It was already late when I got there
Digital mining was already in progress before I joined Phoenix Group, at the time of interviewing, and it was meant to be completed by the time I started. It is basically cloud mining for Bitcoin, sold to family offices, B2B and direct to consumer. It was a huge revamp of the old platform. When I joined I realised this project is still ongoing, and it has not been finished in terms of product readiness.
I got onboarded with an agency, and I figured out that we were behind schedule, with a lot of issues. Everything is dragging along very slowly. There are many things uncompleted, and things are not being delivered as promised, or as per the contractual agreement with the agency.
Then another challenge came into play. The upcoming IPO campaign was coming, and digital mining is one of the verticals we are going to go for in the IPO.
The agency was not going to make it
I tried to fix it with the agency. I gave it a shot, but I didn’t see any major improvement, and that just made it a much bigger risk that we won’t be able to launch on time.
What I did is I brought in a small in-house team:
- A full-stack developer
- A project manager with some development expertise
- A UI/UX designer
We also quickly shortlisted two agencies, and we ended up working with one. We built a team of 25 to 35 people working nonstop. We redesigned the front end and the admin, and also did the iPhone and app designs.
We worked nonstop, including weekends and sometimes through the night, just to meet the deadlines, because there were a lot of dependencies. I built a visualisation for the developers showing them, if one field changes, what other fields got affected. In this way they understood, and they could implement as required.
What had to be rebuilt
How the vertical ran
The calculator
I took over the calculator from Qin, who was a brilliant mind. His calculator consisted of multiple fields, and he onboarded me on it. I had calls with him to understand this way of calculation. Even though I come from crypto, blockchain and Bitcoin mining, it is a niche inside the niche, so it took me some time to understand.
It was frustrating to me that I don’t understand this complicated calculator, but I sat on it. I read it, studied it and did research, and in the end it cracked. After that I simplified it and trimmed a lot of unnecessary fields. I double-checked with the internal team managing the backend, and they confirmed I’m on the right track. After that I rebuilt the whole calculator.
It was never a handful of fields. It is one model where everything depends on everything else: the miner’s power draw and cost feed the cost per terahash per day, network hashrate and the Bitcoin price feed the revenue per terahash per day, the two meet in the spread, and the spread sets what a plan can be sold for and what the customer sees as a return. Change one and the rest move.
So the work was not simplifying it to something small. It was trimming what did not earn its place, roughly 70 or 80 fields down to about 50, and restructuring the formula so it read the same way internally and externally. The admin panel, the checkout and the slider on the plan page are all views of that one model rather than three separate calculators.
We pull data directly through the Antpool API, the hash rate, the difficulty and the price, to calculate in real time and make sure everything is accurate.
The model itself
The legend, and the views it fed
The business model
I also created a business model with visuals to showcase to leadership and get their approval.
Three panels, one plan: 150 TH/s for 30 days, bought for $346 with Bitcoin at $40,000. The only thing that changes between them is where Bitcoin ends up 30 days later.
- Bitcoin flat at $40,000, and the customer finishes with $353, a $7 gain
- Bitcoin up to $50,000, and the customer finishes with $400, a $54 gain
- Bitcoin down to $30,000, and the customer finishes with $300, a $46 loss
Phoenix earns $104 in every one of them, on a $228 cost to mine and $14 in payment fees. That is the whole argument. The plan is sold in dollars against a cost that is known when the order is placed, so the company’s margin does not move with the Bitcoin price and the customer holds that exposure instead. A single panel would have looked like a forecast. Three of them made the approval conversation about the model.
The same plan, three Bitcoin prices
What shipped
Everything was front-end facing, and for the admin we still had a lot of control. We created mobile apps and TestFlight. In the end we launched this project in about 5 weeks, and after that we were still making some changes, but overall the platform was the first cloud mining platform in the UAE. I’m quite proud of this. The domain now serves a holding page, but the platform as it ran is in the Internet Archive, captured in February 2024.
I came to the realisation that it can be quite fun, even if it’s super challenging. I quite enjoyed this. This is my first experience leading such a project end to end and being in charge of a whole vertical, especially before the IPO campaign, as this was one of our verticals to go for in the IPO.
After launch
Shortly after launching I ran some mini campaigns, promoting across crypto networks and engaging with our previous and current customers to drive the platform. Later we added USDT payments, while still having Stripe as the payment processing provider.
The platform went on to sell more than $1M in mining plans.
A buyer could also pay at the Phoenix store in Dubai, which sold mining hardware and wallets. The order was created online and settled in store, with a 7-day window before it expired.
Digital mining was one of the verticals the company listed on. The retail-investor side of that offer closed 180 times oversubscribed.
Read the retail-investor IPO campaign