Google exec shares 5 steps to SEM success in the age of machine learning


June 7, 2019 8:47 am

<a href=Google Chief Search Evangelist Nicolas Darveau-Garneau" width="1920" height="1080" srcset="https://searchengineland.com/figz/wp-content/seloads/2019/06/Nicolas-Darveau-Garneau.jpg 1920w, https://searchengineland.com/figz/wp-content/seloads/2019/06/Nicolas-Darveau-Garneau-600x338.jpg 600w, https://searchengineland.com/figz/wp-content/seloads/2019/06/Nicolas-Darveau-Garneau-768x432.jpg 768w, https://searchengineland.com/figz/wp-content/seloads/2019/06/Nicolas-Darveau-Garneau-800x450.jpg 800w, https://searchengineland.com/figz/wp-content/seloads/2019/06/Nicolas-Darveau-Garneau-550x310.jpg 550w" sizes="(max-width: 1920px) 100vw, 1920px"/>

Google Chief Search Evangelist Nicolas Darveau-Garneau

SEATTLE — Instead of fighting machine learning technology, marketers should harness its capabilities and work to make it smarter, Google Chief Search Evangelist Nicolas Darveau-Garneau said At SMX Advanced in Seattle on Tuesday during his keynote about optimizing campaigns in the age of AI.

The bottom line? Machine learning is here to stay – and marketers need to get on board if they plan to stay afloat in an increasingly automated digital landscape. Darveau-Garneau offered five key recommendations for winning in the age of machine learning marketing.

1. Measure together

Rather than measuring every customer interaction in a silo, marketers should view the data holistically, evaluating each touchpoint as part of a bigger picture. By measuring the ROI on each tactic and stacking it against the ROI across all of Google, marketers will be able to more easily spot opportunities and view the integrated value of all campaigns.

2. Optimize the right goal

Darveau-Garneau used the analogy of car insurance companies to explain how businesses should identify distinctive goals based on their unique audience and market value. The goal for Insurance Company A, he said, might be to cast a wide net and capture as many leads as possible, while Insurance Company B might choose to focus on selling a wide range of policies. Insurance Company C could have a more granular goal aimed at selling good policies. With a clear objective, machine learning technology can help brands pinpoint ideal consumers with greater efficiency.

3. Optimize the right metric

Results can differ dramatically depending on the goal being optimized, Darveau-Garneau said. It’s important to know which metrics to optimize when evaluating the holistic performance of the campaign. Since a machine can only learn from what a marketer tells it, it’s important for search teams to understand which metrics to test and tweak in order to paint a complete picture.

4. Optimize longer term

Brands that are only looking at short term goals and opportunities are missing a key piece of the puzzle, Darveau-Garneau argues. Focusing on the short-term means having to constantly prioritize acquisition. Marketers should integrate retention and loyalty optimization as key components of a sustainable, long-term cross-sell strategy.

5. Acquire the best customers

On paper, it might look like Google is targeting the ideal audience. But demographic and psychographic nuances can draw a bold line between high-scoring, model customers and average, middle-of-the-road leads. The difference comes from what Google knows about the goal, regardless of lead qualification. Finding the best customers requires SEMs to step away from the mindset of a sheet bidding strategy – and look instead at the acquisition process as an audience strategy. We need to be talking about customer buckets in order to determine customer lifecycle value, Darveau-Garneau said.

More from SMX Advanced 2019