying ahead of the ‘curve’.

Both Bose and Carter highlight the ‘collision’ between market research (data at rest) and dynamic market information (data in motion). What is the new role of the traditional market researcher given the type and speed of information we are now receiving through 24/7 technology?

In terms of the need for skills in this area, some universities have responded to this shortage and introduced academic offerings in business analytics. Dineli Mather, head of the School of Information Systems at Deakin University says, “The Deakin course aims to develop business analytics professionals with a broad skill set spanning business analysis, information management, data analytics, business intelligence and decision analytics, rather than ‘data scientists’ who focus entirely on the analytics.”

This growing area also has implications for recruitment Christine Khor of Carrera Partners, who says, “From a talent acquisition and hiring perspective, this adds another layer of complexity. The amount and speed of the information we now have access to is astounding. High performing employees of the future will be those who have the ability to access and assess data but also know how to process that data.  Is it useful? Is it junk? Can it be shared and what are the implications of sharing?

“Years ago companies would have one media spokesperson who was well trained and scripted. In our modern world of data and digital we are all potentially spokespeople for our organisations. The onus of getting the right people into organisation who have the ability to think through implications quickly is even more essential.

“Additionally, the role of the specialist within organisations will increase. With the greater depth of data we have access to the greater the need for specialists understand and draw insights from that data.”

Historically, market researchers and marketers have looked at the correlation between an action and an outcome. For example, when we run an advertisement for diet pills after Christmas, sales increase. The bigger question is why? What are the psychological, cultural and environmental triggers to this purchase and how do we replicate this to ensure sales throughout the whole year? Again, this needs strategic thinkers with initiative and an inquisitive nature to understand and interrogate the data.

From the discussions I’ve had and readings, these are some of the key barriers to an organisation’s ability to effectively use data on a large scale:

  • Political interdepartmental pressures and conflict between marketing and IT (who owns the process?) and marketing research (do we need surveys anymore?),
  • the difficulty of getting business analysts with the right ‘skill/mind set’. Ideally a blend of commercial acumen combined with rigour and patience to go through seemingly endless correlations within diverse data sets in the hope of finding a causal relationship that marketers can use in developing products or evaluating marketing performance,
  • developing a culture that accepts databased logic in decision making, and
  • identifying spurious versus real relationships in data.

 

In summary, we are at the early stages of ‘The Big Data Journey’, and two things appear clear:

  1. Trial and experimentation appears to be the approach, and
  2. some organisations have a board or CEO putting pressure on the marketer to show they have a plan to consider or implement ‘big data’ initiatives, while in other organisations marketers are proactively selling the ‘big data’ business case to the board.

A final thought from Bose: “Big data is about causation not just correlation. It is about answering the question why?”

 

Did you know: in each issue of the print edition, Marketing includes the very best opinion articles curated from our huge industry blogging community, as well as exclusive columnists writing on the topics that matter? Becoming a subscriber is only AU$45 for a whole year, delivered straight to your door. Find out more »

6 ways to stop annoying your customers

by ON
Reply

This advertorial is by SAS. To learn more from thought leaders and industry experts on business analytics, customer intelligence and risk management, click here to visit the Knowledge Exchange.

 

Customer relationships are like personal relationships – what you put in is what you get out. You both have a role to play, and in the case of a company, it’s important to make extra efforts to show you care about and value your customers.

But many companies stick to routines, create monotony, and continuously confuse and annoy their customers. By failing to plan and operate in sync, individual departments vie for a customer’s attention and reveal the company’s lack of understanding of that customer.

Fortunately, customer analytics enable a new and better approach. At the eMetrics Marketing Optimisation Summit in San Francisco. Retha Keyser, business development manager for the SAS Customer Intelligence Global Practice, talked about the steps you need to take to avoid irritating and annoying your customers. Follow these six key steps to become a learning organisation:

1. Collect as much of the right data as possible. “Start with the transactions and the channel systems, and collect as much information as you possibly can,” said Keyser. Make sure you are collecting the right information at the right level of detail. For example, you cannot predict the sale of ice cream very accurately if you don’t collect the daily temperature, and you can’t do meaningful analysis at the individual customer level if all you have from Web and mobile channels is summary-level reports.

2. Create a 360-degree customer view that includes data from every relevant source possible – from back-office systems to online and offline customer contact channels for every business unit, product or brand.“You need to include that channel behaviour, because that is a very important source of behavioural data that gives you great levels of accuracy that you wouldn’t have with just pure demographic data,” said Keyser. Good data management tools and common metadata make it possible to incorporate data from disparate and incompatible source systems.

3. Build customer intelligence on top of the data warehouse, said Keyser. “Without analytics, it is impossible to build rules with the necessary level of specificity, such as ‘This campaign should go out to people who have brown eyes or blue eyes and came to the website from Google and then did this and then did not do this and so on.’”

4. Automate inbound and outbound communications. “When we move to real time, and the customer has opted out, your only chance is to catch them while they’re coming to you,” said Keyser. “You need to know the right offer to make while they talking to your call centre agents, while they’re walking into your store, or while they are navigating your website.” That responsiveness requires automation with real-time decision management.

5. Add scope and analytic sophistication as you go. Continuously keep on building, because as we’ve seen over the past five or 10 years, new channels are constantly emerging,” said Keyser.Add predictive analytics to segment customers on a much more dynamic basis, to automatically detect what drives different groups of behaviours and outcomes. Use text mining to understand unstructured data. Add constraint-based optimisation to make the best choices within known limits set by contact policies and internal resources.

6. Make analytic insights available at customer touch points across marketing, sales and service. “We’re seeing disconnects happening where the service side is not aware of what’s happening on the sales side, and vice versa,” said Keyser. “You have to optimise all those interactions, inbound and outbound, with all your brands, products and across all the different channels. This is not simple to do, but the technology is there to enable it.”

For more, read the full white paper, How to stop annoying your customers.