Photo by Drew Beamer on Unsplash

Product Management in the ‘20s

How to PM in the wake of the tech backlash

Product Coalition
Published in
13 min readJan 28, 2020

--

From where I sit at this decade juncture, it looks to me like the 2010s will be known as the decade of the rise and fall of vanity metrics (engagement metrics, AARRR, etc.). No, they have not yet fallen, but this is a time of prediction and mine is that vanity metrics will fall very soon.

Why? Because they are at the root of the current problems (social isolation, polarization, superficiality, addiction, anxiety, child development concerns…) with consumer technology. It’s not that users don’t have control over their data, or control over themselves. It’s not that technologists are evil and/or greedy. It’s that we (product managers) define and measure success by KPIs like time on site, revenue generated, monthly active users, churn rate, conversions, clicks, and so on.

It’s cliche at this point but that doesn’t make it less true — we care about what we measure. You can’t tell people “meaningful interactions” while holding them accountable for “monthly active users” and hope everything just works itself out.

Or to put it another way, if you measure value creation indirectly (engagement or revenue generated) versus directly (user achieved goal you and she both wanted for her), you end up with the myriad of unintended consequences that are at the heart of the tech backlash.

<An example for those not quite there…>

(If you are already there, skip ahead!)

Let’s take the simple example of a web site that is about which kind of dog is right for you. When you are measuring things that are proxies for value creation (like time spent on-site) versus directly measuring value (the user decided which dog is right for her), as much as you intend to help the user make a decision, your organization will fall into the trap of increasing time spent without consideration for the real goal of figuring out which dog to buy.

It’s just too enticing to keep moving the metric. It gives everyone a jolt of accomplishment. For example: “We added a YouTube video inset for cute dog videos and they are now staying on-site for an additional 30 minutes! High five!”.

But what if your user decided which dog to buy in 10 minutes and then wasted 50 minutes watching cute dog videos instead of working out? Is this what you really wanted for them? No, of course not! Congratulations, you have just added to the problem. And you did it without gathering any user data at all (bonus points!). Just the simple signal of the user wants to buy a dog.

</End example>.

The fall is coming

How will vanity metrics fall exactly? It may be because users finally wake up and demand better products (note that I also predict users will never, en masse, demand more privacy), or because policymakers finally figure out how to regulate the attention economy. It may be due to advertisers boycotting social media because they are sick of being associated with it and some new ad supply arises. It may be because “big tech” gets broken up as promised by [insert democratic presidential candidate], or because new CS grads decide that making an impact is more important than living in the bay area. I don’t know (or really care) how. It’s going to happen, though. And possibly very soon!

So my question for you is: Are you ready to PM in this new world?

The mark of the successful PM of the 2020s will be learning how to create technology that actually makes the world a better place, instead of just making people less focused and less satisfied with their lives. Instead of making them more distracted, isolated, polarized, and outraged; all the while pretending that we are neutral and just giving people what they want 🤷‍♀️.

The good news is I think PMs are going to be essential instead of optional in this new world. Because it will be necessary to really understand the why behind the work, to measure success in sophisticated ways, to thoughtfully prioritize, to do user research… in order to build this post-tech-lash technology. The bad news is I don’t think many PMs are ready.

Here are the skills I think you are going to need in addition to all the ones you already need (sorry, this is a both/and, not an either/or 😘).

Intentionally choosing values replaces being “neutral”

The current stance of technologists is that technology is neutral; it’s the bad actors that are the problem. And yes, the bad actors are a problem. But they aren’t the main problem.

Let’s dive into this idea that technology is neutral.

You first have to see that as soon as you display something, you have expressed an opinion. A button that says “cancel” expresses the opinion that the user may want to change his/her mind about the action. One that says “OK” expresses the opinion that the user may want to do the action.

But those are just two options of many when you think about it. It’s bad UI so it may seem laughable but instead of the standard OK and Cancel, you could have buttons “you decided for me”, “ask me again later”, “pull a tarot card and do what the card says”, “magic 8 ball decides”, etc. You are nudging the user to make a decision about something and your interface provides them with the choices. That is not neutrality, it is choice architecture. You are pushing the action you want them to take. Your bias about what they should do (click OK!).

Still not ready to declare technology universally biased? Let’s look at search — the canonical neutral platform — as another example.

When search results are displayed, they are displayed in order. With the implication that the ones on top are more… what? Google will say “relevant” but isn’t it also “important”? Perhaps even “truer”? By deciding which one is displayed first (or after the ads, assuming the user in question has successfully trained him/herself to ignore those), Google and Bing are choosing the criteria by which “better result” is defined. It’s the entire point of SEO — to figure out what Google thinks is “better” and change your site to meet that criteria.

Can you choose what’s best and claim neutrality?

If your mom hangs up your green dress so its the first thing you see as you are getting ready for church, is she neutral? 👗

The same goes for the FB news feed, Instagram feed, Twitter feed. They are picking for you based on some criteria. They may reasonably object to being biased in a particular way (liberal or conservative, for example), but they are biased.

Hopefully, you are convinced now that there is no such thing as neutrality in technology. What do you do with that?

The PM of the 2020s is going to understand this instead of resisting it and will use it with intention. He/she will lead discussions about values within the team and communicate what values the product is choosing to project onto its stakeholders (not users anymore, see below).

The search PM will talk about the values of the search platform and how ranking can express that. The UI PM will discuss the choices being offered by the interface and the values behind those choices.

The PM of the 2020s will innovate ways to test whether the product is succeeding or failing at these values. Hint: it will almost certainly involve qualitative research.

Finally, he/she will use these values in prioritization discussions and for idea generation. For example:

  • Which of these feature ideas most helps our users connect?
  • What more can we do to help our users broaden their perspective?
  • Which of these features most helps our users stay focused and accomplish their goals?

If you’ve ever been in discussions like these, it’s also way more fun than talking about which feature will generate more revenue. 🙌

Stakeholder analysis replaces personas

You may or may not use personas today. Either way, the new PM will need to get sophisticated about broad stakeholder analysis instead. Personas are an attempt to scale the design thinking step of empathy building. They are a way to help the whole team understand that they do not represent the user. “Soccer mom” and “Grandma” and “Teenager” personas describe people so we can build with them in mind without “being” that person or having direct knowledge.

But the next generation of technology needs to go further to avoid the mistakes of the 2010s. It is no longer just about your users, but a broader set of stakeholders. You will need to consider the impact of your products on the people in the lives of your users: the mom of the teenager addicted to gaming, the person receiving the email or text, the business the electric scooter gets dumped in front of…

And you will need to consider society as a whole as a stakeholder as well. Why? Because the harms that have been created don’t just show up in individuals, but in society at large. Polarization, for example. Or the influencer culture.

Photo by Nicolas Picard on Unsplash

Systems thinking replaces design thinking

I love design thinking. In particular, the first step of building empathy for users before attempting to solve problems for them. But the post-tech-lash world will require systems thinking that does not solve problems in a vacuum (which is what is encouraged by design thinking — sorry IDEO).

PMs will need to have the ability to play out the impact of the technology they intend to put out into the world and watch carefully for consequences that are misaligned with the intended values of the product. Instead of throwing stuff out fast and furious and A/B testing it with success defined solely by vanity metrics.

The canonical example here is the “like” button. The inventors of which have publicly expressed regret for their part in creating our current societal state, aka the “influencer” culture. I appreciate their remorse, but in the future, we will need to hold ourselves accountable for impacts like this, not just feel sorry about them.

Not getting my point? The team that’s great at systems thinking ideally doesn’t ship the Like button because as they analyze what wild success might look like, they see how it changes the shape of what matters to people and see that is inconsistent with their values (meaningful social interactions in the case of Facebook).

That bar (proactive discovery) is high though, and it’s not possible to predict these negative consequences all the time. Systems thinking is more about the understanding that these unintended, negative consequences are unavoidable (because technology isn’t neutral) and a commitment to finding them and rectifying them.

How do you do that? You make this core to your definition of success. You define success by how much positive impact you are creating and how little negative impact you are creating. Instead of engagement, revenue, MAU, etc.

You stop thinking in terms of “trading off” traditional definitions of success and harm creation. No harm is success.

When you have this mindset, if you accidentally ship the like button, once you see the impact it is having, you address it. You remove it or limit its use or… I don’t know the right answer… but I do know that if the smart people on that team have internalized this new definition of success, they will figure it out.

Red-teaming, the process by which you find negative consequences before you ship, will be an area of innovation as well. Teams that find the balance between discovering consequences in advance and getting products out the door in a timely manner will be the most successful. PMs will need to carefully manage this balance.

Expertise in what you intend to improve replaces “move fast and break things”

Technologists today value fast. This value, contrary to the story they will tell you about it, is driven mostly by fear, hubris, and laziness. The fear that someone else will get there first and then “we” don’t “win” (that reminds me, please go read Simon Sinek’s The Infinite Game and perhaps you will agree that there is no “winning” anyway!). The hubris that more technology is always the answer and that if it can be done it should be done. And yes, also laziness. It’s a heck of a lot harder to truly figure out if something is worth doing than to convince investors (or execs) your idea is amazing. Or to just do it and see what happens (ultimate laziness). See more on this here.

In the ’20s, we will start to value thoughtful, informed, and intentional. Maybe even humble. When we consider offering a new product into the world, we will first learn about what we are trying to evolve. Twitter will deeply understand communication, conversation, and the public square. Facebook will deeply understand relationships, community, and connection. Uber will deeply understand transportation. Because a deep understanding of the system you are acting on will be necessary to employ the systems thinking just discussed.

As a result, the systems we impact will more often be genuinely better because of our technology.

Compassion replaces empathy

Hopefully, I’ve convinced you that we can’t just throw stuff out there and use A/B testing to see if it “works” because we need to satisfy deeper needs and values instead of surface wants. And that no matter how careful we are, we may still screw up significantly. So what do you do about this? The answer is compassion.

Let’s go back to design thinking and the first step of building empathy through research. Sadly, this concept is still new or unknown to many technologists. But I can’t help that. It’s not going to be enough anymore.

What we need to build in the 20s is beyond empathy for users. We need to build compassion. Empathy is understanding another’s perspective. Like, “Hey, we are building a transit product but all of us drive to work, maybe we should go talk to people that use public transportation”. The goal of empathy is to understand, gain insight, and innovate based on that insight. All good things, so don’t stop that!

But to get to the point where you are ready to negatively impact your short term business metrics because you know that long term success relies on truly making a positive impact, you need more than empathy. You need compassion.

For example, if you are Instagram, you need to deeply care about the stress of being an influencer and having to pretend your life is perfect. You need to deeply care about the desperation of not being one and feeling like the influencer lives actually are perfect.

You need to feel compelled to change your products when you see pain being caused in your now broadly defined stakeholder set. You need to really want to understand what impact your product is having on society in order to spend the resources it will take to develop this understanding.

This is what compassion is — the need (not want) to see what impact you are having and the motivation to address it. Without compassion, you’ll just pick another metric to optimize that seems better than engagement and I’ll be writing about the fall of those metrics in 10 years.

Just like with empathy, once you have it, you will find yourselves making different decisions. This isn’t about a formula anymore and while it may feel like a leap of faith, the research is available if you look for it. I’m not likely to invent a compassion-driven framework (never say never 😏) to follow so you are going to have to figure this out for yourselves. Which leads me to the last skill that underlies all the others.

In summary, shared context replaces being data-driven

Reading between the lines of the above you may have noticed that I’m not giving a lot of steps or specific actions or tools. Instead, there is a call to curiosity, intentionality, research, understanding, and remaining open. I almost sub-titled this section “intuition replaces being data-driven”, but there is so much disagreement about intuition I thought you’d skip the section.

Today, vanity metrics are popular because they are so clear. OKRs set around vanity metrics are aligning because they are easy to understand. All of this data is just so… deliciously objective. We don’t have to disagree with each other. We don’t have to take the time to build shared context. It all seems so efficient. If you are in an organization that is sophisticated about data and have been grappling with data-driven decisions for a few years, you know this isn’t really the case. But many technologists still think that it’s true.

In this new world where vanity metrics have fallen, we need to embrace ambiguity, shared context, vision, and values. All of which are much squishier things than KPIs and OKRs. Right-brained skills instead of the oh-so-measurable and therefore oh-so-enticing left-brained ones.

This means we have to talk to each other instead of just “doing”. We will need to engage in healthy conflict. Psychological safety, EQ and the ability to be vulnerable with each other will be as important as the research has long been telling us it is. And again, all of this is on top of being a great PM today — business skills, technical skills, project management skills…

I’m not here to tell you any of this is easy. It really isn’t 😳. I’m here to tell you it’s going to be necessary. You can hope I’m wrong about the fall of vanity metrics. Or you can start building these skills. Your call!

Good luck and happy innovating 😘

--

--

Product management leader (Apple, Microsoft) | Mentor | Lifelong Learner