As the largest gathering of product leaders, Amplify is where hundreds of teams go every year to learn how to lead digital transformation and build the best product experiences.
In 2020, this mission took on new meaning.
This year, digital acceleration took hold overnight and affected every company imaginable. Large or small, old or young, digital transformation is no longer just an idea. It’s every business’s reality.
At this year’s Amplify, we adapted the traditional elements of the conference—talks, networking, session breaks and swag—into a fully online experience. Attendees connected on Slack, watched live streams on social media, and gained access to educational sessions on Amplify’s new digital platform.
The live sessions brought together luminaries from companies like Peloton, IBM, Atlassian, and AB InBev to share their stories of adapting—and winning—in a digital-first world. Here’s a rundown of each session and their key takeaways.
The Amplitude Keynote
This year’s keynote began, as usual, with an address from our CEO Spenser Skates. Spenser made the stakes of the post-pandemic economy clear: “Our reliance on digital products has only increased.” He identified three opportunities for business leaders:
- “Monetize the moment.” In 2020, there were more than 130 days where more than $2 billion was spent on ecommerce. Before 2020, there were two such days per year. Are businesses positioned to capitalize on this substantial new ecommerce opportunity?
- “Retain to sustain.” Current customers are key to growth in the post-pandemic climate. How do businesses best retain them?
- “Seize the signals.” Four out of five product managers say users have developed new and lasting habits. How do we assess and build better products around these lasting habits?
We then heard from Justin Bauer, executive vice president of product at Amplitude, and Shadi Rostami, executive vice president of engineering at Amplitude. Justin began by explaining that “the change to digital everything has propelled us into a new universe where product is no longer a channel, but instead product is the business.”
We also heard how Amplitude helps customers break through the builder’s paradox. This paradox holds that developers typically have to either build fast, based on broad understandings of user behavior, or build more slowly, with more granular data relative to user behavior.
Justin and Shadi emphasized that companies can do both—with a little help from Amplitude. We then got a peek at Amplitude’s new product, Journeys, which helps companies see their product as their users see it. Journeys provides a high-level view of how customers move from one state to another, like from activated to subscribed, as well as detailed individual user behaviors.
This, as Shadi highlighted, is key to understanding “what is driving value in your product.” Companies with this understanding can look forward to both building fast and building with the full benefit of data.
“The change to digital everything has propelled us into a new universe where product is no longer a channel, but instead product is the business.” Shadi Rostami, EVP of Engineering at Amplitude
We also heard from two product leaders about how they use Amplitude to drive growth at their businesses. Ivan Galea, VP and head of analytics and data science at Atlassian, told us how Atlassian has benefited profoundly from using Amplitude. Thanks to our technology, they’ve been able to map out workflows, validate hypotheses, and explore adjacent flows to deepen their understanding of their customers.
Kurt Williams, global director of consumer products at AB InBev, spoke about how Amplitude is helping his company—the world’s largest brewer—with the switch to digital. Amplitude helped remove data silos and inefficient processes at AB InBev. With Amplitude, they have better data literacy, see customer segmentation more clearly, and make faster decisions.
Peloton’s David Packles on Building User Understanding
Next up, David Packles, Peloton’s director of product, broke down how to use customer data to build products that appeal to a wide variety of users.
David’s theory of building a product for a lot of users, based on data from a few, was based on the idea of emergent behaviors. Emergent behaviors are usage patterns that appear organically within a highly engaged subset of your community. You can use the data from this engaged subset to improve the product for the rest of your users.
Like many of our speakers, David emphasized the benefits of taking risks and experimenting in this tumultuous period in the world.
David noted, “The beauty of good product ideas is that they sound obvious after the fact.” Getting to the point where the present looks like it was inevitable requires courage. It also requires that a company know how their most engaged members use their product and how to build a product that supports these behaviors.
Panel: Growth in a Pandemic
This year’s panel, “Growth in a Pandemic,” featured Sandhya Hegde, Amplitude’s executive vice president of product. She chaired a discussion between representatives from Shipt, Imperfect Foods, and Shopify:
- Vinay Bhat, VP of data science at Shipt
- Patti Chan, VP of product at Imperfect Foods
- Phillip Rossi, director of data science at Shopify
Shipt, Imperfect Foods, and Shopify are examples of companies that adapted when the pandemic first struck. Vinay, Patti, and Phillip told the audience how their leadership, data, and scaling approaches changed in its aftermath.
Patti told us that at Imperfect Foods, “We got a COVID task force together, where leadership from every department met every day, twice a day, to stay on top of the rapidly changing nature of our business. Our operations were scaling very quickly, much faster than we had planned for.”
Vinay told us that Shipt’s focus was on agility. “Where our process used to be a little bit more rigid [with] not so much human in the loop, starting in March, our forecast team became much more nimble, and there was a lot more human input.”
For Phillip and Shopify, data scientists were the key to adapting.
“We needed to rethink our road map,” Phillip said. “We’d set up a task force of data scientists that normally embed themselves within different business units of the organization to come together and help us tell the story of what was happening as the pandemic played out.”
Each of the three businesses also changed their forecasting frequency. Where plotting using forecast data had once been a quarterly or annual practice, they were now doing it daily to handle the speed of market shifts.
IBM’s Nilanjan Adhya on Driving Digital
Nilanjan Adhya, chief digital officer of cognitive applications at IBM, joined Amplify to discuss IBM’s challenges and successes in transitioning into life as a digital company. He spoke about this legacy company’s need to create and scale an experience-focused business model. The company also faced a large enterprise’s difficulties when trying to deliver those microservices at scale, otherwise known as “going small.”
Adhya also shared IBM’s three transformation principles:
- Daily improvement. “Our first transformation principle at IBM? Improving by 1% every day.”
- Escaping the gravity of decision paralysis. Aiming at “achieving liftoff” on projects and not feeling the need to pre-plan everything before getting started.
- Experimentation in making decisions and approaching problems.
Nilanjan ended by encouraging businesses to build infrastructure that empowers experimentation. This requires integration across the centers of engagement of marketing, sales, product, and customer success. It also requires good data infrastructure and a tool for achieving a singular view of your customer.
Patreon’s Maura Church on How You Can Become a Cross-Functional Data Leader
Maura Church, director of data science at Patreon, emphasized that data scientists should interact with all departments, not just product and marketing.
Your data scientists are fundamental to your operation, but they won’t necessarily have the experience to make significant contributions across different verticals independently. Maura encouraged an approach that involves data scientists sitting in with teams to learn their rhythms and find new points of potential assistance. “Make sure [your] data scientists are sitting with the teams, are embedded in their decisions and in what they’re trying to build.”
Teams can then create more robust, more effective data partnerships by sharing information among themselves, ensuring data relevance, and curating good data storytelling.
Squarespace’s Jonathan Hastings on How to Embrace Hypothesis-Driven Product Development
Jonathan Hastings, group product manager at Squarespace, addressed how the company is tackling the post-pandemic economy by embracing data transparency, data visibility, and data democratization.
Squarespace did so by adopting a hypothesis-driven product development, which allows for a more experimental approach. They realized the flaws in their old method of designing for people who “[were] just like us.” This narrow conception led to ineffective onboarding and lots of missed opportunities.
In response, they developed a new way of working based on shared frameworks, revised North Star metrics, and a design approach based on expected outcomes.
The result? More creative freedom and better outcomes. Sometimes a leap of faith is required to make real progress. “It’s hard to imagine working [in a new way],” Jonathan said, “until you’ve tried it.”
“[Squarespace has] ritualized learning as the key outcome of product development to orient experiments towards revealing information about users and the market,” Jonathan said.
WWT’s Noah Rosenberg on Leveraging Data to Break Out of the Big Box
Noah Rosenberg, director of digital growth at World Wide Technology, explained how retailers can build a frictionless and rewarding shopping experience for customers.
He emphasized the need for brick-and-mortar companies to approach digital with gusto by optimizing customer retention and getting data flowing. Noah said that, in some cases, this requires the development of a new way of approaching things. “Now digital, which was a companion to your physical advantage, [has become] the number one way your customers are going to reach you.”
Noah noted the pressures on non-digital retailers. Suddenly, they have to move from non-specialized “broad-brush” marketing to a more user-specialized approach in order to try to beat the digital natives at their own game.
He encouraged businesses in this position to think about this transition to digital as a journey. Data-driven decision-making means starting with fundamentals and going forth.
His advice for digital growth? Get data flowing by refining your knowledge of your space and customers. Let insights from this data flow bloom, uncovering customer knowledge through collaborative exploration. Then, take targeted action to stimulate growth through improved customer knowledge.
Damien Delautier of Canal+ on Implementing Better Product Frameworks in a Non-Product-Centric Company
On our second day of Amplify, we heard from Damien Delautier, chief product officer of renowned French media company Canal+.
When Damien came to Canal+, it was not a product-centric company. “The role of engineers, the role of products was not really well understood,” Damien said. There was too much focus on how much was being produced, not on how effective products were. “We were not tracking the results of the project we were launching,” Damien said.
Damien surveyed this state of affairs at his new company and asked himself, “How can we innovate faster? The answer is in product management.” He then helped revolutionize Canal+ by redoubling their focus on output metrics and user retention and achieving business growth through a product-led philosophy.
Canal+ then transformed its product culture by strengthening the six foundations of product development: product strategy, product discovery, product delivery, product team, product culture and product intelligence. Damien sought to mobilize skills and specializations from around the company to increase the quality of output. “It’s not only about the product manager and designer that need to conceive and develop something,” he said. “We also need skills from all over the company.”
He emphasized the value of empowering the team through co-creation, maintaining “customer-centricity” and, once again, encouraging experimentation when driving business outcomes.
Pratilipi’s Shally Modi on Building an Impact-Focused Culture
On the third day of this year’s Amplify, we heard from Shally Modi, a cofounder and the head of product at Pratilipi, India’s largest online self-publishing and audio-books portal. Despite hosting over a million users, India’s “YouTube for writers” has a relatively small team. So how did they scale while keeping their team lean?
The key was a combination of qualitative and quantitative decision-making techniques. Some companies prioritize talking to users and taking user interviews over statistical analysis, while others go for the statistics-only approach. Shally highlighted the benefits of taking a combined approach. “It’s not just about looking into . . . what your users are talking about, but also taking those insights and putting our inferences on top of that [via] behavioral analysis,” she said.
Pratilipi’s UX and product design approach stems from the rich data Pratilipi acquires on both analytics fronts: qualitative and quantitative. It allows them to be highly responsive to new user wants and needs while remaining cost-effective. “We take the inferences down, and then we make an informed decision, [an] informed approach,” Shally told us.
Adapting to the Digital Era
At this year’s Amplify, there were a few common themes: Digital transformation is here. In order to forge ahead, retain customers and build the products that win, you need to understand your customers’ changing needs and habits.
Understanding customer behavior is still a challenge at many businesses, though. The newly released Product Intelligence Report lays out this problem: interviews with 350+ business leaders reveal that companies are seeking to act on digital insights and move fast, but are often limited by access to reliable data about customer behavior.