What a trip to Party City taught me about product management
The Party City on West 14th Street in Manhattan has an interesting ambiance. Sure, it says ‘party’ over the front door, but spend enough time wandering the color-coordinated aisles and you’ll soon realize that most of its patrons are quietly freaking out over preparations for their impending events. It really puts a damper on the whole ‘party’ vibe.
I’m a Product Manager at Managed by Q. Now, this story isn’t really about Party City, but thanks to a feature release on a Wednesday morning back in March, I was one of those panicking patrons. And this time, Party City saved my ass.
Instant Booking
Managed by Q is a platform that enables offices to build, design, staff and manage their workspaces. We partner with local business to offer a wide range of office services to our clients through the Q Marketplace.
Earlier this year, my team worked on a new experience that would make it easier than ever for our clients to get light-maintenance services. Light-maintenance services include things like unclogging a toilet, mounting a television, or getting someone to come and unlock the snack closet. The things that keep an office running smoothly day-in and day-out.
These services are interesting because they are frequent, often urgent in nature, and for the most part, are things that office operations teams just don’t want to deal with. Spending time on these services was making it difficult for our clients to focus on the things they really care about. Culture building activities and projects to improve employee productivity always had to wait whenever a conference room light bulb needed replacing.
To solve this problem, my team developed an Instant Booking feature. Instant Booking would allow our clients to submit a few details about their service needs, see an upfront price estimate for that service, and automatically receive a confirmed service provider with an arrival window — no further actions or steps required. Finding the right highly-vetted service provider, a process that could take several hours on the Q Platform (and typically even longer outside of it), would now take just a few minutes. This would free up our clients time, enabling them to move on to other tasks.
In little over a week, we had designed a new booking experience, onboarded a few service partners to the new Instant Booking workflow, and shipped an MVP. We were moving fast.
The Party City Run
One of our first orders came in at 11 am. It was for a ‘Helper,’ which is a service used for tasks like unpacking groceries, furniture assembly, and event setup.
We immediately saw a problem. Our MVP used text fields to enable the client to describe their issue. In this case, the client wrote the following:
“I need someone to go to party city and get me some pink supplies and decorations for a baby shower today ideally i would need someone to go to party city and bring me these supplies ASAP — the baby shower is at 4.”
The order also included a very specific list of assorted pink items. This wasn’t the intended use for our Helper service, but the clock was ticking. Our MVP allowed the submission and now there was a client relying on Q to make it all happen. It was noon and I had a meeting in an hour. This seemed like just enough time to grab the requested supplies, bring them back to our office, and hand them off for delivery to the client. Maybe I could even do it all undetected. At the meeting, I’d happily report that all this Instant Booking was going great.
I left our office and made my way through a mix of hail, rain, and mud that only New York City in the month of March can provide. Soon enough, I was jogging through the store like I was on an episode of Super Toy Run. Luckily, Party City has tailored their shopping experience to people like me and it was easy enough to find the “pink stuff” aisle.
The shopping spree was going smoothly. My one misstep was leaving the 24-inch balloons for last. Balloon orders take far longer than one might think and the proper expectations are seldom set beforehand. As I stood waiting for the balloons, I got the sense that there were some lessons to be learned from this whole trip to Party City.
I resolved to reflect on this later — I had about 15 minutes to get back to our office for a team-wide meeting. I hailed a cab with the wintery mix still coming down, time running out, and a laughable amount of pink accessories. It wasn’t long before I was heading up the elevator of our building, back to Q.
My plan was to sneak in, hand off the baby shower items to the client’s account manager, and slip into the meeting. As the elevator doors opened, I met the perplexed gaze of several colleagues. It turned out my meeting was being held in the all-glass room with a clear line of sight to the elevator. I thought to myself 24-inch balloons are really big, much bigger than 24 inches, i think.
Our expert Account Management team took things from there. Sure enough, the baby shower went off without a hitch later that afternoon.
I realized Instant Booking was a complex problem. It was a new experience on our platform and because it was so new, we didn’t yet have all the information needed to make it great. We would have to embrace failure and try new things in order to learn quickly. There was a long road ahead, but thanks to the Party City trip, I took comfort in having some great feedback for our next iteration.
Instant Booking Today
Thanks in part to the lessons learned on that trip to Party City, Instant Booking became a valuable capability for Q’s clients and partners. We’ve come a long way since the initial release:
1) It’s used a lot. Our 10 most frequently ordered services can now be Instant Booked. The volume of these orders has increased ~3x from Q1 to Q2 2018.
2) It’s reliable. To date, over 80% of Instant Booking orders are fulfilled, well above industry standards.
3) It’s saving time. Clients that use Instant Booking receive a confirmed service provider faster and with less effort, giving them time back in their day to focus on other activities.
Lessons Learned
My experience at Party City also taught me a few valuable lessons about product management.
Don’t wait for perfect information
Colin Powell has a rule for decision-making. He says that you should have no less than 40% and no more than 70% percent of the information needed to make any tough decision. If you make a decision with less than 40%, you are shooting from the hip and bound to make too many mistakes. If you have more than 70%, you’re moving too slow.
With Instant Booking, our team was operating in a brand new problem space. We knew we didn’t have all the answers, but early signal from our conversations with clients helped us make some assumptions and get past that 40% mark. From there, we knew that the faster we got started, the faster we would learn, and the faster we would improve.
Plan for feedback and respond fast
Tolerating failure and keeping an experimental mindset means budgeting time for unplanned work before you even start on a project. Anticipating the unexpected helps set a team up for success in the long run.
During Instant Booking, we held a weekly meeting with product, engineering, design, and all the teams on the front lines to create a faster feedback loop. We talked through any new entries in a running issue log, which made sharing problems, ideas, and updates between a cross-functional group easier and more actionable.
Own your promise to the end user
“Being a founder” is one of our core principles at Managed by Q. We believe that being a founder means taking responsibility for successes and failures, regardless of your functional role.
In the Party City run described above, the Instant Booking MVP gave the client an automatic confirmation that their service could be completed and they expected great service in return. To them, It didn’t matter how long we had been working on the feature or if we thought they were using it “correctly.” They just wanted help.
We could have ignored the order and chalked it up to some early issues with the MVP. Instead, we did what we could to make the client happy, logged the feedback, and prioritized a fix in a later iteration.
