Company On Fire

The Client

Backed by popular venture firms, a Silicon Valley virtual currency startup had set their sights on taking over the online app space across web and mobile. Seeing great opportunity in a market with explosive growth, they knew there was much more potential and security in becoming a utility provider in the market instead of just another app. They saw there was very little communication and exchange between apps which could drive unique value for users and drive engagements for apps. They wanted to become the online virtual currency of the entire app ecosystem. They planned to create a platform where users in one app could earn currency and use it in another app as well as marketplace to exchange currency for things like gifts and cash.

The Problem

Eager to get traction in the market to secure more investment, they launched as soon as they had an MVP. The MVP was built by their internal team and two external engineering firms working together.

Technical Issues

The nature of the platform was very data-intensive. For every user, there were thousands of actions needed to be logged and processed daily. Allowing internal and external logging of user events while continually crunching that data to calculate variable rewards was a demanding challenge. Having a marketplace with auctions that needed to use this generated data in real time amplified the difficulty.

Speed

The site was incredibly slow. They only had 2,000 users in total, but it took 30+ seconds to load a page even with only 1 or 2 concurrent users. In addition, performance degraded exponentially with more concurrent users up to around 50 where the servers crashed.

Corruption

The currency was not being awarded correctly. Sometimes it worked and sometimes it didn’t with no reason why. Gaps would appear in the data with no way to retrieve what was missing.

Scaling

To meet investor expectations, they needed to scale to hundreds of thousands of users. But with such poor performance with so few users, this looked impossible.

Design

Their app design was very basic. It did not convey to users what they were about and what they should be doing so many ended up lost and frustrated.

Investor Confidence

Seeing the issues in production and the issues with their brand, investors were not confident in the executive team’s ability going forward. Their request for another round of funding to churn through their issues was rejected until they could prove they could take care of the current situation.

Company on Fire

With a very poor user experience, users were quick to dismiss the company as a scam. Even though they weren’t, they couldn’t shake this first impression. They received an online backlash and serious negative PR as a result. Too much of their executive team’s time was spent strategizing and fighting back against this misinformation instead of working on their core business. Their company was on fire, and it wasn’t clear if it would make it.

The Solution

We were called in to handle their desperate situation and come up with a path toward success. We had to be smart about our strategy because of the limited funding they had left. Spending all their remaining funds on server costs to house and process their enormous amounts of data on the off chance that speed was their biggest issue was not an option. Few firms could have rescued them from their current situation with their limitations.

Whenever a company positions itself as a utility provider, performance and reliability become the most important things they must ensure in their platform. Just as electric company customers won’t accept the lights going off for no reason, so to will their users. With the brand having taken such a massive hit, the only way to repair it was to rebrand and relaunch with a new platform that worked.

Speed

After analyzing their work, we realized there were many major bottlenecks in their approach. They were many user-facing synchronous processes that could be broken up and made asynchronous with a proper level of coordination. We sped up their site by over 100x from 30+ second load times to less than 3/10ths of a second. Currency processing times that locked the user-facing site were decreased by over 1,000x. Using advanced fragment caching and processing on the back end and front end, the site was able to handle thousands of concurrent users effortlessly with no degradation in load times using the same servers as before.

Design

The site was completely redesigned from the ground up to be more engaging. Being a new company trying to disrupt an industry, the focus was put on showing information to guide users in using their platform correctly. With this, they were finally able to unlock organic growth to make their app viral.

New Concept – Pending Points

Users needed instant feedback on their earned currency no matter how big their user base or concurrent counts got. They needed something 100% future-proof. With some clever backend systems, we introduced Pending Points to solve this. Pending points are currency that has been earned but hadn’t been processed. This way users could see that their activity had been recognized even though it hadn’t cleared in their accounts. This small changed shifted the burden of the technology from constant, instant processing to transparency which what was users really wanted. User engagement soared as users were constantly checking through the day to see how many points they’d earned and when they’d clear. This concept was infinitely scalable and also completely eliminated issues with data corruption.

Scale

With site performance and design in order, they were able to scale their users organically to hundreds of thousands of users very quickly without growing pains.

Brand

They relaunched under a new, yet similar brand. They avoided the PR nightmare and online backlash of the first launch.

Investment

With such a massive turnaround in their businesses, the investors confidence had returned. They secured their next 7-figure round of funding with ease.

The Results

100x

Site Loadtime Speedup

1,000x

Processing Speedup

20x

Smaller Codebase

1,000x

Smaller Database

200,000

Growth From 2,000 Users

1

Damaged Brand Saved

The Lesson

Discovery is Paramount

While getting to market is great, having a working product is much more important. When you get behind on schedule, it may feel like you have to rush to get something out too often the damage is too much. With proper planning, you can avoid all of these issues. This is a phase we call discovery. It is not about just planning what you need to do, but how you need to do it and what you should be prioritizing along the way.

Had the startup partnered with us before working with the other firms, we could have advised them to focus on their reliability to avoid issues with being considered a scam since they were attempting a utility play in the market. Having a technical partner shows you areas you need to focus on that you might not realize so you can avoid major fires.