Data can actually be a source of creativity, not just the answer to specific questions or challenges facing a brand, TBWA Global Data Director Baker Lambert said at Tuesday's Ad Age IQ Conference: Marketing & Technology in downtown Manhattan.

Data has led TBWA to new ways of telling a brand story and new tools to tell it, Mr. Lambert said during a presentation titled "How Not to Suck at Data." The discovery that National Donut Day had a high incidence of social mentions, for example, led TBWA to create a video for its Nissan client showing cars performing donuts in a parking lot covered with 1,000 pounds of sprinkles -- all to cover actual donuts in said sprinkles. The clip became one of the most- shared and most-viewed pieces of content for Nissan last year, Mr. Lambert said.

The presentation was one of three IQ Conference Insiders' Guide sessions, which were meant to give well-versed marketers new ways to take their data, personalization and automation efforts further.

Ralize that "people produce insights, not machines," became a key point at TBWA, according to Mr. Lambert. Marketers too often want to dump all the data they gather into a seemingly magic black box and then just ask the box to tell them what to do.

"What that means is that you are automatically looking at the same data set each time," Mr. Lambert said. "The first time you look at a new data set, you're a genius. Six months later, they've heard it before. After a year, you need to look for a new job," he said.

Marketers need to start with a question and then find the data to answer it, not the other way around, Mr. Lambert argued. "You can spend a lot of time data mining and not getting anywhere," he said. "But if you know what you want to discover, trust me, you can find that data. It's out there."

The most important question about data, however, is when use it, he said: A data-fueled post-mortem after a campaign shows you where you missed an opportunity but doesn't help as much as a data-fueled conversation before the campaign.

Jeriad Zoghby, managing director and global head of personalization for Accenture Interactive, reminded the audience of the 4 R's of personalization: Recognize, Remember, Recommend and Relevance.

Today's e-commerce gives consumers endless virtual aisles of options to shop from, but little to help them narrow their options toward a decision, Mr. Zoghby argued. In a physical store, a salesperson can help guide a customer to the right product or be a trusted resource for a returning customer. Mr. Zoghby said 40% of consumers say they have left a business website because they were too overwhelmed by options, and half of customers say they have never bought anything shown to them by a recommendation engine.

Those results suggest that many e-commerce players aren't doing what they hope they're doing with their marketing technology.

Rather than erecting a new challenge for consumers to struggle through, an effective e-commerce approach helps consumers solve problems by curating choices, Mr. Zoghby argued.

A good data management platform will recognize a customer so she doesn't have to remind a marketer who she is. From there, marketing tech should remember what she bought before and even why she made that choice. Was it the style and fit that sealed the deal, or the color? Experience optimization and testing tools can lead marketers to understand what is most important to the customer, and the relevance of what they find will be much more on target. That approach lets recommendation engines give consumers a hand instead of trying to cross-sell or upsell, Mr. Zoghby told the crowd.

"Successful personalization requires orchestration across channels and within experiences to drive performance and remove unintended barriers to demand," Mr. Zoghby said.

Judy Shapiro, CEO-founder of EngageSimply, walked attendees through the steps of marketing automation and a customer journey. "Marketing automation is the process of marketing, animated by technology," she told attendees. "Martech is the process. Ad tech is the stuff, the platforms, the tools, that enable the process."

Every campaign at every company will look different, but every campaign will encounter what she calls the "Gotcha Gremlins," the setbacks or small failures that will require a campaign to course correct. "'Gotcha Gremlins' are often annoying, but they are never fatal," Ms. Shapiro said, adding that once you know they will be there, it makes it easier to deal with once they pop up.

She then took attendees through a sample customer journey, starting with topic discovery. In her hypothetical, a mom is looking for information because her child wants to play football. Search engine queries lead her to articles describing how to keep kids safe while playing football. That's where SEO comes in. "Topic is closest to real-world intent," Ms. Shapiro said.

After topic discovery, the consumer hits category discovery, roughly midway through the so-called consumer marketing funnel. Here she finds out about safety helmets, and what they can do to help, so she starts reading about safety helmets and reaches out to friends for recommendations. That's where a topic-based programmatic ad can be deployed, by using cookies and profiles. Social publishing, social listening and influencer platforms can also be used at this stage.

The next step is brand discovery. "Here is where the rubber meets the road," Ms. Shapiro said. "This is where personalization kicks into high gear." Automated email personalization and automated web content personalization can be deployed and landing page info can be tailored. At this point, retargeting and refer-a-friend platforms can be used alongside loyalty programs.

Once marketers get to the moment of conversion, mobile ad traffic can drive the consumer to the store, and promotions and loyalty programs can offer what is most relevant.

"Create user journeys that you would love with authenticity, and that's when you reach martech joy," Ms. Shapiro said.