Breaking it down by the numbers, there were four experts from two countries and a couple of hundred audience members at a panel session during Thursday’s Data Marketing 2015 conference in Toronto.

It all added up to a lot of insights on the nuances of crunching numbers to win and keep customers today. We’ve boiled it down to 10 tips on how to use analytics for data driven marketing.

1. Focus on people, not percentiles.

Make sure your customers don’t get obscured by all those numbers, said panelist Jenne Barbour, Texas-based director of marketing at Teradata. “We don’t build relationships with data points, we build relationships with individuals,” Barbour reminded the audience.

2. Ask questions that matter.

“Sometimes I ask (executives), ‘If I got you the number you’re looking for, what would you do with that?’ Make sure you’re asking the right questions,” said Brent Dykes, the Utah-based customer analytics evangelist at Adobe.

3. Analyze your failures. 

Everyone loves to look at why a marketing campaign succeeded. But Dykes said you can learn just as much from seeking data-based insights into where things went wrong. “Go back and look at your unsuccessful campaigns and look at what happened,” he suggested.


Huffington Post's Graham Howes.
Huffington Post’s Graham Howes.

4. Crunch numbers but stay creative.

“Focus on the expression, not just the impressions,” said Graham Howes, director of audience development at Huffington Post’s New York City office.

While it’s easy to get caught up in simply measuring how many online impressions a campaign generates, he said, marketers need to be creative to really connect with customers on a human, personal level.

“The data is our roadmap but it’s not going to get us where we want to go,” Howes said. “How are you going to stand out from the crowd … to separate your (business) from all the noise out there?”

5. Listen.

Instead of looking at data in isolation, Barbour said marketers should listen to what their clients say about their customers, products, goals and services. “Marry the data with what you’re actually hearing from them,” she advised.

6. Consider the context.

It’s tempting to put blinders on and focus only on your client and what they’re doing, Dykes said. But what are their competitors doing? Did one of them go out of business or leave that product stream? Widening the breadth of your analysis adds important context to your data results, he said.

U.S. CMOs will spend an average of 11.1 per cent of their budgets on analytics in 2018, up from 6.7 per cent in 2015. (Source: Aug. 2015 CMO Survey)

7. Aim for value over volume.

“It’s so easy to generate complex reports. But give people what they need to actually do their jobs better,” said Greg Dashwood, Microsoft Canada’s product lead for Internet of Things and advanced analytics.

8. Hire the right help.

Skilled data scientists can really extract value and context from your data, said Dykes – if you can find them, that is. Those people are in high demand.

9. Create value but don’t creep.

Marketers walk a “fine line” between creating a personalized customer experience and making people feel stalked online, Dykes said. “Look at (data) as a precious resource, be respectful of it and it’ll give you a great return over time,” he recommended.

10. Give and take.

Dykes believes the next wave of marketers will view analytics as more of a two-way street. Besides just collecting data from customers for their own corporate purposes, companies will focus more on “giving it back to (customers) with context,” he said. “Not just for here’s how much we know about you, but here’s how we can make your life better. That creates that loyalty and that bond over time.”

Share on LinkedIn Share with Google+
More Articles

  • Ilya Geller

    What are you talking about?
    Oracle already structures unstructured data:
    1. Oracle obtains statistics on queries and data from the data itself, internally’.
    3. Oracle gets 100% patterns from data.
    4. Oracle uses synonyms searching.
    5. Oracle indexes data by common dictionary.
    6. Oracle killed SQL, there SQL either does not use statistics at all or uses manually assigned one.
    Meanwhile IBM said: “Cognitive systems represent an entirely new model of technology which understands “unstructured data through sensing and interaction” – IBM is to start the structuring soon.

    Analytics is over: everything can be easily found at the structured unstructured data.