How Blockchains Will Change Our Understanding of Customer Sentiment

I am giving a presentation on Tuesday at the Sentiment Analysis Symposium on the topic of How Blockchain Technology Will Change Marketing, Sentiment Detection, and Customer Interaction.

The conference organizer, Seth Grimes, is one of the leaders in this space. When he invited me to speak on the topic, I didn’t have a clear thesis about what the answer is. Honestly, I still don’t, but I’m giving the presentation in the hopes of presenting some ideas to the audience and having smart people help me figure it out.

Here’s the rough sketch though. Feel free to chime in.

The key challenge with a blockchain-based world is, in my opinion, the anonymity/pseudonymity of the user. So, getting sentiment of an individual user (as we have become accustomed to in the social media) can prove to be a challenge.

The second challenge is that much of the specifics of a transaction on a network (say what someone bought on OpenBazaar and for how much or whether a Bitcoin payment is for a house or gambling debt are also hidden).

Let’s tackle the first one.In a token-based world, your sentiment towards a decentralized project may actually be easier to measure. If you use the token, acquire more of it, send it to new people (thereby spreading the ‘virus’), and the price goes up,

In a token-based world, your sentiment towards a decentralized project may actually be easier to measure. If you use the token, acquire more of it, send it to new people (thereby spreading the ‘virus’), and the price goes up, sentiment is positive.

So, for example, the fact that DRP tokens (disclosure: client) have seen an average of 2.7 purchases per address during their crowdsale can tell you that the sentiment for the project is higher than another one that may only have 1 purchase per address.

In effect, you can more rapidly compare sentiment between projects than ever before.

Maybe it’s that simple. That doesn’t address how people feel about a website or an Dapp, so there’s probably a usability/UX metric sitting in there somewhere.

Customer interaction, however, may be more difficult. When you buy a token, you don’t release any information. There’s no way for the project to actually interact with you (save those that play the regulatory game in a straight up fashion, but they are in the minority). So, you don’t really have a way to ask people why they buy or gain any insights about different types of customers (people in London or men between 30–45).

That could be a good thing. We’ll move to a true 1:1 marketing world based on account addresses and patterns of behavior (how often that account logs into the app, sends a token, receives a token, etc.) It’s possible that sentiment will be a function of usability, which would fit Nick’s thesis. If you use it more, you like it more. If you use it less, you like it less.

I suppose there’s an angle where an app or a protocol will incentivize people to reveal more about themselves (connected to their address and identity) by paying for it with tokens as 21.co is trying to operationalize, for example.

I could see a scenario where blockchain.info offers me some Bitcoin to connect my wallet to a given social profile (that I can verify is mine-say via OneName ), but it’s more likely that they start building “sentiment clues” into the app. I.e. “if a user of the app does X, he likes us. If not, he doesn’t.”

It’s probably clear that I don’t have all the answers just yet. I’m ok with that. What’s important is to ask the questions and start to think about it from the outside in.

Not “how will I use my current methodologies of sentiment detection in a blockchain-world?”

Rather, “given a blockchain/decentralized world, what does sentiment detection look like?”