From Zero-to-One: Behind the Scenes of a Tokensale Pre-sale (Part 1/2)
This is the first part of a two-part breakdown we offer the basic premise of our pre-sale, and some basic charts and analysis of the outcome.
With our public tokensale quickly approaching and the pre-sale concluded, we have compiled a breakdown of our public pre-sale. This is not a common occurrence, however in both the interest of transparency and pulling back the veil on these kinds of things which happens so rarely, we are publishing a breakdown of what the pre-sale was, how it performed (and underperformed ).
The first part of this will be just a breakdown of the contract, offering, numbers, and overall process of running the pre-sale. Next, I will look into the data about the pre-sale including comparing crypto-fiat price as a real price and a perceived price while looking at how price and various events may have affected the pre-sale.
This kind of debrief isn’t common, nor are we entirely sure this is a good idea, however at Proof we pride ourselves on engaging in an open business structure allowing our community and supporters to have an actual say in the running of the company. In fact, as a sidenote, we open sourced our business plan and summary here: https://github.com/ProofSuite/WhitePaper/blob/master/Proof_Business_Plan.md.
Likewise, as it becomes more security-appropriate, we will be giving a full architecture rundown of the pre-sale over at our proof devblog.
What is a ‘Pre-sale’?
George Carlin once had a bit about the affix ‘pre’ and how utterly useless it is. In the case of a ‘pre-’sale the linguistic rules of Carlin once again apply. That said, the term ‘pre-sale’ is useful, albeit a tad cringe-worthy. Generally, a pre-sale has a discrete start and endpoint and occurs before another related sale of the same goods. But, then, why do a pre-sale at all?
Most Token Sales, heck, most sales events in the financial world, have tiered sales schedules. Day-to-day customers don’t usually see these but they take on the shells of ‘private-offerings’, ‘dark-pool’, ‘wholesale’, and ‘Christmas Eve’. In some cases, they are held for bulk purchasers who will go on to be vendors of a product and in others they reward loyal customers with rebated purchases prior to the public.
Most of the time, the machinations of a pre-sale are kept secret, its underlying contributors often being venture capital, commercial banks, or close supporters of a project. As projects move forward, open pre-sales have become more popular as they are an effective way to drum up interest and build community based projects. That said, this kind of practice has the potential of opening the doors scams and fraud. This is an underlying motivation as well for this write-up: by being open about the methodologies and outcomes of the pre-sale, we hope to set an example of transparency for pre-sales (and eventually token sales) moving forward.
The Proof Pre-sale
As stated before, Proof is a company that owes its community everything. We hire, speak, fight, and work with our community everyday and have even made pivots from the opinions and observations of our supporters. In the same vein, we wanted to allow our community members who believe in the project to purchase our Proof Token first. Beginning with some of our oldest supporters, The Taas team who we have been speaking to and working with for over a year and who acted as our lead purchasers.
Another reason for our less traditional, less insider-focused, pre-sale is that Proof is a developer-centric company. Although our business development capabilities are more than enough for operations (our dashboard has seen steady growth for the better part of a year), in many ways when we first began this initiative we had to spin up our marketing, partnership, advertising, and PR operations very quickly. Because of this, we have been exponentially building on these capabilities as the pre-sale carried on.
Overall, although not an ideal way to carry out the most ‘profitable’ pre-sale, this method has allowed us to create an open and fair way for our community to purchase tokens based off their assessment of that token’s utility, the team’s verisimilitude, and the product’s growth rather than approaching direct sales to high net worth individuals, exploiting FOMO, or creating yet another pump-and-dump mechanism. It also gives us an interesting look at the organic progression of a token-sale.
These rationales notwithstanding, we are aware that the number of actual data points here is relatively small, and any corollaries stated herein should be considered as hypothesis. However, we will be doing an even more comprehensive version of this after our public sale so the combination of these two documents, hopefully, can be insightful for anyone looking to venture into this fundraising technique in the future.
Moving on, let’s talk about the Proof Pre-sale itself. Here are some hard figures:
The Token
Symbol: PPT
Available Supply: 295,257 Tokens
Re-issuable: No
Tradable: Yes
Minimum Contribution: 10 ETH
Wallet Address: 0x3931E02C9AcB4f68D7617F19617A20acD3642607
Token Address: 0x2469f31A34FCaAc0debf73806cE39B2388874B13
The Pre-Sale
Duration: 14th August ~ 21st September (Extended for RoadShow)
Total number of purchasers: 274 Accounts
Total Tokens Sold: 112,387 Tokens
Ether Raised: 5619.3356 ETH
In USD @ $296.55 / ETH: 1,666,414 USD
Avg Purchase: 147.877 ETH
The pre-sale contract allowed for 295,257 tokens out of the total 2,362,062 Proof Tokens to be purchased at a 43% discount. Of these, a little under half or 112,387 tokens were purchased. From this, the Pre-sale raised a total of 5610.3356[…] Ether. In total, the USD price raised based on Ether prices as of 10 October 2017 was 1,666,414 US Dollars. In the second part we will address how the Ethereum price affected the sale.
Purchaser Breakdown
In the interest of privacy, any information about purchases will only derive from actual press-releases and publicly accessible information (such as from Etherscan). No information from community channels, emails, or private messages will be shared here.
Overall, tokens were issued to 274 accounts.
Given that our pre-sale was handled via a smart contract, exactly matching where each contribution and what time (locally) each occurred is almost impossible. As you can see from the breakdown chart though, the pre-sale consisted of a few bit accounts and a very large number of small contributions (all 274 accounts are in the dataset but their proportion of the whole is so small it does not show up). This means that with our relatively steep barrier to entry (10 ETH), our sale consisted of mainly of large-cap purchasers buying in bulk.
Likewise, if we look at the Etherscan chart (Which is truncated since it is generated starting from Late August due to a problem with registering our token which was resolved), located at: https://etherscan.io/token/tokenholderchart/0x2469f31A34FCaAc0debf73806cE39B2388874B13
We can notice an interesting development:
Our token has been submitted to the etherdelta_2 contract. This means that people are exchanging the proof pre-sale token on a decentralized exchange. After further investigation, we noticed there were a little over 1000 tokens being traded on etherdelta. Although not useful for a flat priced sale, we will definitely be doing some stochastic analysis of our token in the future, lining up actions of the company with fluctuations of any prices on exchanges in order to help build a better model for blockchain company fundamental analysis.
Although this wasn’t ideal for the exclusivity side of the sale, one of our ideals is ‘ownership’ so we allowed the Pre-sale token to be tradable from the outset. We will not be allowing tradability until the end of the public sale, however, since the Proof Pre-sale Token was never meant to hit exchanges and having the token tradable on purchase for the public sale would cause issues with accounting and getting the token on bigger exchanges. In the former case we sacrificed user autonomy for later gain. An uncomfortable but necessary compromise.
Pre-sale Timeline
The Pre-sale started on August 14th and ended on September 21st. In the beginning we intended to end the pre-sale on September 15th, however, we expanded it to September 21st so that attendees of the d10e Kiev conference could purchase pre-sale tokens if they so wished. First, let’s take a look at a naive line chart of the sale (simply looking at transactions as line patterns). The prefix ‘tail’ before the lead purchaser spike is left over ‘test’ transactions we performed prior to the opening of the pre-sale.
In this very simple chart showing just showing purchases plotted into a line graph, we can see that the initial purchase from our leads was the largest input which is unsurprising. As we moved forward, there was a bit of a gap in large purchase, followed by a significant number of smaller purchases until a few larger inputs came in. Now let’s look at a scatterplot with a date-time x-axis and the initial lead purchase removed:
As you can see, we had some sort of purchase each day of the sale (for the sake of scale we have omitted the initial purchase of 870 ETH). Likewise, you can see that the highest concentration of purchases occurred between August 21st and August 30th while large purchases of 100 ETH or greater did not occur until the second half of the sale. Likewise, as the number of larger sales (40 ETH and above) grew, the concentration of smaller inputs had reduced.
Given a limited amount of actual data points. I think it is safe to assume that as the number of purchasers wanting to place the minimum into the sale decreased and we hit the outreach threshold of our marketing and outreach while that threshold began to grant exposure to wealthier purchasers. Another interpretation of this as well could be that larger purchasers began to take notice at the tight cluster of smaller groups and decided to take the leap. And yet another interpretation could be that the Pre-sale simply dangled out there long enough for some people to take notice, which would make sense since this pre-sale lasted a bit longer than many others. Of course there is one more theory I think this sparse but clear data shows us that we will discuss in the Narrative section.
Purchase/Gap Clustering
The great philosopher Lisa Simpson once said, “It’s not the notes you hear, but the notes he isn’t playing” in reference to the late, great, (and fictional -ed) Bleeding Gums Murphy. In this spirit, the final bit of interesting data to mention isn’t the points themselves, but the gaps between them.
Gaps can drive a team mad. Waiting for the next purchase, plugging away at ads and staring at conversion numbers, the space between each token being issued by the contract can create a bevy of conversations about messaging, outreach, and even conversations about pivoting a product itself. When things are moving fast, it feels like they will stay that way forever. When they slow down, you question every decision made up to that point including how you spoke to someone at a conference, an adjective on a product description page, and whether or not you should have even gotten out of bed that morning. In fact, the emotional rollercoaster of running any token-sale at all is very much akin to crypto-participation in general.
Here are some base numbers about the gapping that occurred during the pre-sale:
Minimum Gap: 0 hours
Maximum Gap: 37 Hours
Avg. Gap: 2 Hours
Median Gap: 1
Most Common Gap: 0 (101 occurrences)
The reason to look at gaps for this kind of pre-sale is to measure it against exposure initiatives and to see what kinds of issues may be happening in the smart contract. For instance, there is a large cluster of long gaps (a large arrangement of plots high up at a specific time), this could mean a long string of errors in the smart contract, a possible MIM attack on the support sites, or a strong lack of interest.
Luckily, the Proof Pre-sale had a pretty steady input of relatively large inputs (note that the sale minimum was at least 10 Ether, a not insignificant amount). Another point of note, when placing the Gap Chart and the Sales Chart over each other, Some of the largest token purchases occur after long gaps. What then follows is a fair number of descending larger-to-smaller purchases (characterized by the ‘descending ramp’ pattern seen in the naive line chart at the beginning of this document).
Next up, the good stuff! We will be posting another post giving a little bit of a narrative to these numbers. We will be presenting a cluster analysis of the relationship between ‘perceived’ price and gaps while giving a better sense of the ‘perceptive value index’ of a token sale whose value is so tied to such a volatile currency such as Ethereum. We will also be aligning our pre-sale against others to see if there is any correlation between big news items, both foreign and domestic, and the purchasing habits of customers. Finally, the last part of the post will be our methodology and a warts-and-all assessment of our own pre-sale.
