Dealing with FUD
FUD /fəd/ — Acronym. Fear, uncertainty and doubt
A bi-product of a successful crowdsale is that a new community can grow as rapidly as the project does. In our case, we had unprecedented traction and from a standing start we collected 250,000 people across social networks and our website, all actively interested in DADI Cloud.
The majority of the community have proven to be grounded, lovely people with genuine interest in the project. But there’s a significant minority who risk ruining the space for everybody — and as the blockchain community grows we believe steps will need to be taken to deal with this.
We’ve seen death threats, racism, sexism, extortion, impersonation, scamming — and a level of vulgarity that would be unacceptable in any context. Some of this has been straight up defamation, and we’ll be dealing with those cases through established channels. The rest of it simply exposes a dark side to a young industry that needs to be stamped out if it is to mature as we all want it to.
Here are a few examples of what we’ve had thrown at us, from abusive language to racism (they happen to be from Telegram but this kind of behaviour was not limited to one social channel):
Why so aggressive?
This is hard to say. Perhaps it begins with the depersonalisation that social networks can provide. People behave in a manner they would rarely have the courage to if face-to-face (or indeed use their real name) — and indeed often when presented with a reasoned response change their manner almost immediately.
Then there is a sense of entitlement. We clearly wanted as many people as possible to participate in our project but could not accommodate everyone. Those who didn’t make it have not technically lost anything, they simply missed an opportunity to gain. Many seem to react by demanding compensation for this when really it is their own emotional investment they are mourning.
Of course we regret that some were unable to take part — and we encourage them to remember that we will listing on official exchanges very soon, and of course there will be an opportunity to earn DADI tokens by contributing to the network later in the year — but beyond this we cannot be reasonably expected to ‘airdrop tokens’ or ‘refund gas’ to contributors that due to the nature of any crowdsale process did not make the cut. We did also take as many steps as we reasonably could to make the process fair, plus redress any areas where we felt there was an issue our side.
How we dealt with it
1) We have a great team
Our team has provided 24-hour cover on our social channels since before Christmas, helping to grow a community from a few hundred at inception to nearly 250,000 before the sale.
We had:
- 10 community admins, most of whom joined us from the community through natural enthusiasm for the project
- The whole core team pitching in, with staff accepting assignments across three continents to ensure we could offer round-the-clock coverage
- 24/7 cover for all channels
- Almost everyone who works for DADI being online during the sale periods
And of course we tried to remain responsive and professional, even in the face of tens of thousands of users and messages across multiple channels. We only banned users as a last resort.
2) We made a bot
During the crowdsale period it became apparent that the Telegram channel would need a dedicated application to handle repetitive tasks and allow busy admins to focus on questions that weren’t already answered in pinned messages or documentation.
Enter the @DADIHelperBot, taking shape as a Node.js application utilising Telegraf to interact with Telegram’s Bot API. ‘She’ was tuned to handle the influx of pervasive images and links spamming the channel by both removing the content, and informing the sender as to its actions.
We soon added slash commands for commonly asked questions, such as /admin for a list of admins and others to cover token prices, links to resources, KYC deadlines and cap metrics. The bot also weeded out spammers (without penalising those users that habitually split their questions over a number of lines or common phrases such as ‘ok’).
Just before the presale started we noticed an influx of users trying to impersonate admins by using identical profile images and similar usernames. To safeguard against scammers, we instructed the bot to listen for new chat members, identify similarities and remove these users.
Our bot also auto-posted important messages throughout the sale at varying frequencies according to how busy the channel was and filtered out bad language (including attempts made by some users to replace letters with numbers, add spaces between characters, or even split the profanity across multiple messages).
Here are some stats from the sale period:
- 18,794 wallet addresses removed
- 380 fake admins removed
- 642 spammers warned
- 212 spammers blocked
- 262,357 images deleted (mostly pornographic)
- 41,360 slash commands called by users
3) We leant on the limited tools our channels provide
Not all social channels make it easy to filter out the noise. On Reddit, we filtered out poor quality submissions by requiring accounts to be established before contributing and requiring posts to have substance. We also increased security by detecting and removing ambiguous URLs such as misleading links or URL shorteners.
We reported inappropriate users to Twitter and Facebook and in some cases are pursuing further action. We have also taken steps to detect impersonator accounts to prevent scamming (and, while we mention it, have acted to remove copycats who have popped up via websites with similar domain names, fake email addresses and even in imposter Google docs).
But it’s the few and not the many
As we said at the start the vast majority of our community are supportive, engaged and enthusiastic — and we are grateful both for the means to engage with you all and the fact you are there to do it. Perhaps the most effective weapon we’ve proven to have during the last weeks has been the community itself, with many quick to step in to defend us against FUD and noise.
We’re sure we’ll continue to work together to weed out the detractors as the DADI network grows over the coming months and years.
Written by Paul Regan. Paul is the Product Director at DADI and is responsible for the strategic development of the Web Services stack.
