The power of recommendations

Recommendations help us navigate choice in a world of abundant options. When coming from trusted sources, such as friends, they bring a sense of excitement around discovering something new or relief in not having to evaluate more options. Digitally, recommendations are at the cornerstone of some of the largest businesses in the world, such as Google, Amazon, and Netflix. Without the highly personalized recommendations they provide, we'd drown in a sea of options, struggling to know what to buy, what to watch, or who to connect with.

Navigating choice at scale is a superpower that unleashes our curiosity, awareness, and knowledge.

Take Spotify's Discovery Weekly as an example. Each Monday, Spotify recommends a set of songs that you might like but haven't listened to before. This is a mind blowing feature when you think about the amount of work it would take you to generate the same thing, if you could at all. Across so many aspects of our personal lives we now exert next to zero energy to connect with relevant products, people, and content — even though we have more options than ever.

Unfortunately, our professional selves lack this superpower. Better, faster decision making based on all the data we have at our fingertips still eludes us. Even after adopting digital tools, it remains frustratingly difficult to, for example, find the right content across them or to find the best person to connect with regarding a topic. This is all because we lack a recommendation engine that ties all our tools together, profiles us, and then offers personalized recommendations based on the data we couldn't possibly rationalize ourselves. Without recommendations, we're still reliant on "analogue" methods such as leveraging our personal networks or brute force searching to find the most relevant people and content. In a sense, at work we still make our own playlists by asking friends and doing research rather than Spotify suggesting one for us.

This is why recommendations are the critical factor that will unlock the true potential of enterprise digital transformations and allow us to optimize individual productivity.

Businesses will view recommendation engines as enabling human capital transformation, while we'll view them as catalysts for becoming the best version of ourselves. They'll allow us to maximize time spent on deep, thoughtful collaboration with a team and minimize the time spent wrangling the right content and team members together. Automating menial tasks has always been a key promise of AI; good recommendations are perhaps the most profound manifestation of this.

If recommendation engines are so powerful, why don't we see them today?
The answer is simple:

Your company doesn't truly know who you are. Your reactions, comments, engagement stats, etc. are all spread across dozens of apps and no one app has enough data to generate personalized recommendations accurately enough.
The ubiquity of the platform business model strategy exacerbates the above point. This is because platform companies are insular in their thinking and focus almost exclusively on making their platform as attractive to users and developers as possible. They're not focused on connecting data across tools to generate recommendations. To them this is someone else's job.
Business workflows are complex and almost always require several apps and a number of different people. Social apps aren't like this. They're more similar to appliances like toasters that do one thing really well. For example, Twitter generates a feed of tweets based on your interests, engagement, and followers. While still not easy, its scope is at least narrowed to the feed.

Based on the current state of the market, I predict it will take another 2-3 years before we see compelling examples of recommendation engines transforming businesses. In the near term, integrated vendors, such as Salesforce, Microsoft, and Google will act as pace setters and trickle out recommendations slowly. Microsoft provided an example of this last month when they announced improved Microsoft Teams search that:

"...makes finding messages, people, answers, and files faster and more intuitive...with AI-powered relevance based on the people and content you engage with most in Teams and other Microsoft 365 services."

Ultimately, to realize the power of recommendations, enterprise CIOs must mature across the following areas:

Digital transformation (most enterprises are here)

Covid-19 has accelerated the adoption of cloud based productivity tools across all industries. What's important to understand though is that modern cloud applications capture all kinds of data that will be used to produce recommendations. Data, such as reactions to chat messages, documents you share and with whom, time spent reading a document, who you meet with and for how long, etc.. As we adopt these tools and mature in our use of them, more and more valuable data will be generated that can be used to profile us.

Application modernization

Hybrid / multi-cloud is more than a contender for 2020 tech buzz-word of the year. The adoption of containers and container orchestration platforms such Google's Anthos or Red Hat's OpenShift, represent a likely 20+ year foundational architectural decision for businesses. The ability to write, test, deploy, and operate in a single OS, yet being able to operate on any is huge. Goodbye enterprise IT snowflake environments and hello rapid IT innovation. But what matters in terms of recommendations is that this modernization lays a common platform on which data that resides anywhere can more easily be unified.

AI platform

Enterprises will adopt an AI analytics platform that virtually connects data across clouds and on-premise systems / databases. Snowflake is a popular example of this type of platform, while IBM's Cloud Pak for Data is another. These platforms will contain the cross-app interaction / event data that's needed to generate meaningful recommendations.

AI based recommendations

With AI platforms in place and user, interaction, and content data unified, developers will be able to profile us and recommend things like:

how best to tailor your communications with a person, team, or organization
who best to work with
who to follow
what to learn next based on your skills, needs of the business, and/or what people like you are learning
who the experts are on a topic
who the influencers are in your company
what the most relevant information to you is
and others

In the end, recommendations are akin to the pot of gold at the end of the digital transformation rainbow. Enterprise CIOs and business leaders need to understand this and plan for it by adopting digital tools, modernizing their applications, and choosing an AI platform. The sooner they do this, the sooner we all gain our superpower!