Skip to Content

Joshua Spanier is Google’s global marketing VP for media. He leads teams around the world who plan, buy, run, and assess media on behalf of Google’s brands. Here he shares his reflections on the past year of marketing in the pandemic.

Out of crisis comes learning. After a year of living through this pandemic, I’ve been reflecting on the marketing lessons that will stick with us. Like every brand, we’ve had to radically reassess our go-to-market execution. Early on, we developed principles that served us well as shutdowns began. Most useful was a back-to-basics digital marketing pragmatism that we leaned into as restrictions continued.

Now, as I reflect at this one-year mark, there are a few clear lessons from the pandemic that we’ll keep in mind, long after all of us have gotten the vaccine.

4 icons that represent digital marketing lessons from the pandemic. Toolbox: Embrace infrastructure and tools. Microchip & light bulb: Boost creative with insights and ML. Piggy bank: Win before spend. Looping arrows: Invest in and cultivate agility.

Embrace infrastructure and tools

Centralization, automation, tools, best practices — their mere mention can elicit cringes of corporate speak, red tape, and bureaucracy. Yet, when I look back on the past year, the infrastructure and processes we developed before and during the pandemic are what helped us pivot, relaunch, scale, and continue flexing quickly and efficiently.

Take campaign tracking, for example. Before COVID-19, we had upward of 20 different campaign trackers for our media teams all over the world. The quick pivots we needed weren’t going to happen with that disparate system, so we consolidated those trackers into one global view — a single source of truth.

Bureaucracy is bad. But infrastructure, scaled systems, automation, and tools are good.

At the same time, this pandemic has forced us, like so many marketing teams, to reckon with resource allocation. Looking to the future, we realize there’s more we can do to automate our processes to free up our humans to do what humans do best — drive insights and innovation. For example, we’re building a tool that will help us automate channel planning and publisher selection for our campaigns. Why? Our people spend lots of time crafting media plans for every campaign. But, at the end of the day, most media plans look a lot alike, and this past year has certainly made us question if they’re worth it.

Bureaucracy is bad. But infrastructure, scaled systems, automation, and tools are good. When we delineate those, we help teams see the difference between time wasted and time invested.

Supercharge creative with insights and machine learning

Sometimes there’s an allergic reaction to putting “creative” and “machine learning” in the same sentence. And believe it or not, this conversation even happens at Google. What we’ve found through lots of experimenting is that machine learning can actually make creative more effective, and more efficient.

Last year, with people staying home, digital shopping exploded, which presented an opportunity for our hardware products but at a time when shooting new creative was tough. To capitalize, we started with performance media and worked quickly to automate the build of thousands of creative assets. We tapped into first-party data and trends from Search and YouTube, incorporating those insights into our creative. For example, seeing that search interest in things like “cloud bread” were on the upswing, we automated our Google Store campaign to nod to those trends in the creative.

Un anuncio de The Nest Cloud Bread Buddy de la tienda de Google presenta una foto de producto con una imagen en su pantalla de las manos de una persona amasando pan. La copia del producto dice: "Disfruta de un pedacito de cielo". encima de un botón azul"

Similarly for Google Search and Maps, in less than six weeks, we were able to deliver 30 contextual videos and over 110,000 dynamic banners in support of more than 8,000 local businesses. To produce those kinds of assets, we built a machine learning–based feed that dynamically creates relevant and personalized variations (numbering into the 1,000s) based on dynamic levers, from geolocation to busy times to affinities. For example, if you lived near Chicago and were interested in dancing, you might have seen the ad below that featured local dance studios near you. That volume at that speed simply wouldn’t be possible without machine learning.

‘Win before spend’ is the new motto

With more scrutiny on every media dollar spent, this past year forced us to put more planning and tenacity behind our work, a lesson that’ll keep paying dividends long after the pandemic. Our teams developed and conducted preflight media checks to evaluate our campaigns against a list of best practices that had to be met for a given channel, tactic, or format.

For example, we know strong creative is critical to campaign success, and, like many advertisers, we conduct precampaign creative tests. We’ve gone one step further by developing a creative score card to track creative quality and the percentage of budget spent against creative that drives lift. This helps ensure we’re investing our media against the strongest performing creative. We also want to ensure we’re investing in finding the right user enough times to drive lift. Through the preflight check, we track our buys at a platform level to ensure that we have sufficient frequency to drive cut through.

Over time we proved that these checks can forecast the likelihood of success for a given media plan and deliver incremental lift. And following through on these checks can ensure up to 5X more likelihood of a campaign delivering impact.

Invest in and cultivate agility

The biggest takeaway of all? Resiliency for brands requires investing in experimentation. If we weren’t already comfortable pushing the boundaries of what it means to be digital-first, this past year would have been much harder.

I recall vividly a particular meeting that started us down this path. Six years ago, a creative agency came in to present work for an upcoming Google brand campaign. They slid TV scripts across the table and a 45-minute conversation ensued. I couldn’t help but imagine what our work could have been if we’d spent as much time centering the work in the promise and potential of digital marketing instead of the one-size-fits-all TV spot we were trying to amplify via digital channels.

If we weren’t already comfortable pushing the boundaries of what it means to be digital-first, this past year would have been much harder.

That meeting was a catalyst. It made us ask ourselves: How do we leverage digital marketing for what it is — flexible, contextual, fast, data driven — versus trying to make it fit into the traditional marketing model, centered on broadcasting messages in mass mediums?

From that point forward, we started building what we called “Team Adrenaline,” a team of creative strategists, digital evangelists, and innovative thinkers who help us make better work at the intersection of creativity, digital technology, and media. We set aside a budget for them every year to partner with internal marketing teams to push the boundaries on marketing as usual.

And the experiments born from this pivot over the years were what buoyed us when the world and our typical ways of working were turned upside down during the pandemic. For example, when shooting new creative wasn’t possible, Team Adrenaline stepped in to help us deploy YouTube’s Director Mix to make dozens of adaptations of video ad assets for our products, like Stadia, that happened to be more in-demand than ever. If we hadn’t already built up that muscle, we would have been starting from scratch in a time of crisis.

Relying on what’s worked before, without setting aside budgets and resources for experiments, is a dead end. And the pandemic made that more real for us than ever. With vaccine news that’s more promising by the day, I’m hopeful. But I’m also holding tight to the marketing lessons we learned from this crisis to carry us forward.