The ICO Handbook

This page is available via www.theicohandbook.com . This essay is part of The Token Handbook, a book on tokens, ICOs, and venture capital written on Medium.com.

If you’re planning an ICO, this is a handbook, guide, and some color from our ICO in July, 2017. Nothing in the ICO world goes as planned. Ours didn’t. Yours won’t. A successful token sale requires skill, game theory, fast reflexes, and luck (mostly luck).

This work is in five parts:

• The story of the evolution of the idea
• The real-time story of our ICO
• This ICO Handbook (this essay)
• A list of ICO resources (to come)
• There is also a recorded webinar on ICO principles and the Pillar ICO

Table of Contents

Part I: Planning

• Offering Terms
• Communications
• Social Media
• Advertising
• Public Relations
• Hackers, Phishers, Spyware, and Thieves
• Smart Contracts and Multisig Wallets

Part II: The ICO Process

Part III: After the Sale

Part IV: Running the Project

Summary

A word of warning: a token sale to the public is not to be taken lightly. There are maybe 200 things listed here that you must do; you could do 199 of them perfectly and then lose all your money and all your customers in an instant. You are always one small step away from disaster, even long after the fundraise. A token sale is a very long-term commitment.

Disclaimer: The information in this article is for informational purposes and does not constitute legal advice or instruction. You take your own risks with cryptocurrencies and tokens. I try to provide the best information I have but cannot guarantee you will be successful when acting according to my recommendations.

Part I: Planning

What does the token do?

It keeps coming down to this single question: what does the token do? As William Mougayar explains, it helps to break this down into three parts:

• Role of the token — how does it work in your system?
• Purpose of the token — what objective does it accomplish? Why is this token the best way to accomplish it?
• Features — what features does it have that make it the right tool for the job?

It’s easy to get sucked into having quick answers to these questions, but if you are just having a token sale to get money, people will smell it a mile away. The vast majority of token sales don’t reach their cap. More than 50 percent barely get over their minimum threshold, and you simply don’t read about them in the press. Keep asking the hard questions: what does the token do? How does it do it?

People are writing about token models. Vitalik recently published a piece about token models and offerings. I have my own framework, which I’ll link to here when it’s out.

Challenge your token

If you really believe in your token, it’s now time to bring in the best devil’s advocates you can find. Have them challenge your assumptions and force you to defend them. If they tell you it doesn’t sound as useful as you think, pay attention. As a result of many back-and-forth battles on what the Pillar token should do, we finally decided to call the unit of one one-millionth of a pillar a curmudgeon, for the two guys who forced us to make sure the token had purpose, a clear use model, and value in the system. Ask:

• Who will want to buy and hold this token?
• Why would people want to use it?
• Under what scenarios will people need this token?
• Can the same job be done better in a different way?

Tokens are Project Finance

Tokens are nothing like stock. Anyone wanting to substitute tokens for the usual fundraising paperwork doesn’t understand what tokens are. Tokens are not a good choice for venture capital or private-equity funds (yet); tokens are not for start-ups. Most tokens are more like tradable Kickstarter commitments for projects — you’re selling a future product to early buyers, and the future token has utility to the buyer. Just as with Kickstarter, the platform or format of the sale does not magically bring buyers. You have to get every buyer yourself.

For example, even though the Pillar project raised over $20 million, the parent company, Twenty Thirty, has almost no assets. Not a single ether we raised can go toward the company — everything has to go to the project. In fact, as soon as we raise some equity money, I’ll go off the project’s salary to reduce drag. I’ll talk about spending money later.

Token Income is Regular Operating IncomeJust as with Kickstarter, you’ll be taxed at the end of the year on your total income minus expenses. This is one reason why so many companies and orgs are located in Zug, Switzerland — corporate income tax is around 14 percent. It’s also a good reason not to raise money after July. The earlier in the calendar year you raise, the longer you’ll have to spend it in the first critical year before having to pay taxes. The government contributes to your spend by reducing your tax liability. This is especially true for open-source projects that have no business model or future income in sight. You’ll want to accelerate expenses if you have no income for the next several years.

Consider this scenario: You sold your tokens when ether was high, and you have to calculate your income based on the average price during the sale. Later, at the end of the year, ether has gone down, so you have an unrealized loss on your holdings. While you consider ether to be “cash,” the tax authorities don’t. You need to sell ether and buy it back just to realize the loss and reduce your taxable income for the year.

Assuming you raise more than a million dollars, and you have an open-source project, the first thing you want to do is make sure to form a tax-exempt entity of some kind before the end of the year. If your token sale is coming up and your project is open-source, form the entity before the ICO. Too many people think about it afterward, but heed my advice: form a tax-exempt entity before your ICO if you’re open source.

Who is the Audience?

There are four main token buyers:

1. Underwriters
2. Investors
3. Traders
4. Natural buyers and users of your future system

You may be aiming for the big whales who have taken many of the early ICOs, but don’t count on that. Fewer and fewer of them are playing the ICO token-flip game. My advice is to aim your entire campaign at your natural buyers and be happy with a lot of them. If you’re the big deal of the day, your ICO will be over in minutes, but consider that a bonus and don’t plan on it.

Going after your natural buyers means about a three month campaign — two months to get the word out and one month of sale. It’s hard to estimate how many man-hours this will take, but it’s well into the thousands.

Is it Legal?

Depending on whom you talk to, it is or it isn’t. There is much discussion of this. Current best practice is to not offer any tokens to US citizens, simply because the SEC is impossible to second-guess and there could be problems later. Recently, the SEC ruled that tokens of The DAO were securities, but those tokens were securities! Very different from the kind of system tokens most ICOs offer. This is a tricky and changing landscape. See the links at the end for many online opinions.

Standard practice is to blacklist American IP addresses and send them to a non-sale site, exclude them from your newsletter, and say you’re not offering tokens to Americans. Certainly, this make sense if your token looks or smells a bit like a security.

However, during the Pillar token sale, I decided this was a wave of fear that was completely unjustified. As I wrote about in my post-mortem, I don’t think a standard ERC20 system or utility token is a security, and I don’t think the SEC even has jurisdiction over it — that would go to the CFTC. So I stood up and said I was willing to accept the risk and open our site to Americans, which we did. I’m not big on shared delusions. But you should consult a lawyer and make your own decision if you’re willing to accept the liability, and it could potentially be significant.

How is it Structured?

Is it an ERC20 token? How many are there? Are they divisible? The best way to answer these questions is to look at other token sales. In general, yes, more than you think, and to 18 decimal places.

What are your Assets?

Think of giving a score to all these things and add up the score …

Running code is a big deal. If you have something working, people will want to try or inspect it.

Customers are a big deal. If, as the Brave browser did, you have millions of happy customers for your free product, you have a great chance at issuing a token and raising serious money.

Open-source. If your token is going to be valuable and there are network effects, you will have to defend not going open source. Most projects are. We knew we wanted open-source code, and we sought out an open-source consultant to help us make sure we were doing things right. We’ll continue with him as our project develops.

Commitment. Your token buyers are looking for your commitment to realizing the goals of the project. If you give your word, they expect you to keep it . Assume every word of your white paper will be read to you in a courtroom. We committed to an $8 million minimum, and we had a very active team on all fronts.

What are your Liabilities?

If it can be said online, it will be said. If there are any liabilities, they will be found. You probably can’t avoid people making fast assumptions and jumping to wrong conclusions. But try to list any liabilities and address them ahead of time.

MarketingThe most important aspect of marketing is to make sure the token has a legitimate purpose and fills a market need. Then you need to communicate that and add social validation. Almost all of the marketing effort takes place before the actual sale event. Here are some of the marketing elements. I’m going to show you the steak as well as the sizzle because, well, it’s marketing, and marketing is mostly sizzle …

The pre-sale is both a signal to the main buyers adding social validation and a way for you to get tokens to people who really want them but may not have the muscle to buy in the main sale. You can see how others have done this — it helps to sell some reasonable but small fraction of all available. You can read about our crazy presale story — it had a lot of moving parts.

Advisors are people who should believe in your project. You shouldn’t pay too much (in tokens) for advisors — they should want to help you anyway. Use your silver bullets and credibility you’ve built up to ask people to join and join for the right reason. I’ve heard some people saying a phone book of advisors smells like shilling, but I’m not convinced. The more people you have on your team, the better, as long as they are legitimate and interested in the project.

Partners help a lot. Companies that are full-fledged partners are a huge asset. Companies that are willing supporters are helpful. Even people who are willing to be on your customer council will help if you can mention the names of the companies they work for.

Road show. If you’re going after the Chinese market, you should go to China. I decided not to, and it probably hurt our sale. It’s hard to know, but live events could be an important part of your communication strategy.

Logos add validity, often tricking people into thinking a sale is more legitimate than it is. I’ve seen ICOs list the logos of CoinBase and Kraken on their front page , not because they are partners or the token is listed there, but because you can change money there! Everyone knows you can change money via these exchanges, but the logos are recognized and lend faux legitimacy to the project. The same goes for logos of press that has mentioned your project, no matter how little — it’s common to see CoinDesk and other press logos if any of those publications has been (paid to) run a piece on the project. The best logos are of real partner companies if you can get permission.

Foreign language translations require commitment. If you are going after a serious foreign market, you’ll need them. Translating your white paper and web site will be several thousand dollars, not to mention creating and maintaining the site. You’ll be making changes and announcements in that language constantly. You’ll need help with social media in that language — there will be a lot of Q&A.

Bounty programs.

You can run a bounty program for signature avatars (people’s “footer” lines in Reddit and BitcoinTalk), tweets, YouTube videos, articles, talking in forums, and more. You can often pay in tokens or a mix of currency and tokens. There are people who help organize bounty programs — better to know what you’re doing and set expectations correctly, because you want those people to feel good about it after the sale. It’s not unusual to see an ICO dedicate 2–4 percent of tokens to payment of bounties (under the marketing budget).

However, many bounty programs pay people who don’t disclose that they are being paid to make content. On the one hand, a bounty program is usually “create whatever you want, be honest about your views,” and those who pay don’t put any restrictions or say not to disclose the payment. But on the other hand, we all know that you get more tokens if you say good things about the project. The United States has a “truth in blogging” law, and that should apply to all social media, especially YouTubers who make money video blogging. It’s fine if you pay people and encourage them to say what they want, but rather than having a “laissez-faire” approach, you should require them to disclose any reward upfront.

We didn’t require disclosures, and we should have. Honesty in media is a serious issue — it gives regulators another reason to do it to us if we don’t do it for ourselves.

It’s not a black/white situation. Some YouTube vloggers were so enthusiastic about our project that they made their own videos, and we decided to reward them afterward with tokens.

Vloggers, tweeters, and other influencers — protect your reputation and include a monetary-incentive disclosure in any content you make.

Online events. I do a lot of webinars, and I scheduled token-specific webinars during our token sale (including one on our own ICO process). The key is to provide good value, not advertising. We ended up doing several team videos using Zoom, where our audience could see us all discussing the project as a group. These turned out to be popular. You can set up and advertise events, and that gives you both a real-time audience who asks questions and an asset that stays online working for you.

Understand buying behavior. People will watch the price of ether and generally buy at the last minute. So if you have discount windows, you’ll sell most of your tokens at the very end of each window. You’ll have a different audience on the weekends than during the week. People are generally followers — they are looking for others to jump first.

Offering Terms

You’ll have to decide how many tokens to issue. Then you need to decide on the offering period. Because people buy at the last second on any deal, and because so many deals were being snapped up in minutes, I made the decision to run a four-week campaign and a 60-hour sale event, to give people a chance to buy on the weekend and weekday without dragging it out. I thought this was going to be the way others do it in the future. I now believe that was a mistake.

I believe it’s better to go after your natural audience and give them time to learn and buy tokens. I would run a 6- to 8-week campaign with 2–4 weeks of sale period.

You should only have one pool of tokens, priced only in ether. People will change other cryptocurrency to buy — you don’t have to worry about that.

The PreSale

The pre-sale should take place a few days before the main sale, because you don’t want the price of ether to move too far between the two. You can price the two differently, but it’s ideal if the price in ether is the same and then in the presale you get extra tokens (buy 10 get 1 free, for example). This makes all tokens the same after the sale.

We planned to sell only about 2 percent of our tokens in the presale, but the demand was so great that we took it up to 8 percent and sold it out in 34 minutes. i would say it’s fine to offer up to 10 percent, and if you only sell $500k worth that’s a pretty good presale. Ours was exceptional.

I was surprised how big a discount certain larger investors were looking for — in the range of 50 percent. We didn’t end up doing one of those deals and didn’t need to, but serious token-hunters are looking for deals before ICOs. However, at least some of those are going to be interested in selling soon after.

Communications

Whatever you’re planning, you’re not prepared for the communication effort. Here are some of the elements:

An explainer video. You need at least a semi-professional 2–3 minute video with professional voice talent. This is not an option.

A white paper. White papers are hard. You can’t farm them out. Our 35-page white paper was easily 200 hours of work in a very short time, and I’ve written a lot of them. If you have a white paper, you must take ownership of it and work with a professional writer to make sure it’s readable.

Articles. The more articles the better. Many publications are pay-to-play, so you can have a front-page article any time you want. We didn’t do that except in China, but you can assume there are almost no “pure journalism” sites dedicated to crypto. I won’t name names, but the “crypto news” sites are all eager to monetize their platforms.

Interviews. I believe interviews are critical to your communication platform. I set up interviews with key journalists and video-interviewers months in advance. They took the interviews because the project was interesting. As the campaign unfolded, video bloggers wanted to interview me and I took every chance I could get. They will continue long after our ICO, and yours should, too.

Social Media

You must take social media seriously. Here’s a breakdown …

Twitter is critical. Not only do you need full-time 24/7 coverage on Twitter, you need to beware of similar Twitter handles and accounts using your graphics to attract your followers to their account. This happened to us.

Slack is both a blessing and a curse. There is the amazing CoinFund slack, where many of the smart token investors hang out. On the other hand, many Slacks are now overrun with scammers and phishers. You must lock down your slack before you announce your token sale.

Facebook is not my first choice, but it probably helped us. If you’re trying to reach people who care about your project, messaging hundreds of thousands of people for not much money could be a win. On the other hand, these could be exactly the people wanting to sell their tokens immediately after the sale.

LinkedIn articles are great if you have thousands of followers, as I do. Plenty of people contacted me on LinkedIn saying they had bought our tokens.

YouTube is awesome for grass-roots campaigns. If you really want to connect with your audience, find ways to encourage people to make videos and make plenty yourself. More on this below.

Reddit and BitcoinTalk are critical. Our project had a 98 percent approval rating on Reddit and BitcoinTalk, thanks to the hard work of the Vindyne Group, who started channels, ran bounty programs, answered every single comment, and were on guard for trouble in two shifts, 24/7. We paid them well and they had perfect integrity, making sure our message got out clearly and wasn’t distorted. They were even interested in our project. They said we made their job easy by being honest and keeping our word. I wouldn’t run a campaign without them.

Advertising

On the one hand, when you see ads for ICOs, do you click on them? I don’t. On the other hand, if you pay $1,000 for ads and get $5,000 in token sales, is that a win? It’s a personal choice. I strongly believe that the big token buyers aren’t clicking on ads to find their next token. On the other hand, the more people see your name out there, the more likely they are to finally decide to check it out, and you need people checking out your token sale!

There are direct-response ads and brand-building ads. Each has a different purpose and different metrics. You may want awareness ads at the beginning and move toward response ads as you see them working.

Personally, I don’t like advertising. I think it shows desperation and attracts token hunters, not natural buyers. In the end, we did run a few banner ads as experiments. I think they helped marginally. But I’ve been wrong before. You should consider advertising and experiments never hurt.

Public Relations

I consider PR far more important than advertising. We worked with the amazing team at FamousPublicity (they have very good systems for managing and tracking everything), and if we had been bold enough to do PR in the United States from the beginning, I would have worked with David Wachsman and Vindyne. You need PR help in your target markets.

We got a lot of articles out and it helped. People respect good, honest writing on Medium.com, LinkedIn, and other platforms. Tell a story. Tell them about their world with your token in it. It’s very important not to communicate about how your token will go up in value but rather talk about the value of your future project and what it adds to society. You’re not selling a token, you’re selling the system. Same as KickStarter.

Customer ServiceYou might think you can handle customer service. We got 10,000 emails in ten days. We had to answer thousands of emails manually. This took almost three full-time equivalents. We stayed up until 3am answering emails from people who were told and promised things we never said or committed to. Much of it was damage control, as people were upset about various messages we sent out, misunderstandings, going too fast, and falling prey to hackers.

Getting them into the newsletter was almost impossible, because after people sign up for a newsletter, they confirm, and their very first issue goes into their spam or promo folders, where they never find it. We thought we would just update our FAQ daily, but we never had time. We focused on the newsletter, but I’m sure hundreds, possibly thousands of people never saw it. We had a slightly higher than 50 percent open rate for people who had signed up — I am sure that can be improved by telling people how to sign up for a newsletter properly from the beginning.

It would be smart to build a customer-service system up front, so you could make sure your most frequent questions are easy to answer. We found ourselves writing “Please see the newsletter for your answer, thank you.” hundreds and hundreds of times. Thank god for CopyClip!

You’re going to end up doing tech support for several wallets. You must be very clear which wallets you support and which you don’t. We only supported MyEtherWallet and said so, but dozens of people wanted help with various other wallets.

You can’t please everyone. Some people will send nasty messages no matter what you do. You will also get thanks, praise, and love from people who are really attracted to your project. Treat them like gold tokens.

MistakesMany people will make mistakes trying to get into any ICO. People will send their ether to almost any smart contract, making them easy prey for phishers. They will then send their tokens anywhere, practically giving them away in an effort to send them to an exchange or another wallet. They will send your smart contract ether directly from an exchange. Even though we made videos showing people step by step how to do everything using MyEtherWallet, plenty of people ignored the videos and most people didn’t follow them exactly.

Your communication here is critical. You have to be careful and clear how you’re dealing with problems and what your policy is. Even though we stated our policy in writing and on video, many people still thought we owed them money or tokens after they had made mistakes.

Hackers, Phishers, Spyware, and ThievesHackers are attracted to concentrated wealth. Hackers love chaos and are willing to help create it. An ICO has at least the following vulnerabilities:

Social media phishing: sites and Twitter accounts that look like yours. If you have a popular ICO, they will “extend your sale” afterward and give people the address to send to. This works. I just want to show you what a phishing site looks like:

This site looks exactly like MyEtherWallet.com and even includes a phishing warning at the top. This is what a phishing site looks like. It looks identical to MyEtherWallet.com. I’m now going to give you the URL of this site, but don’t think of going to that site, because your browser will remember (I did it on a browser I don’t use). Here is the URL:

See what’s going on? A capital i looks like a lower-case l. In a text link on a web page or a fake email newsletter, you’re not going to know the difference. This is causing people to lose millions of dollars’ worth of tokens and ether.

Web sites and newsletters that look like yours. Several ICOs have been hacked exactly at the moment they thought they were going to change their web page, only to find themselves locked out of their own site and a look-alike page with another address inserted. Tell people they should use bookmarks or type in the URL. There’s also a Chrome plug-in called EtherAddressLookup that supposedly validates domain names, and now MyEtherWallet has a Chrome plug-in — if only people would use it.

Your web site is a big vulnerability. You need Cloudflare, two-factor authentication, hard passwords, and only a very few people running the show. Our web site had so much traffic to the domain that we couldn’t change the site — a better way would be to have the ICO site ready to go and change the Cloudflare or domain pointers. I’m not sure how others have done this, but it needs to be done right, because thousands of people could be coming to your site in those first minutes.

Domain names. Hackers will get domain names that are confusingly similar. One new ploy is to use the .cam TLD and your name. For a big ICO, it makes sense to buy a few dozen similar domain names and don’t let them point to your site, so people know they have to spell your domain name properly.

Slack seems to be full of phishing messages. Lock down your slack before you announce your ICO. Remove anyone you don’t know. We were smart to create our own ICO slack with just the people needed, so the main company Slack could go on as usual. We were careful not to share non-public information outside our Slack. If you’re using SlackArchive, disable and delete.

Keyloggers and spyware. Several of our token buyers had their tokens disappear. I guess they had spyware that watched them make their wallets.

EtherScan. You can search for keywords on Etherscan. What does this mean? People can type the name of your project into the search window and find your smart contract by keyword. This way, they can also find the addresses of your holding wallets. If you had not thought of this, you should stop everything now and get professional help.

Presale. We couldn’t tell buyers when our presale was coming. We had so much demand that if we had given an exact date and time, the hackers would have been ready with their fake sites and tweets. This also confused our buyers, many of whom were upset that they missed our surprise sale. A presale should be done as privately as possible, using only your official newsletter and nothing else.

Here’s a list of precautions buyers should take before/after buying tokens.

Smart Contracts and Multisig WalletsYour entire fundraise relies on smart contracts and wallets to store your funds. If you do them 99 percent well, you’ll lose everything. Not only must you have people experienced in these exact types of smart contracts, you must also audit, test, and possibly run a bug bounty. You may want to start with Zepplin’s open contracts and truffle. You may want a one-stop service like TokenMarket or Waves — there are more and more of these now.

No matter what, you need your smart contracts audited and tested. There are several good groups that do this . We worked with DLT Labs, and they were not only exceptionally professional, they also charged a fair price. I would definitely work with them again. I’ll list them on the resources page. They all say the same thing — get your smart contracts ready MONTHS before your sale, because the audit process will almost certainly involve rewrites, and you want the smart contract in place well before the due date. Keep your business logic simple. The fewer lines the smart contract has, the less chance it has to go wrong. Price for an audit should be in the $5,000 to $20,000 range but could potentially be more. Some people are big believers in smart-contract bug bounties; I found the audit and test to be thorough enough to instill confidence.

For example, the presale and sale should be for the same tokens — the tokens should be interchangeable. But both your presale and main sale will use different smart contracts, so you will need to coordinate that. We moved the presale addresses into the main sale contract by hand, which was inefficient and caused delays. Ideally, you use one single smart contract that has different availability on different days, but that also has more complexity and therefore more risk.

Most smart contracts now are set for a certain number of blocks, rather than time-based. This requires a decent estimate and may need adjustment. It’s not practical to estimate how many blocks are left in the next 10-plus days, so you publish the day and time, and then about 24 hours ahead of the end, you estimate the correct block and publish that. It’s better to publish an ending block than an ending time.

It has become common practice to put the testnet address of the smart contract on your web site, so people can find and inspect your smart contracts. Do not do this. Buyers will send money to any contract address they find. If they send money to your testnet address (as many did with ours), they will lose their money. Put up a note inviting technical people to contact you for your testnet address.

I recommend keeping your ether/bitcoins in a multisig wallet, which is now standard practice. Unfortunately, the Parity multisig wallet had a serious flaw and many groups lost significant money to an attacker. For our ether, we use the Gnosis multisig wallet, which we are very happy with. It takes us an hour or two to move any cash out of it, which is a good thing. You can learn more about atomic ownership and multisig wallets in my description of The Lykke Platform.

PART II: THE ICO PROCESS

Everything is ready. It’s time to go live. If you’ve done your job, a lot of people are about to hit your server to discover the public page with your smart-contract address. This is also the time hackers have been waiting for. Chances of something going wrong in the next several minutes are very high.

Chances of something going very right, as you have read in the press about recent ICOs that have been sold out in minutes, are very low. If it happens, congratulations, but chances are this is day one.

Your first job is to protect your buyers from scammers. They will come out at the same time. Your best tool is your newsletter, which has been coaching people for days on how to get into your ICO. A few tips:

• Before you release it, make a video showing how to buy your tokens with the real address. Keep this video away from YouTube on another site or server, where you can keep it private and change the settings just before you release everything else.
• Have people set a bookmark to your site. Tell them to type it in. Warn them that clicking on a link could take them to a page that looks right but the address is wrong. Educate them about phishing more than you think is necessary.
• Tell people that the address will come simultaneously via your newsletter and on your web site, but that the newsletter is far better, as your newsletter is your number-one source of truth. (Make sure that account isn’t hacked.) Make your newsletter message as simple and mistake-proof as you possibly can.
• Respond on Twitter as soon as you see fakes.
• Tell people to set their parameter at 60 Gwei and about 200,000 gas limit and put in their order as soon as possible, and that’s it. Don’t order again. Don’t try a small amount and see (as long as they have the right address). Don’t flail around trying random things. Just order and let the blockchain do its thing. Do not email you asking where their tokens are! Tell them what to expect many times, so they don’t email you asking where their tokens are.
• Unpause and start the smart contract when the time comes. Do not think you can leave it sitting on the blockchain without a pause function.

I can’t give you a playbook for what happens next. Assuming you are going for a “natural buyer” audience, you’ll want to schedule events and generate news to continue to build momentum. It may be fast and furious or it may be slow and steady. But the more buyers you have, the more buyers you can attract on top of that base. It makes sense to have bigger discounts early, but I don’t think the discount should be the draw. It should be the project. The reason your sale is open for several weeks is to continue to get the word out and get new customers who hear about it from those who have just bought. Rally your new base and ask them to help. If you have a great project and are connecting with your audience, they will.

During this time you will be hand-holding constantly, answering the same questions over and over again, being dismayed at how many people don’t open and read the newsletter but instead send individual emails, and generally making everything else harder. It is during this time that you must remember you are actually selling a product and they are your customers, rather than you’re getting money from them to do your project. This may be harder than you realize.

At the same time, all your media channels must be working overtime, building on each other. Use your advisors and their contacts to set up as many interviews as you can get. Watch to see which messages turn into sales. Use A/B testing and different messages on different audiences.

We had live events almost daily, where we would all be on a Zoom screen, livecast to YouTube, and we consistently had 200–500 people watching and asking questions. We had a chance to introduce all the team members and give people a chance to get to know us. Because we were always thinking one step ahead, people had something to look forward to.

To close out the sale, we had a “control room” broadcast and counted down the last minutes. We were surprised how many people watched that! We thanked them and told them that our big kick-off broadcast would be the following day, keeping the momentum.

Part III: After the Sale

Assuming your sale has reached its minimum, there are now two questions coming in as fast as people can type:

Where are my Tokens?

This is easily the number one question. People will have sent you their ether directly from an exchange like Coinbase, in which case you have to tell them their money is gone forever. People will have wallets you don’t support. And it turns out that MyEtherWallet has what could be called a bug or a feature — you may have to click on the white space where the number should be and then it magically appears (people don’t know this). Not to mention that 50 percent of people aren’t opening their newsletters, they prefer to fire off an email. You will be cutting and pasting a lot. Better to put the answer on a web page and send them there than to keep answering the questions one by one.

What Exchange is Your Token Listed On?

At least a few of the people who just put all that effort into buying your tokens are now just as eager to sell them. At the same time, all the exchanges are being bombarded with requests to list tokens, and they are just as picky as they were earlier this year, when they weren’t listing very many tokens. So the chicken/egg problem must be solved again. You must find an exchange to list your token.

Most exchanges have an application page. They want to see that you raised quite a bit of money and there is strong demand for trading. They want to see that your smart contracts are verified on Etherscan. They want to see your tokens on a blockchain explorer. They have other requirements you can’t meet until you’ve issued, so you have to communicate with your buyers while you apply.

At the same time, various exchanges will list your token for sale anyway. EtherDelta is a peer-to-peer exchange, where anyone can offer any tokens or currencies for sale. Even before the sale is over, you’ll see people selling tokens for lower than the listing price. Why? Because thieves and phishers have managed to get your tokens for free from unsuspecting buyers (who at this moment are emailing you, asking what happened to their tokens), effectively dumping them for any price they can get — it’s all upside to them. Meanwhile, your unhacked token owners will email you asking Whiskey Tango Foxtrot is up with people dumping their tokens — is it a crap token? Have they been duped? Should they be trying to sell now, before the price collapses completely?

I wrote in the newsletter that this below-price selling of stolen tokens isn’t surprising and isn’t worrying. In fact, we also found people who purchased in the main sale selling their tokens at a loss. Once these early sales are absorbed, natural market forces will take over. You want demand. You want a bit more demand than supply. But you don’t want too much demand, or speculators will take your token for a ride. And you have to estimate this about three months ahead, without knowing the cryptocurrency market dynamics or market size on the days surrounding your token offering.

This may give you some appreciation for what happens, but if you’re on the front lines answering all these emails, it’s pretty frustrating. You just wanted to raise money for your project. Now you’re being accused of price manipulation, fraud, theft of the tokens, and being rude, insensitive, and dishonest because you won’t agree to give new replacement tokens to people who had them stolen. These are your customers. You sold them a product that they may not be happy about, regardless of your integrity and intentions to build a great service later. I recommend having customer-service people who can cheerfully and patiently answer emails (and work on your communication to reduce the need).

Part IV — Running the Project

Congratulations. Having run the gauntlet above, you now have to answer to token holders, employees, partners, and future customers. Let’s take each in turn.

Token HoldersYou will have two kinds of token policies, each of which has an effect on the token supply:

Fiscal policy is the buying, spending, icing, discounting, and burning of tokens. If you retain a lot of tokens in the treasury, people will wonder when those will flood the market. If you pay people in tokens, they will dump them. If you pay in locked-up tokens, you will also have to supplement those with some cash. All of this requires clear communication, so people know what’s coming.

Monetary policy is the overall set-up: how many tokens there are, how divisible, and the inflation scheme. If there is a fixed supply forever, as with most tokens, then you can control the price from the treasury by buying and selling tokens as a liquidity provider on several exchanges. I don’t think it’s a good use of raised funds, but if your token is on a big exchange and trading is thin, you may want to purchase tokens when they come on the market. Ideally, you find a liquidity provider to do this for you.

I will talk more about inflation schemes in my upcoming token-model series, but almost all token sales go with a fixed number. We did something a bit different. A token sale is a one-time event affected by hundreds of factors that can result in a lackluster sale or a blow-out. You have to plan what to do with unsold tokens. The first thing that comes to mind is to burn them — get rid of them so the money supply doesn’t have any extra tokens waiting to come back into circulation later. But I think there’s a better way: put all your unsold tokens on ice for 5–7 years. This has a lot of advantages. In case the project does well, you’ll have a good supply to sell into future demand. You’re sending the right signal and saving seed corn for later at the same time. You’ll have tokens to seed good projects that want to build on your platform. And you don’t have a market-cap rebalancing problem, where people’s tokens are now suddenly worth more because the supply has gone down. There are other philosophies, but I think icing unsold tokens gives a good balance.

EmployeesYou may have a lot of money in your wallet, but you are now running a very public project. You must make all expenses very transparent. Imagine you’ll soon be interviewed and asked about any expenses that look suspicious. Salaries, in particular, should be reasonable. I think they should be tiered — we are shooting for four salary levels. It may or may not be legal to pay people in tokens, but if you do pay in tokens, employees must book that as regular income for tax purposes. Make sure employees understand the liquidity situation and don’t dump their tokens. Better to pay people in cash than in tokens they are just going to sell.

CustomersYou also have customers to answer to. If you can make customers visibly happy and communicate that to token holders, your token may rise slowly but steadily. This has happened to most of the famous projects — it’s about showing the progress you’ve made with the money and setting expectations. Don’t focus as much on the technology and product as you think — you should also focus on market adoption, messaging, and pivoting often so you don’t get too far out over your technology skis looking for product-market fit.

Token StewardshipBefore and during the sale, you don’t want to talk much about the tokens. You want to talk about the project. Now, after the sale, you must be careful what you say and do. On the one hand, what’s good for the token is good for the business — anything that increases the value of the token should be the right thing to do. But this is not true. This is what you should not do. In general, you should ignore the token price and focus on achieving visible project milestones that everyone can see will lead to the network effects you promised.

You can’t control the price, but you may be able to control a) the supply to some extent, and b) the expectation of supply. These fiscal measures should be very clear and communicated long before you do anything. The more you have “skin in the game” and don’t cash out, the more people will believe in you and your mission. Projects where the majority of tokens belong to the founders are problematic.

About LiquidityPeople don’t appreciate the role liquidity plays in markets. It seems like a good idea to list your token on as many exchanges as will accept it, but think about liquidity. We have a real problem in the crypto-economy: the bigger exchanges get bigger, while smaller exchanges must struggle to get customers. And bigger exchanges don’t necessarily have a better experience or lower spreads, they are just bigger, and bigger is better.

This is a problem Omega One and the Pillar project hope to solve. Until then, you should list your token on the largest exchange you can get it on, simply because that’s where the liquidity is. Having your token on several exchanges fragments the market. I wish people would appreciate the liquidity problem more — it’s great for the big exchanges, who are printing money, it’s bad for everyone else.

Money ManagementAssuming your token sale has been successful, you will have to manage the funds. Neither should you cash straight into fiat currency, nor should you stay 100 percent in ether. You may be sure ether will rise, but you’d be foolish to keep all your eggs in that basket. Here’s how to build a smart, diversified portfolio:

Fiat currency. If you really have a lot of cash, you should probably only be about one third in crypto altogether. The rest should be extremely diversified in ETFs and indexes across currencies, locations, and strategies. A smart-beta and barbell approach are a good combination.

Petty cash. You should cash out whenever you see a decent rise in the price of ether or bitcoin. Not much, but enough to meet your projected needs for the next three months. If the price goes down, hang in there and draw down your petty cash, and if the price goes up, replenish your three-month supply. If you have raised quite a lot, change that to six months.

The token economy. You already probably have a pretty large exposure to your own tokens, as you should. If you can, and if you’re building a platform, give or award some of your tokens to other start-ups to finance their building parts of your ecosystem. Obviously, they shouldn’t dump the tokens all at once — you could have a pay-as-you-go funding approach. There are good reasons not to invest in other ICOs but rather use your money and tokens to get direct results that apply to your project.

Your native cryptocurrency. Our native currency is ether, which is very volatile. Yet long term we believe in it. We plan to retain a significant portion of our holdings in ether.

Build a diversified portfolio. A decent percentage of your holdings in crypto is okay, as long as it’s diversified. That’s why we have built the CryptX index, which you can read about at TokenFactory.io. We plan to be the first customers. There are other indexes, but I haven’t seen one that is properly diversified and weighted.

The number one thing you should have is an investment committee with diverse, outside views and some qualified statisticians/quants. A large enough portfolio should include hedges. It’s easy to get too far out over your skis and not manage your holdings well. Don’t trust anyone who is paid when you trade. Please see my webinar on postmodern portfolio theory to learn more.

CommunityAny good open-source project has a strong community of supporters. In the same way that Kickstarter can help start companies and build brands, your token sale can bring you hundreds or thousands of active supporters. Get them into an online community (we use RocketChat) and engage them! Ask them to help you execute, hire them, use them as beta testers, and find ways to use them as early indicators for your market. I’d rather have $20 million and 10,000 active supporters, which is what we got, than $30 million and no supporters. I’m not just saying that — we expect our supporters to help us solve the chicken and egg problem when we go to market — something investors and traders won’t do.

Summary

Every ICO story will be different. The market is changing. By the end of 2017, I don’t think it will be anything like the beginning. We could see a wave of huge ICOs in the coming months, or at the beginning of the year, or something could happen to ruin the party for everyone. It’s a complex adaptive system, an arms race that keeps everyone going up to the next level, including the hackers. There are no rules, but there are many warnings. I end by telling Nassim Taleb’s tale of the turkey.

The turkey lives on a farm and clucks around all day, a pleasant life, and a pleasant farmer who feeds him, gives him water, cares for him when he is sick, and cleans up after him. The turkey has everything he needs, including female turkeys and plenty of corn to eat. He lives an uneventful life. Until one day, when the farmer carries him into a shed and hacks his head off with an axe.

Whatever happens to you in this black-swan, nonlinear, fat-tailed world of ICOs and cryptocurrencies, you most likely won’t see it coming. Good luck out there.

This essay is a chapter in The Token Handbook.

David Siegel is a serial entrepreneur from the United States. He is the founder of Twenty Thirty AG and the Pillar project, both of which have newsletters you may wish to subscribe to. 20|30 provides consulting services and is building the CryptX — a broadly diversified portfolio of cryptocurrencies in a single token. His full bio is at dsiegel.com.

Disclaimer: The information in this article is for informational purposes and does not constitute legal advice or instruction. You take your own risks with cryptocurrencies and tokens. I try to provide the best information I have but cannot guarantee you will be successful when acting according to my recommendations.