The story of my startup
Simplifying e-commerce with
community driven product reviews
Spoiler alert: the story of VillageHunt isn’t over, and to be totally honest I’m not sure how it will turn out. I’m extremely proud of what we have made, and believe in this product’s vision and potential. But it’s one thing to be stoked about something you’ve crafted, and quite another to see people take it up and make it their own. We’re still figuring out the latter.
So, this is a story about the start—about seeing an opportunity and diving in after it. But most of all, it’s a story about working hard to put a product out into the world that, with a bit of luck and a lot more hard work, has the potential to become something great.
It’s never the right time to start a startup
Before starting my own startup, I led the design group at a well funded startup in Tel Aviv, worked at some of Seattle’s best design agencies, and was part of the in-house design teams at Microsoft and Expedia. I quit my corporate job at Expedia, just as I was about to get promoted, to fulfill my startup dream. And even though this dream turned out to be the hardest time in my professional career, I don’t regret any second. I feel I got promoted far beyond any job could ever grant me.
Building a startup I learned a lot, more than I could have imagined. It’s insanely hard, not only because you are creating something from nothing, doing crazy things to get noticed, working long hours, or being without a salary. The even harder parts are the extreme ups and downs day to day when in reality very little has changed, thinking how close you are but realizing how much more work lay ahead of you, wondering why you started in the first place, having no safety net, and feeling that no-one you’re close to truly understands what you’re going through. Despite these challenges, and maybe because of them, I have become a better designer. I wrote about what I learned as a designer entrepreneur here and here.
If it’s so hard, why do it?
One motivation was to truly test myself as a designer. Working at large companies, I’d always ask myself what was my part in the success or failure of whatever we were building (the answer to this goes way deeper than ‘role x’). I wanted to put myself in a situation where the quality of the UX is life or death (of the company).
Another motivation is having a voice in the world. I wanted to make something good for people, and I saw an opportunity to do that.
Lastly, building something on my own that would impact millions has been a dream of mine ever since I discovered design. But I was always scared of jumping in. So many things could go wrong, and of course so many things did go wrong with VillageHunt. But nonetheless, while we were doing it, it felt amazing, we were building our dream, and we weren’t giving up, even at the hardest times. We didn’t impact millions, not even close. But we touched the lives of a small community we care deeply about, new moms. New moms are super humans in so many ways, but even super humans can’t do everything alone.
It takes a village
A little help and shared knowledge from their community can go a long way. We were on a mission to make their lives a little easier by letting them see what their friends and community think of a product they are considering before they buy it. And we knew how to do it, we knew this community very intimately, as we were among them. ‘We’, are myself and my partner to change the world, Sam Hamaoui, a good friend I’ve known since high school.
In the picture - A group of wonderful moms I hung out with while creating VH (I'm 5th from the right), and learned so much from.
The Problem we wanted to solve
We started VillageHunt to solve a problem we had, that became even stronger as we became parents. Buying a product that is new to you is extremely hard and time-consuming. Any product search brings back tons of options and information. Now go research 😩. So you start to read reviews, which you don’t know who wrote, and why. Barry Schwartz, author of The Paradox of Choice, writes: “The Internet created a massive choice problem and then started to solve it with reviews and recommendations… But now there’s a new choice problem: choice of review. Too many reviews are as bad as no reviews.”
As a new mom myself, suddenly needing to buy so many products I knew nothing about was overwhelming and time consuming. I worked at Microsoft at the time, and there was a massive spreadsheet going around the company among new moms, with an ongoing list of products we’d need. Each mom, after delivering, would add products to the list and note what she’s learned. This exhausting list was so needed, yet so hard to consume. Still, it was better than searching for the answers online.
Shopping online is hard
Product reviews are long, anonymous, often contain irrelevant information about the seller or the shipment, and most important, they are not personalized. For a product that’s new to us, we have to read MANY of these reviews.
Surprisingly, product reviews haven't changed in more than 15 years - they look the same as they did back in the days before E Commerce grew massively and with it the amount of products and information, before we all had smartphones and before we used social media.
In our research, we asked people if they write product reviews. Almost everyone said no. The question that comes to mind is, if no one writes reviews, who are those people who write reviews? And what is their motivation?
There is no shortage of headlines like this, which hint to the answer -
There has to be a better way
Sam and I thought the best way for us to make a decision about a product we are considering is to know what worked or didn’t work for our friends and people we trust. We wanted to just ask our community and get back a list of products in order of priority 🤔
So here we go
The day after I left Expedia was a scary day. We were on our own, and we had to make this work. It was all up to us, yet I had no experience building or running a business. So instead of thinking about a ‘business’, it was easier to think about a first small step.
We started with research, something I knew how to do very well. We spent a couple of weeks just talking to people in our target audience, validating this was a real problem, learning if and how they solved it today. We had no money to pay the people we talked to, but we found creative ways to engage the community, and it worked.
In fact, to my surprise, so many people replied positively to our post, voluntarily offering to talk to us and share their experiences, we couldn’t even fit all of them in the time we had. New moms are sharers, and love to help each other. Just another reason to love this awesome community 🙂.
Here’s a video testing one of our earliest prototypes -
Jump forward a few months and we launched our first Minimum Viable Product. We invited moms from Seattle to join. The idea was very simple - rate products you’re familiar with to help others, see top products among moms in Seattle.
Moms loved this! The engagement rate was very high (67%) and the feedback we got was encouraging.
We launched a few more MVPs after this one. We tried different product categories, to test if and how we can expand our product in the future. We launched the same idea for sewing products, and invited a community of seamstresses from the UK. Neither of us has any experience sewing, and we thought it would be interesting to test the method on a category we know nothing about. Again, engagement rates were very high. Apparently the type of needle or scissors you use can make all the difference!
Conceive. Make. Improve.
Following the success of our MVP we started to prototype the ‘product’, and initially got caught in the trap of making it too complex.
Of course in the process of building, we thought every feature was important to make the experience better. But with each iteration, we ended up reducing features for the sake of simplicity. At this point our CTO decided to leave the company (which felt like a breakup to me, I was heartbroken), and we decided to outsource development. Still being bootstrapped, this forced us to (a) prototype and test A LOT before putting anything into code and (b) remove any feature that wasn’t 100% necessary.
The problem we were trying to solve wasn’t new. Shopping is hard, and many tried to offer solutions before us, startups as well as large companies. At times we found that features we thought about were very similar, almost identical, to what has been tried before (for example, asking users to categorize their product collections into buckets). And we also stumbled upon startups that came after us, making the same mistakes we made in the beginning. We didn’t want this thing to get full of user choices (read: complexity). It can be hard to see stuff we designed, and even built, not go live, but these features needed to die so the product could grow through simplification. Any functionality that did not add a great amount of value, we eventually got rid of.
Focusing on one goal:
Get users to contribute content
Our site had the classic chicken and egg problem. How do you get users without content? And how do you get content without users? We needed to fill the site with valuable, trustworthy content, before focusing on anything else.
In a market where people get paid to write product reviews, gathering this content wasn’t going to be an easy task. We wanted content what was authentic and trustworthy. Reading a review on VillageHunt would feel totally different than reading a review on Amazon.
We didn’t optimize for retention, shares, or boosting the product with enhanced features like user generated product pictures. All of our efforts - the way we approached marketing, the way we designed the onboarding, and the product - were focused on getting users to add content.
Engaging with moms through marketing
gave us amazing product insights
We didn’t have marketing resources, and no experience with content marketing or blogging. We experimented with different types of content, placing links in different places, most of them without spending any money. We analyzed what is working for other moms and parenting blogs. We followed moms who got several hundred likes for their posts and asked ourselves ‘why?’ I reached out to a few influencer moms asking if they’d share their story with me. Luckily some agreed and told me why and how they got so many followers, where they post, who in the community they ask for permission, how they phrase the ask, and more. Their advice was gold.
The thing that ended up working best for us was content marketing in Facebook Groups. The stories and posts we wrote came from content people shared on our site. It was unique content that couldn’t be found anywhere else. For example, surprising (negative) opinions about an Amazon best selling product.
We also figured out how to convert people to become VillageHunt users without making it feel like we are selling something. This was very challenging and took many tries to get right. Facebook group admins are super sensitive about the content they allow in their groups. In mom groups the admins are the queens of the group, they worked really hard to get so many people to join and keep the group active, and upsetting them isn’t a good idea 😉. In our experiments I was banned out of a few groups forever. But after many failures we got the wording right, some admins gave us their blessing, and we were able to onboard hundreds of users within a couple of hours from just one post. A good post mentioned the admin and thanked her, included something personal + a picture, and encouraged users to comment, which kept the post on top of the feed for about 24 hours (depending how active the group is).
If you clicked the link from the Facebook post, onboarding has begun.
Many things prevented the onboarding from continuing smoothly. Since the user pressed on a link coming from Facebook, the site opened with the Facebook chrome around it. Pressing the back button threw you out of the site completely. And, in order to sign up you had to remember your Facebook password, which hardly anyone does. This was a critical step to design, that could make a difference between someone trying again or closing the tab, walking away, and never coming back.
Messaging was tricky. I can’t even count the number of times we went back and forth between ‘moms’ and ‘parents’. When we used ‘Moms’, many moms were offended saying they’re not the only ones in the house making purchasing decisions; however the word ‘Parents’ - felt less unique. We wanted people to feel like they are entering a special community, and ‘Parents’, although more inclusive, did not give that feeling.
Using tutorial slides to explain what the app is about before signing up didn’t work well for us, they gave a feeling that there’s too much more to consume before actually using the product. So we removed any barrier to sign up and made it just one click away. Instead of the slides, the entry page had a short paragraph explaining who we are and what’s unique about our community. If you still weren’t sure and wanted to learn more, there was a link with screenshots of the product (a very small percentage of users clicked the link, most just wanted to go straight in). The rest of the onboarding happened after you were signed in, and was shown in context.
We ended onboarding with this text: ‘A great way to start is by helping others - just click on a product you’re familiar with!’. This closing was key to achieve our goal of content first, and it was indeed the first thing users did, they added content to the site before exploring it further. 54% of our first time users rated products, with an average session duration of 8 minutes. Another advantage we got by doing this, is that contributing content left users feeling invested in their first minute of being on the site.
After dismissing this last sentence, users arrived at a page showing a grid of products. This was a very risky decision, having a landing page that doesn’t add value, but only extracts value from users. We were very hesitant to do this. Why would anyone add their own personal experiences to a site they had just entered, without even exploring it first? We made it work by choosing wisely which products appeared in the grid, and in what order. We knew, through research, which products triggered an emotional response, such that if you have a baby in a certain age, it’s hard to resist commenting how this product helped or didn’t help you.
A small change in the UI can completely change the way people behave
In an app where people come in order to see what their community thinks about a product they’re considering, being able to add a product is important. It serves two needs: either to ask others what they think of it, or to criticize/praise the product for others to know.
We experimented with many ways a user can add a product, each time seeing big shifts not only in behavior, but also in how users talked about VillageHunt and explained it to others.
Adding a Product
In the beginning we put the ‘Add product’ button at the end of search results. The scenario made sense - you search for a product, see the results, and at the end of the results, if you still haven’t found the product you were looking for, you can add it.
Not many people were adding products with this design, but that wasn’t our main problem. Look at the bottom tabs -
The three tabs on the right are pretty self explanatory, search, notifications and profile. The first one on the left is the VillageHunt logo, the home tab, it’s where users can see what products people in the Village are researching, and what they recommend. But how did the products get there? With this design some users were left confused.
We decided the ‘Add’ functionality shouldn’t just be a button at the end of search results, but instead have a whole tab dedicated to it. Of course in reality most of the products were added by us, but we believed this would make the app as a whole more understandable, answering how products got added in the first place. We thought a lot about what content this tab should include, and decided to turn it into an ongoing feed of products being added, so that it’s interesting for users to visit even if they don’t have a product to add.
With this design the percentage rate of added products went up significantly. But still, we had a lot of problems.
⓵ The actual functionality of adding a product was only a small area on the top of the page, easily being overlooked in the overall layout.
⓶ The name of the product was too large, even though it’s not a very interesting piece of information, and most importantly,
⓷ The ongoing feed in the ‘Add’ tab didn’t give enough value, it didn’t show information about why these products were added (the only information given is the little icon to the left of a user’s name).
Improving this as part of a redesign of the entire site
Now that we had more content, we could use it to deliver more value, faster. With this change in focus, the importance of ‘Add Product’ reduced significantly.
The main use case of returning users to the app is to look up a product they are considering and see what their friends and community think about it. But having the Search as a tab at the bottom, even if it’s the middle tab, does not encourage that.
When search is part of the navigation bar, it’s at the same level with ‘Home’, ‘Profile’ or any of the other tabs. It isn’t as discoverable as a search bar on the landing screen, even if it’s easily accessible, one click away. A tab also implies ‘Search’ gets an entire screen for itself. This screen has a search bar on the top and the rest of the screen is populated with data that would either aid the user’s search or would help the user explore content on the platform. This facilitates an exploratory search for a user who does not have a clear intention yet. But our users do have a clear intention.
So we changed the hierarchy, and instead of having tabs on the bottom with equal importance, we moved important functionality to the top, showing the user only what they need to see at each given moment.
Putting the search bar on top of the landing page makes it easily discoverable as a main scenario, easily accessible, and a core action. With this change ‘Add’ is one of a trio of buttons meant to assist the functionality of the app as a whole, but not appear as a main feature, a tab that’s equal to the home page.
More users leads to better content shared
The two main tabs on the homepage are now ‘Recently Researched’ and ‘Recently Recommended’, they are the essence of the app, and tell the story well. The latter tab contains the most interesting content, and eventually they need to get switched.
When we just started, the content of this tab showed most recommended products that relied on very little data. In the good case where a user had a friend in the app, and if the friend recommended a product, she’d see that. But it wasn’t very interesting. With more content, we could show community top rated products by age, recommendations by category, useful information about products the user is more likely to be interested in, and reviews the community marked as helpful.
The story isn’t over
We now know vastly more about personalized shopping than when VillageHunt was just an idea in our minds, but we still have so much to learn. The journey is exciting. As I said in the beginning of this article, one reason I quit everything to start a company of my own was to test myself as a designer. Our product does not have any exceptional development or unique patents. If we win, it will be because of good User Experience - from marketing to execution to product.
In the future VillageHunt will learn the personal priorities users have, and will let them discover products based on their needs, their preferences, friends recommendations and what’s popular among their community.
There is still a long way to go. We feel good about what we've done, and even more excited about what we want to do.