• bookmark
  • print
  • 0

  • share

    0

3 steps to successful data-driven marketing

Sudhir Hasbe is no stranger to the field of analytics and data-driven marketing.


Publisher Image

Published by Questex Asia
on 25 Feb 2019

3 steps to successful data-driven marketing

Sudhir Hasbe is no stranger to the field of analytics and data-driven marketing. Before taking on his current role at Google where is he responsible for data analytics, Hasbe was the vice president of software engineering at US online retailer zulily.

With as much as US$100 million in annual marketing spend and nine thousand new products listed every day, zulily was the epitome of a data-driven organization. In Singapore earlier this week, he spoke to CMO Innovation on what organizations must do to find success with data-driven marketing.

Build a data-driven culture

The first step towards successful data-driven marketing is to establish a data-driven culture, he says. This isn’t something that the CMO or marketers can achieve in isolation but must start right from the top with CEOs and business leaders who genuinely care about using data as the basis for business decisions – and who expect their employees to do the same.

“Building a data-driven culture is about challenging people with how they can measure success,” Hasbe noted. “The push is always: ‘Can you measure it?’ If you are going to spend ‘x’ dollars on Google ads, the question would be: ‘Can you measure if this approach succeeding, what are the metrics, and are the metrics improving?’”

Unsurprisingly, relevant metrics will vary, and Hasbe says it is up to the organization to identify what matters the most to them: “[At zulily,] it was lifetime value of the customer. For others, it could be the conversion rate or improvements in margins. You need to figure out what metric you care about.”

Prepare for increasing data volumes

As organizations see more value in making data-driven decisions, expect the volume of data that is collected and processed to grow rapidly. It hence makes sense to have scalable systems in place to cope with a potential influx of raw data.

“As we see more value in data, we will collect and process more data over time. When I joined [zulily], we used to collect 100 million clickstream events a day. We were collecting 5 billion a day by the time I left,” he explained.

This did not take place overnight, but in fits and spurts. One instance that Hasbe related was when marketing wanted insights into what buttons customers are clicking, as well as which images are more appealing and led to higher conversions.

“After we started tracking [clicks], we realized that what we thought worked were not the ones that were clicked the most,” he said. “The size of data will grow. And you will make more decisions based on it. [Be sure to] invest in systems that can scale over time. You want to be successful; you want to build systems that are scalable.”

Hire the right people

Finally, Hasbe says organizations need to hire the right people to drive data-driven marketing initiatives. While he confessed that he is no expert in this area, he did offer a couple of thoughts.

“Look out for skillsets that allow you to leverage infrastructure and derive insights, and who are open to learning new technologies. [Technology] will keep evolving. Look for people who are curious and can be trained. If you pick the right platform [and can figure out how to use it], you can do a lot more stuff.”

Hasbe also emphasized the importance of problem-solving skills, which can be tested at the interview stage. “We give you a problem, about, say, designing a better bottle. And work with the person to understand how they break down the problem, how they solve it. Find curious people who do interesting problem-solving.”

“Ultimately, data-driven marketing is also concerned about rapid execution. Delayed decisions can result in you losing money, and a marketing campaign may well become obsolete if you don’t optimize it over time,” he summed up.

 

This article was first published on Questex, on 14 February 2019. Information is correct at the time of publication.

Last Modified Date: 25 Feb 2019