Context and Objectives
The steem blockchain, backbone of the Steem ecosystem, has been recently upgraded to a new major version, called "Hard Fork 20" (HF20 in short). This post looks at:
- the upgrade process, and
- its outcome
in order to reflect on the future of the Steem ecosystem itself.
In the context of HF20, this reflection is guided by a set of objectives that are made explicit: the success of this Blockchain Community. How does one measure success? When talking about steem, we benefit from a "unit of account" - a significant number among us would agree that STEEM at $10 (or more) is something close to success.
In the next chapters I'll try to briefly explain why I fear we are advancing in a direction which is not the best, and what I think we should rather be doing.
When talking about "achieving objectives in a given context", one usually starts by looking at strategy, then tactics and then execution but here I prefer to do it the opposite way.
I make no secret of it, I believe in this case the path taken by Steemit Inc. is not the best on all three levels, so in a sense this post is a "minority report". I would like readers to see it as a "sanity check" made with the intention of helping this community progress.
Assuming advancing to HF20 was the right strategy and introducing RC (resource credits) was the right tactic, how was the execution?
Well, I don't think you'll find many people defending the botched way in which the HF20 has been executed. It was hurried up in a rather amateurish way, insufficiently tested, and resulted in some serious problems for the whole community, including inability to interact with the blockchain for several hours.
So execution was really, really poor, but what about tactics? The idea of the RC is to reduce spam, to give more granularity over resource control by the witnesses who have to pay for them. Spam has consequences in terms or blockchain bloat and therefore it must be discouraged and should have a high price. However spam did have a price already, in the "bandwidth" parameter of the HF19, only that price was way too low.
Could spam have been addressed by tweaking the bandwidth system and increasing the bandwidth price of certain interactions or for certain spammy accounts? Perhaps, but that is beyond the point. The reason is that both alternatives (tweaking the bandwidth cost or introducing a technically better alternative with the RC) are misguided as they address the consequence (spam) rather than the cause: lack of personal responsibility
See, spam doesn't just "appear" out of nowhere, spam is (ultimately) produced by people. People who spam do so for financial reasons: they want to game the system in order to make money. Of course the first reflex is to "fight fire with fire" and counter their manoeuvres with market mechanics: trying to make spam unprofitable. It might work but I sincerely believe it is shortsighted.
The other way for countering spam - i.e. "bad, anti-social behavior" is to use "social mechanics", thus stregthening what in my opinion is the greatest differentiator and source of value of this blockchain, the community ties.
Back when I joined Steemit, in august 2017, I was encouraged to post a picture of myself holding a paper with the name of the site ("Steemit"), and the date. Such an approach to user onboarding is encouraging users to, later on, behave responsibly. It is offering "social incentives" to interact in a constructive way and cultivate one's reputation.
Of course, it would not have been easy to return to a more "social" approach after so many thousands of spammy accounts were delegated steem power by Steemit itself to come and circle-vote, but it would have been worth trying.
There is a large body of scientific research that tends to show the limitations of purely financial approaches and the benefit of complementing (or even replacing) them with social cues. For those who want to investigate this topic further, I recommend "Predictably irrational" by Dan Ariely
Being Paid vs. A Friendly Favor
In chapters 4 and 5, Ariely speaks in great detail of the differences between social norms—which include friendly requests with instant payback not being required—and market norms—which account for wages, prices, rents, cost benefits, and repayment being essential.
He also explains how combining the two can create troubling situations. The author comments that people are happy to do things occasionally when they are not paid for them. In fact there are some situations in which work output is negatively affected by payment of small amounts of money. Tests showed that work done as a "favor" sometimes produced much better results than work paid for.
For example, some lawyers were asked by AARP to provide needy retirees with services at a cost of about $30. The lawyers did not accept the offer. However, when asked to offer services at no cost, they agreed. Experiments also showed that offering a small gift would not offend anybody (the gift falls into social norms), but mentioning the monetary value of the gifts invokes market norms.
Ariely talks about how social norms are making their way into the market norms. To illustrate, State Farm's slogan, "Like a good neighbor, State Farm is there," provides an example where companies are trying to connect with people on a social level in order to gain trust and allow the customer to overlook minor infractions. The author concludes that "money, as it turns out, is the most expensive way to motivate people. Social norms are not only cheaper, but often more effective as well."
We have established that the tactic chosen to fight spam on the blockchain is at best debatable and at worst misguided. But what about the strategy? Should we, at this point in time, take so much risks (which materialized in a botched execution) in order to fight spam? Was this the number 1 priority?
Let me be clear: yes, I do agree spam here is a serious problem that needs to be addressed. However, as Michael Porter, one of the greatest corporate strategists, used to say, "The essence of strategy is choosing what NOT to do"
The reason for saying such a thing is that "doing something" is costly in time and lost opportunities at the very least.
With a steem blockchain still largely underutilised (according to the blocktivity.info figures) yes, I believe that "fighting blockchain spam" was not the most urgent thing to tackle. That doesn't mean it should not have been tackled at all, but it did not deserve, right now, the status of "number 1 priority"
[source](Sept. 29, 2018)
If not "fighting spam", then what? What should have been the "number 1 priority" in our view?
First of all, let me summarize:
- in the current context, and assuming we agree that we all want the value of STEEM to reach at least $10, if not more
- given the numerous prior hints at how risky the execution of a HF is under existing circumstances (lack of a proper methodology / process as argued by @reggaemuffin and @ura-soul)
- considering the rather debatable nature of the chosen tactic (fighting a financial exploitation with a financial rather than social approach)
I come and say that the best strategy for increasing the value of STEEM is to increase the number of Daily Active Users (DAU).
Steem derives its value from the number of interactions and "value transfers" that this blockchain enables and facilitates - posting, commenting, upvoting, resteeming, following, transferring steem and SBD, powering up, all these actions which compete for the precious time of the users with everything else the world has to offer. The DAU metric captures this "collectively invested time".
It is the collectively invested time that lends value to the steem cryptocurrency.
The tactics for increasing DAU are numerous but most of the changes introduced by HF20 are actually expected to have the contrary effect of decreasing DAU.
Part of the good tactics to increase DAU can be, and are being deployed by dApps. That fact does not in the least exonerates Steemit Inc.: they are the ones with most to gain from an increase in the value of STEEM.
Thus saying "we focus on the nitty gritty of the blockchain because user acquisition and engagement is dealt with my the dApps" is very reckless.
I would like to invite all the witnesses and everybody else who is interested and invested in Steem to at least approach this "minority report" with an open mind. Maybe I'm wrong, but what if I'm right? If I'm right, if indeed the changes brought about by HF 20 decrease the "collectively invested time" and lead to a decrease in the price of steem ... then the leadership of Steemit Inc. is called into question.
The matter is important. How can we tell, what should we monitor in order to determine if I'm right or wrong ?
It is unlikely that the price of STEEM is a good enough indicator in the short term. This price still depends on too many extraneous factors such as the price of bitcoin and general sentiment toward cryptocurrencies. However the evolution in the next months of the number of daily active users and the number of transactions - corrected of the price of STEEM (yes, some co-variance statistical analysis might be needed) - could give a fair indication.
If we see the number of transactions and the number of daily active users picking up when compared to HF19 at a similar STEEM price level, then I stand corrected and I will be relieved!
What I fear is that it will not be the case. That the number of daily active users and the number of transactions will continue to slide (at a comparable steem price). And that the price of Steem itself will not recover to 0.3 mBTC or more as seen only a few months ago.
If that happens, are we ready to have a serious discussion about the leadership of this unique blockchain ecosystem?