The Complex Psychological Hooks I Used to Build a "Simple" Side Project
I recently shipped a product for Valentine’s Day called Good Morning Love. It’s a seven day drip email campaign where you write lovey-dovey notes that get sent to your partner every morning from Feb 14 - Feb 21, 2015. If you want to sign up, it’s not too late - click here. The concept might seem fairly simple, but there were a lot of decisions that went into shaping the product and user experience. I want to go behind the scenes to share how I built psychological hooks into the product and why it’s important to do so.
Why nuances in a product name matters
I got the name of “Good Morning, Love” almost immediately. Names can be tricky, but luckily this came to me almost right away. The name works because it’s evocative of a scene that is in many romantic comedies and touches upon sweetness and tenderness between a couple. It implies that if you spend the night at someone’s apartment, that they are there in the morning cooking you eggs for breakfast.
I didn’t want the name to be overly romantic, because I wanted to have male AND female users. If it’s TOO romantic, it can be barf-tastic to guys who think, “God, why does my girlfriend make me do things like this?” I aimed for a name that would appeal to both genders, but was still reminiscent of loving feelings.
Calling someone “love” is also sweet without being overly sugary. The phrases would have been too predictable:
- Good morning, honey - a middle aged person would say this
- Good morning, sweetheart - again, above
- Good morning, sunshine - an overly energetic early bird PE teacher would say this
- Good morning, sugar - a waitress at a diner would say this
- Good morning, love - Hugh Grant would say this to you. In his kitchen. Wearing a heather grey t-shirt and glasses.
Also, it’s easy to imagine a British guy calling you “love,” and that’s always romantic and just a bit unexpected.
How your logo affects user perception
There are a few choices here:
- Do you want an icon, yes or no? I.e. Nike’s swoosh or McDonald’s arches. My answer was no.
- Do you want a logo? The logo is the typeface. Yes, I wanted a logo that could be a visual identity for the product.
With the logo, more choices:
- Sans serif (like Arial) or serif (like Times New Roman)
- All capitalized, all lowercase, normal capitalization for each word
- Rounded, narrow
- Aesthetic: bold, energetic, confident, whimsical, delicate, old world, hefty, etc
Originally, the logo was in all caps, sans serif. The look of the font here was confident…but too confident. It wasn’t approachable enough. It’s like a finance douchebag who won’t hold your hand in public.
Then I changed it to all caps, serif. This softened the edge a bit. The logo font now was more traditional because of the serif. It was more pensive, like a guy in glasses.
But the real win was when I realized that the font should be in all lowercase. It was an epiphany and a glorious moment. Changing the font to all lowercase made a huge difference. Suddenly, the brand became more approachable while still confident and understated.
I also chose colors to communicate the feel of the site - that this product was for young couples and Gen Y. The charcoal grey for “Good morning” was less severe than black for viewing on the web as a general design best practice. I chose a cool medium pink reminiscent of Skitch’s pink - bold but not neon. Confident but not over-the-top. Not too soft like a baby pink would have been.
Deciding who your product is for
I decided that the target market would be Gen Y, self-identifying nerds, marketers, and people at tech startups. This was an important decision because it drove the snarky tone of the copy, the references to startup and marketing industry terminology, the length of the form (assuming people don’t have all day to do it), etc.
Deciding “who is this product for?” is a crucial decision. You have to be clear on who your product is serving. This allowed me to better process feedback from early users who tested the site before we launched.
For example, because I was clear on who my audience was, I was able to ignore the feedback from one person who said the copy could be offensive. I knew that that person was in the minority and that Gen Y prefers brands with personality rather than something bland and forgettable.
Also, I had no interest in pleasing everyone. First, you can’t please everyone so it’s futile. And second, I only wanted to cater to my target audience of users.
It’s okay if some people don’t get your tagline
The tagline was “Nothing says love like a drip email.” This is very specific to the target audience. Not everyone knows what a drip email is, and not everyone would think that mentioning marketing automation was clever.
But I had a clear definition of who my target audience was, and I wanted the product to speak directly to them and cut through the clutter.
Use analogies whenever possible
When a user encounters your site or product, they subconsciously ask themselves, “What does this remind me of? How am I supposed to use this?”
There are usually many cues that answer those questions. The goal is to make sure the cues are in place, and that they point in the direction that you want the customer to think.
In this case, I explicitly said, “Does this remind you of something? Why, yes, it does. This is like one of those Pinterest ideas where you put cute notes into a mason jar, and feelings of love will burst into sparkles.”
I used the analogy that this is like a digital version of a popular Valentine’s Day art project.
This is a powerful and convenient format to use to describe something quickly. Instead of describing what a pomelo is in scientific terms, you can simply say, “It’s like a grapefruit, but bigger.”
You see this format used a lot in the tech startup world. It’s almost a joke now to say that your startup “is like Uber…for X.” But in general, the fastest way to quickly describe how your user should think about your product is to give them something to compare it to.
Guiding user behavior
Whenever you create a product, the overarching questions are: What does this product do and how do you want people to use it?
There are literally thousands of ways to answer these simple questions. Choosing a consistent approach is the hard part. Once you choose, you have to then design it so that people will actually use it the way you want.
I had two primary concerns:
(a) Easy enough: I wanted the product to be easy enough that people would want to do it, and not be intimidated or overwhelmed.
(b) Hard enough: It had to take enough effort that it felt fulfilling for the person to fill out the questions. Also, I wanted the recipient who got the email to feel like the sender took time to write thoughtful answers, rather than trivial and thoughtless ones that took two seconds to quickly do.
This is a hard balance to strike. Get it wrong, and you lose users.
My first iteration was 365 days of love notes. However, there was another concern that trumped this: I wanted people to do it in one sitting. You can’t type out 365 reasons in one sitting.
If you asked people to do that, they would abandon the site and say “forget it.”
How long should this take? 60 min vs 15 min
I had to decide: what was an appropriate amount of time that I could envision a user spending on this in one sitting? 60 minutes? No, that was too long, especially for short attention spans on the internet or mobile, when checking your email is one click away.
I decided on 15 minutes.
Therefore, I trimmed the number of days for people to fill out answers down to 2 weeks. Then I trimmed it again to 1 week.
Why 1 week versus 2? You could probably do 2 weeks - or 14 answers - in 15 minutes. But they would be fast. The person might get stuck on question 8 and abandon. If it were 7, the user could spend longer on each question and give a more thoughtful answer.
In the end, I wanted it to be a positive experience for the sender and the recipient, and having a thoughtful answer to questions was a main priority.
People like us are nerdy
Nerdiness is something I intentionally emphasized with the positioning of the product.
Being a nerd is cool and aspirational, and people are comfortable calling themselves nerds - it’s even a badge of honor in many places like Silicon Valley and SF.
“Have you ever subscribed to an email list…where you’re the only person on it?” and “an email list for one” are explicit signals that this project is for tech-minded individuals.
Not everyone knows what a drip email is. But the right people will.
How to be snarky in the right way
What does it mean to be snarky? It’s easier to illustrate than to describe. Let’s take a look:
“Because no one loved us enough to create one. UNTIL NOW.”
Strategic use of all caps. It’s like shouting. Normally you’d avoid using all caps, but here, it plays into the tongue-in-cheek tone of the site.
“Feelings of love will burst into sparkles”
Totally over the top, but totally in line with what the target market would at least chuckle at.
“No, this is even better. People recycle mason jars. You can keep these emails FOREVER.”
I bring up a cherished concept of a jar of love notes, and then poop on it, claiming that my product is even better. This is borderline offensive, but the confidence is appealing.
“Do I have to write something for all 7 days? Yes. Don’t be lazy — you can think of 7 things. We’ll give you question prompts too.”
Again, telling someone not to be lazy can be risky. But it was in line with the target audience and I knew that they would “get it.”
Choosing the right form provider: Wufoo vs Typeform
Choosing the right form was an important decision. At the end of the day, this form was what people were interfacing with when doing Good Morning Love.
I originally looked at Typeform, a hip new startup that makes slick forms. Initially, I loved the user experience of their form - it was clean, beautifully designed, great typefaces. It looked professional and legit.
However, Typeform made the user go from field to field to enter information. Meaning, you’d type in your name, then it would pop you to the next field to type in your email, then pop you into the next field…
A user couldn’t see the whole “bird’s eye view” of what the form looked like. This was problematic because the form was, in fact, kind of long with 7 entries.
My concern was that a user would start filling out the form, and not be able to see how much he/she had left to do. If you’re filling out the answer for Day 3 and you can’t easily see how many you have left to do, the easiest thing to do is to say “Meh, never mind” and abandon this altogether.
I didn’t want that to happen. I wanted people to see that the entire form was not that long, and that they could do it in one sitting.
I chose to use a Wufoo form because it allowed people to see the entire form in one view, and therefore made people more likely to complete the questions.
Even the size of the form text response box matters
It would be anti-climatic if the recipient got a note from her partner, only to have it be one word. Yes, some people can fill it out this way, but I wanted to encourage users to type at least a few sentences.
Originally, the default was a single line text box for responses to questions. This would signal to users that they should type one word or one phrase. I changed the box to be a “paragraph text box” because I wanted to signal to users that they should type more.
Then I straight-up included this in the directions that they should try to write a few sentences.
“You COULD write one word answers. But it’ll be more fun if you write a few sentences. You could really go above and beyond if you write a short paragraph."
Again, the goal was to guide people into writing something that the recipient would be delighted to receive, while streamlining the sender’s experience.
Removing friction from the question prompts
The question prompts were perhaps the most important part of Good Morning Love. There were several iterations of this.
(1) Have people list 7 reasons why they loved their partner. This was very freestyle and people could write anything they wanted, at whichever length they wanted, on whichever topic. This was very flexible - too flexible.
(2) Fill-in-the-blank questions with very short and simple answers:
- Favorite body part
- Favorite personality trait
- Favorite thing you do for me
- Favorite thing we do together
This was good because it was fast and frictionless, but also felt commoditized. The sender could fill this out quickly, but the person receiving this wouldn’t get a great experience or feel like the notes were consistently meaningful. So I scratched this idea.
(3) Then I considered asking more nuanced questions. But the problem with these questions were that they put the person on the spot. What do you appreciate most about your partner? It’s a pretty heavy question, and one that you want to get right.
There’s pressure. I wanted to remove the friction and any unease someone might feel while filling it out. The answers should come really naturally and easily.
(4) In the final iteration, the idea became "Complete the sentence” type questions. More details below.
Phrasing questions to optimize responses
In case you haven’t seen them already, here are all seven questions from Good Morning Love:
- Day 1: I appreciate when you…
- Day 2: I don’t get to tell you this enough, so I just want to take the time to say…
- Day 3: I knew I loved you when…
- Day 4: You’re obviously awesome inside and out. Focusing on the outside for a second, you look so good when….
- Day 5: Before I met you, I was a quivering pathetic bowl of jello. But NOW. Now after you, I’m a better person because….
- Day 6: I’m really drawn to your values. I love that you….
- Day 7: You probably don’t realize I love this about you….
This list is a critical piece of the product and it worked for several reasons.
First, I came up with questions that allowed people to answer in one phrase, but also have room to elaborate. This was important and gave people the flexibility to write about what they wanted to, without feeling like they had to write an essay.
Second, I chose a mix of questions. There’s one about physical looks, one about the person on the inside (those two balance each other out), one about appreciation for what the person does, one that encourages a quick anecdote and trip down memory lane, one of how the person transformed you and impacted your life, one about the person’s values and character.
The diversity of the questions allows the sender to compliment their partner from all angles, rather than just focusing on one thing (like looks, values, things the person does, etc).
This is a feel-good thing to read because it shows that your partner notices all these different things about you, that maybe you didn’t even realize that they notice. That’s huge.
Third, choosing to put ellipses after the prompt was itself a strategic decision. I could have used a colon, but that would have made it feel harsh and direct. It would have put the person on the spot.
“I appreciate when you: ____” demands an answer and almost feels like answering a question on a test. The approachable tone of “I appreciate when you…." invites users to answer and feels more like a conversation.
Lastly, I wanted to surface things that people might not say to their partners on a daily basis. For example, day 7’s question was something that you love about your partner that they probably don’t realize. How special would it be to receive a note from your partner with the answer to that question? It would be a surprise and a delight.
It’s also one of the toughest questions on the list, so leaving it for last meant by the time you encountered it, you were already pretty committed to finishing the questions. If I had put it earlier, it might have caused people to quit.
Set expectations to avoid problems down the road
Since this was a side project, it meant that I didn’t have a tech support team, developers, or a community manager to handle incoming inquiries if there were any.
Therefore, my goal was to:
(a) minimize incoming inquiries and prevent people from having questions
(b) proactively anticipate questions and answer them, for instance, suggesting that the recipient might want to check his/her spam folder.
(c) set expectations from the beginning so that users would not be disappointed when they encounter an issue and realize that there’s no tech support
Repeat the same information if it’s important
Along the lines of setting expectations, I created a detailed FAQ section.
It was actually just three pieces of information repeated 15 times in different ways. I knew the questions users would have (How does it work? When does it start? What time is the email sent?), and kept answering them in different formats so that people could recognize their question instead of having to do any extra thinking.
This worked surprisingly well. I’ve barely gotten any emails with questions about how Good Morning Love works, because the FAQ section and the rest of the website was optimized for clarity.
User flow and double opt-in
Once the sender submits their 7 days worth of love notes, I had to decide whether to email the recipient automatically, OR have the recipient opt-in to the email.
I decided to have the recipient opt-in. The recipient would get a note saying that their partner had signed up for Good Morning Love, describing what Good Morning Love was, and stating "if you’d like to receive these notes, click the ‘add me to this list’ button.”
The risk here was that not everyone would subscribe. Most people are used to seeing “add me to this list” buttons for email lists that include thousands of people for promotions. But this was different in that it was an email list where all the content was created specifically for one person. My partner helped set up this MailChimp hack.
I wanted people to subscribe so that they would know what Good Morning Love was - AND so that they had a chance to sign up too. This would double the number of people doing it, if the recipient also had the option to write love notes for their partner.
Another idea that I briefly considered was sending the initial opt-in email and saying that a secret admirer had signed up for Good Morning Love. This seemed like a good idea at first because there would be an element of surprise, and it wouldn’t raise expectations too much about what these emails would look like when the recipient finally got them a week later.
However, I thought the secret admirer approach was risky because people might think the email was spam or an advertisement. If the recipient found the “secret admirer” thing sketchy and didn’t opt-in, he/she would never receive the notes and that would be problematic.
I prioritized having the recipient opt-in to the email list that their partner created so that they could receive the notes. This meant actually including the sender’s name and email address in the opt-in message to make it feel more legitimate and thus safe to opt-in to.
Staying mum about ourselves
(you, your partner, W, J)
I decided to stay mum about who the project creators were. My partner helped with some technical implementation of MailChimp, and I did everything else.
I took into consideration that the topic of love and Valentine’s Day is, of course, incredibly personal. Good Morning Love is literally a middleman between two lovers. I decided to be semi-anonymous and sign each email as “W & J” to not put the focus on the creators.
I wanted the focus to be on the couple (let’s call them Natalie and Jeff). When Jeff wants to tell Natalie something loving and personal, they shouldn’t feel like there are two other people in the room.
If it’s Natalie, Jeff, Winnie, and Jason, that just seems like a lot of people.
If done right, Good Morning Love would not take center stage. The words of each partner would take center stage. I wanted Good Morning Love to be a trustworthy yet discreet presence, and signing as “W & J” allowed me to achieve this.
Show people what it would look like
In general, people like having some certainty as to what something will look like when they engage with a product. If you buy something on Amazon, you’re used to the “add to shopping cart” process. It’s the same on JCrew or Nordstrom. You know that you’ll fill out your billing address, shipping address, confirm, etc.
For something like Good Morning Love, it’s not as clear what the email that you’re sending will look like. And if you’re attaching your name to it, and spending time to fill out this form, you want to at least have an idea of how this content will be presented to your significant other.
Recognizing this, I took a screenshot of a sample email from our hypothetical couple Natalie and Jeff, so that users could feel “safe” knowing what their partner would get as an email.
I’m often asked, “If I’m supposed to build things that might not work, why should I think through all these components? Can’t I just run with it? Even after considering these issues, how do you know if an idea will work?"
Here’s what I say: you won’t really know if something will work. But you have to get the core part right, or at least have a clearly defined assertion of what you think will work. From there, you can expect to tweak and adjust once your product is live, based on how it interacts with the market and with users.
Good Morning Love was a blast to work on. I hope this post helped you think through some of the challenges and solutions that I faced in shipping it.