Published by admin
Posted on July 16, 2017
Wow. No plan survives engagement with the market! We had no idea this was coming. I think anyone who has run an ICO could write a book on how not to run an ICO. This is mine …
Twenty years ago, I had an idea for the open web, where all the information was live and you could find anything, rather than having to go to dozens of web sites and search each of those silos. I tell the story of this journey up until July, 2017, in The Story of the Pillar Project. In this essay, which is part of The Token Handbook, I’m going to tell the story of how the Pillar token sale went down. Not the way we had in mind.
The last eight weeks have been a roller-coaster ride. We have no employees. Everyone is a volunteer. We have been getting more than one new person every day joining the team. Everyone is working to build this dream for free. Here’s the dream:
That video was made by an awesome team in Tel Aviv. They had never touched cryptocurrencies or heard of tokens, but they agreed to be paid in a mix of cash and Pillar tokens. The cash part was paid for by an early investor who wanted to be repaid in pillars. So along with having a few dozen amazing volunteers who are aligned with the mission of the project, we have financed the project entirely with the future tokens that don’t yet exist. And thank you to the more than 60 people who have been involved so far.
The plan was simple: reach out to our friends and some family offices and hedge funds to take about $1m in our presale, then focus on large investors to take down the main sale. I’ve written about this, saying that large token investors play the role of underwriters and help a project get going when there’s higher risk and much to build. We hoped that the Pillar token sale would be one of those that gets crushed by huge orders in a matter of minutes, and that sometime next year our natural buyers and supporters would try our early product and the enthusiasm would build from there.
Most token sales are open for 4 weeks. Ours is open for 60 hours, ending July 18. Why? Because we expected to run a “wholesale” campaign, where big buyers take most of our offering in minutes, like other recent big ICOs. Instead, it’s been the opposite — eager new and small “retail” investors piling into our project because they believe so much in our mission and want to support us. In retrospect, it would have been much better for us to run an 8-week campaign, because something happened that changed everything.
No sooner had we released the web site than Joe (JSnip4) released a video saying how badly he wanted to get our tokens:
It turns out that Joe has 45,000 followers on YouTube. Within minutes, our web site had hundreds of visitors, and emails started pouring in, saying “I’m a JSnip4 follower, please get me in on your presale.” It was pages of emails like this. Within 24 hours, we had 7,000 emails.
So we immediately started making mistakes. We started replying by hand to each email, then in groups, then we maxed out our email limit with Gmail, then we built a script to capture the addresses and suck them into MailChimp and started our newsletter. Here’s what happens: people subscribe to your newsletter, confirm, and then Gmail immediately puts the incoming first newsletter conveniently into your Promo folder. So we started getting hundreds of emails about where’s the newsletter? People asked if they could wire us money. They wanted to get in on the presale before the presale. We were just trying to get the newsletter to work under heavy fire, and we sent out a few newsletters that had small mistakes that caused big problems later.
They were so eager to get our tokens that they went to our web site, saw our smart contract testnet address listed (this is common practice), and just started sending ether to our testnet address, even from exchanges, incinerating at least $60,000 worth of ether in more than 200 attempts to get our tokens. Some people just sent ether from CoinBase, somehow hoping to get tokens back. Some people blamed us for putting a testnet address on our web site (labeled TEST NET), and they were very upset. We told them we’d make them whole if we hit our maximum during the main sale, and they said that was fair.
People somehow got the impression that Joe had worked out a “deal” to guarantee all his followers that they would get presale tokens. A few quotes from those few days last week:
Okay, so this was a different world. I stayed up several nights for hours answering emails by hand, saying I’m sorry you sent your ether to a random address, etc. Joe and I even made a video to explain what people needed to know:
And they still didn’t pay any attention, they just wanted our tokens.
We got many more like that. We had planned to offer $1.2 million of tokens in our presale. In response to the demand, we took it up to $4 million. It wasn’t enough.
By this time, we were all just trying to get the newsletter and the smart contracts set up and prepare for the storm. The presale took place last Wednesday. It was over in 34 minutes. Here’s the message I sent out to the angry emails I got immediately afterward:
I’m sorry. It happened so fast that MANY people didn’t get tokens. It was more like a lottery than a sale — we were completely surprised. We had 1 successful transaction for every 2 unsuccessful. We did not promise anything. This was what many people thought after Joe made a video about “pulling some strings,” but in fact there never was a reservation system and no token sale so far has had one. It was a misunderstanding based on something a blogger said, not a communication from us. We tried to explain it — check previous emails. Joe set the wrong expectations for 45,000 people and we’ve gotten hundreds of such emails. I’m sorry, but it won’t matter much if you can get tokens in the sale on Saturday — 9% difference will be nothing if this really goes the way we hope.
Thanks, and please consider being nice to us — we’re working very very hard with all volunteers. Most people are being very understanding, a few are very angry. We are doing the best we can under tremendous pressure. We could use your support and hope you get your tokens on Saturday.
We kept making small mistakes. For example, when I made the how-to-purchase video for the presale, I said two-hundred-thousand gas limit, but I mistakenly added an extra zero in the video. Believe it or not, hundreds of people did this, so not only did they not get our tokens they incinerated probably a further $10k wasted on gas for tokens that were no longer available.
Not only that, but peope demanded to see their pillars in their MyEtherWallet, even after we told them several times they won’t get them until after the sale. I would say this has caused at least 600 frustrated emails. And lovely emails, like this (excerpt):
The just kept coming, like we were a retail counter …
Our PR group in London, FamousPublicity, this was their first ICO. They were so excited, I think they didn’t get much sleep for about ten days. They jumped into action, producing our blog, notifying journalists and editors, pumping out press releases and tweets — it was awesome to see, because we’re paying them in pillars, not cash. They managed to connect me with a writer at Wired, who put the Pillar project as the lead in a story on ICOs that is still on the front page today.
Scammers were setting up fake Twitter accounts and web sites and sending out messages saying “Pillar Project Token Sale is Live!” giving an ether address, and I’m sure they got some that way — more money up in smoke. We changed our passwords. The Slack had been locked down already for a few months, and we kicked some unused accounts out of our slacks. We reduced permissions. We set up two-factor authentication, a WhatsApp war room, and sent out phone numbers in case. We had different people auditing and testing our smart contracts, changing little details to get them ready (you definitely want audits and tests, probably also bounties on your smart contracts before putting them online for something like this — plan to spend at least $8,000 on audits of a simple set of smart contracts, and the team at DLTLabs in Toronto really came through for us, hats off to them). Thank god for a little app called CopyClip, which let me manage all my clipboard cuts and pastes and saved me hours in replying to people.
Saturday the 15th was the sale. We are about 30 volunteers working around the clock. Overnight, I think we got at least 500 emails like this:
People are starting to make their own Pillar fan videos now. Jack, our Twitter/YouTube guy, was working 20 hours a day building playlists and answering questions. I don’t know how many times I’ve replied “Please slow down and read what we actually wrote in the newsletter.”
We knew there would be a lot of activity at 8am. We knew we had to guard our smart contract carefully. We put it up around 11pm the night before, so we could test it and I could make the video showing how to buy Pillar tokens. But shortly after that, someone found it, by typing “pillar” into the search window at EtherScan.io, and started sending ether to it. We turned off the smart contract immediately. Luckily, that person didn’t release the address — he was just trying to buy tokens for himself.
At 7:30am, we all got on Zoom and started to coordinate the smart contract, newsletter, and web change in english and Chinese. It was supposed to all happen at 8am. It was like mission control. We were ready.
08:00: I tried to switch the web page, but I couldn’t. We had xxx sessions running on our server, and I couldn’t get WordPress to publish the page. We tried and tried, but the server was showing errors to people, which made them keep refreshing over and over, and soon the server had no chance. We had to upgrade our hosting account and restart everything. Meanwhile we managed to get the email out, which reduced the load, since people had instructions there on how to purchase. The address was immediately on Reddit and BitCoinTalk. We decided to release the address on Twitter, knowing the dangers of doing so. It was going to get out anyway. Finally, I was able to change the web page. The next 30 minutes were spectacular …
We collected about $12 million in ether in 50 minutes.
We made the Mission Control feed from Zoom live on YouTube, so people could see us behind the scenes in action. At one point, over 600 people were watching us manage the deluge. Soon, it became clear to me that we would not have any big “whale” investors at all. What was supposed to be a good mix of small, medium, and large investors turned out to be a grass-roots celebration by small investors — over 7,000 transactions putting in an average of 15 ether (and median of 10 ether), with very few transactions over 500 ether.
I considered renaming the pillar token to be the “Plan D” token, because plans A, B, and C never materialized. So what I thought would be a fairly fast fill-up became more like a telethon. We spent five hours on YouTube, live, telling people about the project, introducing ourselves, and asking them to go ask their friends and followers to purchase tokens. At one point, I put my mic on mute, laid down, and took a nap while the team continued. I think the mood improved then, because I was hoping to see some big transactions.
Something happened on Saturday. We started seeing videos. Not hundreds, but tens of videos coming onto YouTube in a steady stream (now 27 videos and counting). People were making their own YouTube videos and telling their followers. It kept building throughout the day. We made a playlist to track them all.
We ended the day with about 80,000 ether, which steadily dropped in dollar value as the ether price continued to deteriorate. Anyone who had purchased tokens at the beginning had an immediate gain, simply as a hedge against ether. At 2am, I checked in to see what was happening, only to be invited to a Zoom marketing meeting to discuss the plan for Sunday. The day was won by the little investors getting a piece of history:
Our emails showed that the majority of our buyers were Americans. At that time, around 2am, I decided that I no longer wanted to participate in this charade of “best practices” around excluding Americans. Most ICOs now redirect American IP addresses and show them a dumbed-down site with no token offering, and we had complied with legal advice to do that. After I read what Peter Van Valkenburgh, a researcher at Coin Center, said in the Wired article, that it all doesn’t make sense, that most utility tokens aren’t even under the jurisdiction of the SEC, I said forget it, take down the wall. We made it so Americans can come to our web site and see the offer.
Joe also made an excellent video thanking the people at MyEtherWallet for their free product. We agree — it’s the state of the art for token sales, until we can launch the easier-to-use Pillar wallet.
Overnight, we learned that an exchange called EtherDelta was selling our tokens — just purchased that morning — on their exchange for about ten percent under our price. It’s mysterious to see a token listed that is still in an ongoing sale, because anyone selling such a token can only lose money. We also heard that HITBTC was selling futures on pillars.
Sunday — an unfolding miracle. People continue to send ether and buy our tokens. Someone bought 1899 ether worth this morning. Our newsletter is no longer full of mistakes. The complaints stopped coming in. We still have people making mistakes — someone sent his pillars back to our smart contract because he was trying to move them to another wallet (we’re trying to rescue them). Someone else simply sent us 21 ether to our multisig wallet (which is hard to find the address of — you would need to dig it out of our code) and got no tokens in return (we’re working on sending it back). Several people rushed and just sent ether from their exchange account to our smart contract — those tokens are likely lost. It’s just unbelievable what people do and how much of a rush they are in. One person put his private key on a USB stick and then managed to destroy it without a backup — in a single day. It’s hard to destroy a USB stick, especially one with a lot of money on it. People struggling to get their tokens onto a Trezor hardware wallet. And much more. You have to communicate over and over and over — I don’t know how many times I’ve sent the message “Please read your newsletter, please.”
We still have to honor a dumb promise we made to some people about giving them a special discount back in the early days, when we had no idea what would unfold. Despite the many mistakes, we now get about 1 nasty message for every 100 nice, supportive messages.
We’re not going to be anywhere near our cap of $52 million, but you know what we have? We have a community of people who care. We have 80,000 ether in our wallet, and that’s just enough to go to Slovakia, plan and hire, and all converge on London to get started on this amazing journey to build the Pillar wallet for the next several years. Having these people on board is worth at least $20 million in “whale” money, probably more. We’ll be careful with what we have and grow a strong community around the ideals and passion of the Pillar project. We’ll make more mistakes, but we have something many other tokens don’t have — a community of true believers.
In fact, we have financed this entire project — more than 100 people involved, entirely with Pillar tokens, not a dollar of cash. That’s pretty remarkable. If entrepreneurs have to solve the chicken-and-egg problem, I would say we did it with a common purpose and pulling rabbits out of hats every day. Nothing has gone as planned, yet the result has been better than we could have dreamed.
A token sale for an open-source project is a one-time chance to raise almost all the money the project will ever need. Since it has no natural source of income, this money has to last a long time, until you can set up a foundation and collect dues from members. The Ethereum Foundation has done an excellent job, focusing on core code and building community to help flesh out the 90 percent of the ecosystem they don’t build themselves.
Here’s a video I made at Midnight with Mike, a video blogger based in Slovakia, who wanted to help us:
The sale is over. Final numbers: 13446 transactions buying 227,348,800 PLR tokens, raising 113,674.4 ether. Cash value at the time of sale close: $20.3 million.
Mean: 12.75 ETH per buyer
Median 4.35 ETH per buyer (this is insanely low — almost all first-time token buyers)
Only 55 people spent more than 200 ETH
Maximum purchase: 1899 ETH
After the first 30-minute presale and the first hour of the main sale, which together raised about $16 million, we pushed hard for two days and managed to raise another $2 million. The rise in the ether price did the rest.
The minute the funding ended, we heard from someone who received a scam email saying the Pillar sale had been extended two more days and to send ether to their address. We sent a security bulletin to our newsletter around midnight. It’s a guess, but I expect somewhere around $100k to $200k was lost to mistakes and scams in this token sale. And on Monday, we heard that the CoinDash site had been hacked and lost $7 million instantly. I’m very happy how well the team worked to secure our servers and smart contracts — you can screw up a lot of little things, but you have to be vigilant and get the big things right.
Thank you. Thanks to our amazing team of volunteers, who continue to pull rabbits out of hats even as I type this. I cannot say how much this funding event has meant to me and the Pillar project team — it’s a once-in-a-lifetime chance to change the world. Now the project begins. Please come to our web site (open to Americans!) and sign up for our newsletter — we’ll try not to screw it up.