6 Learnings & Stats from Kifi’s Launch on Product Hunt


When you launch a product, you have to decide how to tell the world about it. With every investment of time and energy to publicize your launch, there is an unknown ROI. It’s even harder to guess what the ROI might be at a startup, where there isn’t a long track record to look at. The last thing any PM wants is to spend a bunch of effort and not see any returns.

I hope sharing the lessons I learned from our latest launch campaign helps other PMs. Our latest launch, for Kifi for Teams, was a success. We brought in hundreds of new users. In true PM fashion, I had several hypotheses I was testing.

We were launching Kifi for Teams the week before Thanksgiving. It was one of a few dozen medium+ sized product launches I’ve been responsible for, after stints at LinkedIn, Myspace, and Zynga. Kifi has limited resources, like just about every Silicon Valley startup I know. So we were thinking it over: how to get the most for the least amount of team resources spent?

We had reached out to a few people in the media by that time. But, let me put it this way: I’m a product manager. I like things I can control. You can’t control whether the media writes about you or not, let alone what kind of story you’re going to get. I decided to focus more efforts on things I could control.

We had always planned to put Kifi on Product Hunt. But about two weeks before launch, we decided to focus our efforts on Product Hunt. We had always been nervous about whether or not our product was ready for the Product Hunt community, but I remembered Reid Hoffman saying in to a group of LinkedIn PMs:

He is quite a wise man. It was time to rip off the Band Aid and get first-hand feedback.


We chose to launch with Product Hunt and put the bulk of our resources into it because the community is a great fit. Many people who love Kifi are developers and techies in the broadest sense. They play with software in their spare time, and Product Hunt is one of their daily destinations .

We also wanted to get feedback from users on Product Hunt. Every piece of feedback we get is golden, because it helps us understand more and get better. Engineers who have been working on Kifi for years could hear from real people using Kifi’s newest iteration. We also knew that existing Kifi users also like Product Hunt from our internal data so we were hoping they would be excited to see us on there as well.

After we decided to focus on Product Hunt, I read a few great articles like the ones found in Maker’s Success and Benjamin Hoffman’s post. I created a checklist of the things I could control with this launch and that would have an impact. Here’s how the items on my checklist paid off. Our goal, of course, was to get new users on the platform, creating knowledge repositories for their teams. We weren’t shy.


There are lots of great checklists and posts with advice out there. Below are a few new or noteworthy items worth sharing:

  1. Custom landing page banner featuring a $150 credit that people who came through Product Hunt could put toward a premium team. (KIfi for Teams is free but there’s a paid premium product). WORTHWHILE? ABSOLUTELY: The landing page converted 61% better. It’s hard to tell how much to attribute directly to the credit vs the banner vs the traffic source. Product Hunt has a great community of quality users — we might have gotten that many new users anyway. But my gut is that the banner helped a lot. Who can refuse an adorable kitten wearing Google glasses?
  2. Finding a Kifi user with a lot of followers on PH to ask them to post for us. When someone with a lot of followers posts, all of their followers are automatically updated. WORTHWHILE? Helpful. Thanks to Ouriel Ohayon’s help, this tactic got us higher placement earlier in the day. About 7% of our upvotes were followers of the poster. Taking a look at Ben’s findings; there are clearly diminishing returns. With 25k followers, Ashton posted Neighborly. A similar upvote rate would have generated 1,750 from Ashton alone. Given Neighborly ended with an overall total of 580 upvotes the day of their post, it’s clear they didn’t get 7% directly from Ashton, but I’m sure his influence didn’t hurt :) In the end, it supports what Ryan Hoover says, which is that a good product will do well on PH … regardless of who posts.
  3. Prepping the media assets, including the video. WORTHWHILE? NOT SO MUCH We only saw a little over 100 views on it during the week of launch. If you have a video already, post it, but don’t spend a ton of time or money to make one.
  4. Make it exclusive. Product Hunt asks you to fill out this form in order to have an exclusive launch on PH that includes an extra special offer for Product Hunters. WORTHWHILE? YES, BUT: We applied about four business days before & didn’t get the exclusive. On the day of launch, we saw that a startup in a similar space had an exclusive. We’re guessing our competitor on the platform might have something to do with the fact we didn’t get the exclusive.
  5. Get your teammates listed as “makers.” Makers have the ability to comment on the product’s hunt and you’ll want people from different disciplines within the company chiming in on the conversation. WORTHWHILE? YES I must have applied to be a maker of a previous Kifi hunt 5 times over the last 3 months but hadn’t heard anything. We ended up Tweeting at Ryan the day of launch, and he responded immediately to help us get sorted. I was busier than expected that day so it was helpful to have extra eyes that were able to stay on top of the Product Hunt comments.
  6. Find friends/users with comment capabilities. I poked around PH profiles to see to see which of my friends had submitted before. I assumed they would have capability to comment and get the conversation started. We found a handful of friends who got the conversation started by sharing their personal thoughts about Kifi. WORTHWHILE? YES. We’re unclear what this drove, but it certainly made me and the team happy, as we saw all the love.


Product Hunt Stats: We received 280 upvotes and 30 comments on day one, with lots of great feedback, in particular this comment from PietroBlu Giandonato.

Amazing! Kifi is Delicious on steroids! It works like a charm, I really love the way it’s smoothly integrated in web search and navigation, as well as its powerful sharing features.
Moreover, as a teacher I imagine some awesome ways to integrate Kifi in classroom work with my students, to improve their collaboration in groups and to share knowledge with them.
Kifi is really powerful, definitely recommend it.

Teams Created: We also brought in hundreds of users over the next few days, who created hundreds of teams. Welcome to: Wsywig, AppDirect, Red Frog Rhythm, & ZooStudio just to name a few

Retention: While you might expect that users coming in from a soft external source would bring down your average new user retention, in our case new user retention stayed constant.

Virality: We brought in some viral and influential users. We saw a substantial percentage of the users coming in from Product Hunt and a substantial percentage of people inviting their coworkers to the service, mostly on the second day. Perhaps Product Hunters are early adopters that take the time to investigate before inviting others. They wanted to play around with the service for a bit / set up their team and then decided to invite.


I joined Kifi to hep lead their product efforts a little over 2 years ago. I love working on the problem we’re focussing on because I get frustrated that nearly every knowledge management tool recognizes the importance of web content enough to handle it gracefully, but no one is focussing on making it an organized and central part of a company’s knowledge base.

Kifi connects people with knowledge. With Kifi for Teams, we are taking on the challenge of organizing web content and resources in ways that make it accessible for people at work every day. We also give web content its proper due: With our tools, you can integrate it into your knowledge management and client services. We just launched Kifi for Teams service into public beta.