Customers sign up and don’t use the product

Justin Jackson posted a great thread around the idea that people’s motivation to use something is outside of our control. His point is that there isn’t much you can do within a product to get people to use it on a regular basis.

He’s not saying that you shouldn’t focus on great UX or improving your onboarding experience, just that none of it’s going to matter if the underlying problem is around motivation.

A product cannot motivate people to use it.

A company called Marketing Experiments made the same point with this little heuristic:

Conversion = 4 motivation + 3 value prop + 2 (incentive - friction) - 2 anxiety

The single biggest factor in this formula is motivation. It’s more important than having a great value proposition, an incentive, being easy to use, or clearing up any reservations around a product.

Basically, it doesn’t matter how good the food tastes if no one’s hungry.

I agree with the importance of a hungry crowd. I also disagree with the idea that there is nothing you can do in a product to motivate people.

There’s a whole industry called product development built on the premise that design and messaging can move people to use something. Also, anyone who’s seen ‘The Social Dilemma’ on Netflix would probably disagree with the hungry crowd hypothesis. If products can’t motivate people then there would be no dilemma.

So, if products can motivate people then how exactly does that work?

Well, the first step is to establish a baseline. How many people use the product on a regular basis right now? Without a clear reference point, it’s hard to know if more people are using the product or not.

Establishing a baseline#

Measuring how many people use your product should be straight forward but it gets a bit messy when you start trying to define what “usage” means.

The obvious thing to do would be to measure any type of usage. If someone logs in to close their account down or update their password should that count as usage? Or are we just talking about meaningful usage? Meaningful to the business or meaningful to the user?

Somewhere down this rabbit hole, you have to arrive at a core action connected to the value you deliver to your users. It will never be a perfect measure of the value you provide but that shouldn’t stop you from trying to quantify it.

Let’s keep this simple. If you’re a food delivery app let’s say the core action is ordering a meal. I know, what if the meal was ordered but never delivered? What if it was delivered but they didn’t enjoy it? If we just measure orders with 5-star reviews, does not rating it means they didn’t enjoy it? Rabbit hole 🐇 We need to keep it simple.

To establish a baseline for how many people use your product you measure out the total number of people that signed up in a day (or a week or a month). Then you measure how many people did your core action one month later (or whatever time frame makes sense for your product). You average these numbers out for the past year and plot them on a curve like so…

Screenshot 2021-10-03 at 3.11.53 PM.png

This is called a retention curve and it’s the industry standard for measuring how regularly people use your product over time.

In the graph above, on average, only 33% of the people who sign up are still using the product 3 months later.

The shape of this curve gives you a sense of the overall health of a product. If the curve drops to the floor it means that everyone who signs up eventually stops using your product. The flatter the curve the healthier the product. A flat curve means that a certain percentage of your users use your product on a regular basis. This is more traditionally called product-market fit.

Screenshot 2021-04-19 at 4.26.57 PM.png

Right, now we have a baseline, how do we improve it?

Figuring out the job-to-be-done#

The next step is to figure out what people are trying to do with your product. Sure, you might have a food delivery app, but what drives people to order food at home on an app? Is this just people who are too lazy to cook? Do people mostly use it as a catering service when they’re planning a party? Is it more of a family treat when people get bored of home food? Depending on the scale of your business, it’s probably going to be all of these to different degrees. It’s important to figure out what the main use cases are for two reasons:

  1. If you know why someone started using your app you can better fit the experience to what they’re trying to achieve.
  2. When someone has developed a habit around one use case you know what other use cases to gradually introduce them to.

Understanding why people use something is a combination of looking at usage data and uncovering behavior patterns, as well as interviewing people one-on-one and understanding why they started using the product.

Finding the right level of abstraction when dealing with motivational research can be tricky. On the surface, people order food because they are hungry. If you keep digging then everything we do is a means toward deeper spiritual or emotional fulfillment. Somewhere in the middle are jobs that people are hiring your product to do. Knowing what they want allows you to intentionally design your product to accomplish that job.

Better is a relative term. You can’t improve something when you don’t know what better means. Better for the home party crew might mean a wider range of options. More options might just make things confusing for college kids who just need a quick delivery. Understanding what people value is a prerequisite to making something more valuable.

Now that you know why people order food, what do you do with that information?

Funnels, Features, Notifications & Segmentation#

If you have a core action defined and you know what people are trying to do then you can just plot out all the steps they have to take to do the core action and see where people drop off.

You build a funnel and see where the biggest dropoffs are. if people have to do 5 things before they can order some food (like sign up and add their credit and so on) then you can measure where most people run into problems.

You come up with ideas about why people are dropping off at a particular step. Conduct a bunch of research and interview people if you have the time and resources. Then test out a bunch of solutions. If you have the infrastructure to AB test product changes then you can see what works and what doesn’t and iteratively fix product problems in this way.

You start with low-hanging fruit like a feature in the product that is not working or working but really slowly. You can progress to better messaging and guidance in the product. Sometimes this means design tweaks or better wording around things, other times you have in-app messages and notifications to help people along or you reinforce things with emails and reminders. Eventually, you hit a ceiling with this iterative approach, and fixing things just leads to diminishing returns.

This is when you reach for your jobs to be done and ask yourself if there is a better way to help people do the thing they are coming to your product to do. This might mean completely redesigning a part of your product or investing in a whole new feature set to help them get some aspect of a job done. Eventually, you hit another ceiling, and there are only so many features you can build that will appeal to everyone in your app.

This is when you start to look at the different types of jobs people are trying to do, or the different types of people that are trying to do a job, and you start to build more custom solutions for more focused subsets of users. Segmenting users in this way means that some users experience feature sets or emails campaigns that other users don’t even know exist. Product upgrades can no longer please everyone but certain tweaks make the product much better for some people. Who you focus on and how you prioritize your resources become the foundation of your specific product’s strategy.

How far is too far?#

Whether you’re smoothing edges and filing down points of friction or building new features, at some point you can cross an ethical line. You might make a product easier to use than it needs to be. When you start out you’re just using basic game mechanics to make things clear and engaging. You send out notifications to remind people to use something at just the right time. You make the point of a certain feature really clear. You use a beautiful, memorable design to show how something is done. Then you reinforce the behavior and you congratulate them when they do it right. Then you show people how many of their friends are doing the same thing. You start to vary the reward so people don’t always know what to expect.

At some point along the way, we’ve moved from being helpful, clear, and instructional to manipulative and deceptive. It’s a spectrum and the industry as a whole is too young to have arrived at a clear consensus on what’s ok and what’s pushing the line. Harry Brignull’s work on Dark Patterns is a prime example of the kind of conversations we need to be having as a product design community.

Use case transitions#

The problem with the hungry crowd hypothesis is that it doesn’t account for junk food. If food can’t make you hungry then why do cinemas pour insane amounts of salt into popcorn. Perhaps it’s so that you need that refreshing beverage to quench your thirst? A beverage so sweet that you just need a little more popcorn…and round and round we go.

Dark patterns and dirty little tricks aside, products can educate people. Sure, you signed up for a food delivery app because you’re too lazy to cook, but maybe you’d be interested in a 30 day cooking challenge where we give you a shopping list for the week and we show you how to cook a meal in 20 minutes each day. Products can transition from one initial problem to solving a much larger, or just different problem as people get further into the product experience.

A less dramatic example would be the fact that most of us started using Uber to get to the airport. Over time they managed to transition us to using Uber to get around town, some people even use ridesharing to commute every day. Starting with one problem and then gradually educating and helping people solve related problems is one example of an ethically defensible way products can motivate people to continue using them.

One could argue that the motivation to solve the related problem had to exist in the first place. Uber can’t make you want to commute, you’d only consider the use case if it was a problem you already wanted to deal with. But now we’re getting into semantics. At this point, we need a more precise definition of what ‘motivation’ and ‘intrinsic’ mean.

I do agree with the idea that the motivation to use something is largely outside of our control. However, motivation also comes from competency. If you help someone understand how to do something, and you help them get really good at it they become motivated. This kind of work is subtle and comes into clearer focus when you’re dealing with large numbers of users. On a more practical level, if I had to choose between working with the best Chef in the world or the guarantee of a hungry crowd everyday, I’d agree with Jackson and pick the hungry crowd every time.