New Dimensions for Scale and Learning

In the world of Customer Acquisition 3.0, no longer will scale represent only the traditional value of achieving cost leadership and optimizing the provision of a stable offering. Instead, scale will create value in new ways across multiple dimensions: scale in the amount of relevant data companies can generate and access, scale in the quantity of learning that can be extracted from this data, scale to diminish the risks of experimentation, scale in the size and value of collaborative ecosystems, scale in the quantity of new ideas they can generate as a result of these factors, and scale in buffering the risks of unanticipated shocks.

Learning has always been important in business. As Bruce Henderson observed more than 50 years ago, companies can generally reduce their marginal production costs at a predictable rate as their cumulative experience grows. But in traditional models of learning, the knowledge that matters—learning how to make one product or execute one process more efficiently—is static and enduring. Going forward, it will instead be necessary to build organizational capabilities for dynamic learning—learning how to do new things, and “learning how to learn” leveraging new technology and vast data sets.

Today, artificial intelligence, sensors, and digital platforms have already increased the opportunity for learning more effectively—but according to BCG, competing on the rate of learning will become a necessity by the 2020s. The dynamic, uncertain business environment will require companies to focus more on discovery and adaptation rather than only on forecasting and planning. Companies will therefore increasingly adopt and expand their use of AI, raising the competitive bar for learning. And the benefits will generate a “data flywheel” effect—companies that learn faster will have better offerings, attracting more customers and more data, further increasing their ability to learn.

However, there is an enormous gap between the traditional challenge of learning to improve a static process and the new imperative to continuously learn new things throughout the organization. Therefore, successfully competing on learning will require more than simply plugging AI into today’s processes and structures. Instead, companies will need to:

• Pursue a digital agenda that embraces all modes of technology relevant to learning—including sensors, platforms, algorithms, data, and automated decision making

• Connect them together in integrated learning architectures that can learn at the speed of data, rather than being gated by slower hierarchical decision making

• Develop business models that are able to create and act on dynamic, personalized customer insights

Never before have marketers had access to more customer data. The first-party data companies collect with user profiles can go beyond basic name and demographic data and might include downstream rich data points on engagement, retention, monetization, and much more; companies can use this to build great user segments for running prospecting and retargeting campaigns for growth teams. Ingesting and processing all this first-party data from brands layered on top of the existing rich user data enables these media partners to perform sophisticated modeling and analysis with machine learning that wasn’t possible even a few years ago. This results in better targeting with new insights and data analysis.

If you are still manually optimizing campaigns the same way it was done half a decade ago, you may find yourself among a quickly disappearing breed in the customer acquisition game.

If you are still manually optimizing campaigns the same way it was done half a decade ago, you may find yourself among a quickly disappearing breed in the customer acquisition game. Any manual process is likely much less effective and far more prone to human error than the new solutions quickly emerging to attack inefficiencies.

AI and Customer Acquisition

The accelerated adoption of AI for customer acquisition by major media platforms like Google, Facebook, programmatic ad networks, and many others represents a fundamental and pivotal transition in the way that marketing dollars are invested in mobile marketing campaigns. No longer do growth marketers have the ability to choose where or how their ads are shown to users—instead, algorithms decide these logistics, guided by few inputs, such as bids and budget. While that may be good for most growth teams, some of the most intelligent growth marketers in the industry are looking beyond the obvious ways AI can improve results to focus on the cutting edge “out of the box” ways AI can turbocharge their paid user acquisition performance. Companies like IMVU, Netflix, Lyft, and others are pioneers leading the way on the AI frontier, both on their core offerings (entertainment, recommendations, efficient routing, etc.) and on the customer acquisition front. They’re at the forefront of using intelligent machines to automate actionable insights to fully manage their paid acquisition campaigns with fewer human dependencies.¹

It’s Time to Turn on the Intelligent Machines

At the end of the day, the best way to evaluate any emerging technology is to figure out its practical use in your business or industry. Just like good user experiences are personalized for an individual’s needs, the future of scaling customer acquisition will be won by startups who can adapt each platform’s out-of-the-box artificial intelligence solutions to fit their needs, objectives, and goals. Successful startups have learned the importance of focusing on the right metrics and key performance indicators (KPIs), which are measurable value that demonstrates how effectively a company is achieving critical business objectives. Examples of KPIs are customer acquisition costs (CAC), return on ad spend (ROAS), daily active users (DAU), monthly active users (MAU), retention, churn rate, and so on.

AI-powered machines (which we’ll explore in Chapter 5) can help orchestrate acquisition campaigns that more efficiently move toward these goals compared to the relatively brittle process of manual campaign intervention. This requires a holistic cross-channel approach, which massively increases operational complexity—from data-driven targeting to creative proliferation to attribution and performance optimization. And with complexity comes exactly what you don’t want: risk and uncertainty.

Sooner rather than later, your customer acquisition efforts will rely on artificial intelligence, machine learning, and automation (which you’ll learn about in Chapter 4) to adapt, customize, and personalize cross-channel user journeys and deliver optimal results in ways that would be impossible using last-generation business intelligence and dashboards. Managing complex, cross-channel campaigns with multiple targets, creatives, and sequences will require an intelligent machine operational layer above the out-of-the-box solutions to deliver great results—or you may have to settle for being average.

Most companies find a comfort zone with one or two major channels and skip the rest. But each of the big platforms have different advantages:

• Snap skews younger.

• Pinterest has a higher composition of women in their audience.

• LinkedIn is where people conduct business activities.

• Instagram’s core audiences are highly engaged and tend to interact on the platform, which is great for educating consumers and building audiences.

• Search is all about lower funnel intent.

• Demand-side platforms (DSPs)² to reach people outside of the all the other platforms listed here.

Taking these factors into account is important as you develop your strategy. I always recommend managing a broad, diversified mix of different platforms for your customer acquisition, to reduce your business risk of ever being highly dependent on any one single source like the duopoly of Google and Facebook. Too many startups spend their entire budget on the Google and Facebook black boxes with very limited visibility and poor understanding of how the algorithms work or how they change. Imagine what would happen if all your budget was going into only Facebook or Google and their algorithms changed without any notice in such a way that could significantly impact your ability to acquire new customers?

Always continue to invest at least 5% to 10% of your monthly user acquisition budget into testing new channels every month. By leveraging a portfolio of platforms for prospecting, you are able to get a varied mix of users into the system, which the artificial intelligence can then manage for retargeting—a funnel filled with much less effort across a variety of platforms.

It is also important to note that there will be distinct and shifting bid dynamics on different networks even within month-long campaigns (of course, they’re all subject to seasonality). It’s very hard to take full advantage of these shifts in bid pressures across the entire user journey from prospecting to retargeting without leveraging your own fully customized intelligent machine that can be trained to operate and automate budget orchestration a layer above each individual channel for truly dynamic cross-channel optimization. No human user acquisition managers and growth marketers can ever outperform a fully automated intelligent machine because the machines can process, analyze, and take the predictive actions 24-7 without ever needing to take a day off. Humans are prone to making mistakes, changing jobs, and/or needing to sleep, which always gives the advantage to the intelligent machines. The AI machines would only get better with more data to help perfect the training of their algorithms over time.

You must learn how to train the algorithms to control your key campaign optimization levers (Chapter 5), focus on creative and strategy (Chapter 8), and turn the drudgery and math over to the machines to get data-driven results far beyond manual capabilities. The rest of this book will show you how by drawing on examples from IMVU and other cutting-edge startups who are successfully doing this now and reaping the benefits.

For example, IMVU has strict KPIs around CAC and ROAS. By leveraging AI, we were able to take advantage of cross-channel efficiencies, and improve the KPIs dramatically across the board. While channels will vary in their CAC and ROAS every month, by leveraging smart machines to manage against our goals, we were able to effectively allocate our resources to take as much advantage of each channel while keeping performance within our key metric zones. By prospecting across the different channels, our retargeting rates for incremental in-app purchases has increased dramatically.

The future of Customer Acquisition 3.0 rests on the shoulder of intelligent machines, orchestrating complex campaigns across and among key marketing platforms—dynamically allocating budgets, pruning creatives, surfacing insights, and taking actions autonomously. These machines hold the potential to drive great performance with a far more efficient Lean team, hands-off management approach powered by artificial intelligence.

Now that you understand of Customer Acquisition 3.0, let’s talk about one of the most fundamental aspects of Lean AI for marketing: automation.