on
Twitter's PM Woes
It’s been a few months now since Elon Musk took Twitter private, in a much-hyped, mesmerizing, fashion. I think it’s time we take a good look at the disaster this has been.
You can argue there are two major stories here: (i) the commercial angle, whereas Musk revokes bans and goes on random tirades, which led to advertisers fleeing or (ii) the actual product, where decisions on the experience are made without, well, product managers.
I’ll stick to the second, as it’s closer to my expertise. Let’s break each one down…
Checkmarks and Twitter Blue
Where to begin. Before Musk, blue checkmarks served as “official” validation, in a sea of anonimity. Verified users were lighthouses, where you could anchor your trust while the winds brushed ashore all kinds of crypto scams and conspiracy theories. It was actually fun for the rest of us.
But more importantly, checkmarks were coveted by narcissists because they were rare and there was no oficial process you could follow (or pay) to get one. It was like an American Express Centurion Card.
Musk’s idea was to try to capture this value by actually charging for it. The problem is that it’s value is derived from scarcity. As soon as the floodgate is opened, it’s value drops drastically.
Some mid-tier product manager would have argued that it’s actually better to do nothing with the checkmark. You could get creative around the edges, maybe with a new, lower-tier of badging such as a blue circle in your profile picture - earned after verifying your account with an ID. But this second-tier would obviously be less scarce and valuable. It would not detract from the checkmark’s value and would even have the side effect of decreasing anonymity, in theory increasing civil discourse (we would need to carefully test this assumption).
Not having a well thought-out roadmap for the checkmark meant improvisation. Today there is a gold, silver and blue checkmark, which adds confusion and clutter. It also means that the value of a checkmark is now negligible.
Another of Musk’s idea was to pivot to a subscription fee for revenue. Not a bad one in it’s own right, but you need deep product knowledge to understand what users actually value and how much they are willing to pay for it.
As of today, for the insane price of 99€/year, Blue users get:
- Prioritized rankings on conversations and search
- Half the number of ads (aprox)
- Ability to add bold and italic text in tweets
- Post longer/higher-res videos
- Edit tweets, bookmark folders, early access to new features
- Obviously, your blue checkmark
Let’s assume some users (or businesses) will pay for reach, that makes #1 sort-of useful. It’s a small niche, but I grant Elon it’s there.
For #2, why not go all in with no ads? You get this with Youtube, Spotify, Netflix. It adds complexity for users to quantify (did I actually see less ads? what about the quality of ads?). I am sure no one ran the math because there is no way Twitter is extracting more value in ads per user than the subscription they are charging (at a user level).
Then, there is #3. Even if you are adding value for a user, it does not mean you are accretive to the entire site. The complication with social media is that there are so many network effects that changing one thing has multiple ripple effects (some that are unforeseen). It’s a nightmare for a CEO used to working with machines where you tweek an input, and the output changes as expected. I can foresee bold and italics in paid tweets to be perfect for scammers, pyramid salesmen, porn accounts. This is basically making the site worst for everyone else.
Number 4: who cares?
Number 5: Bookmarking might make a good use-case for the type of users who are collectors (a lot which gravitate towards Twitter) but you would need it be a core feature of the subscription, with an interesting experience and maybe cool-er UI for your stored tweets. Even now, there are bots that remind you of tweets, let you schedule, save tweets to other tools like Notion/Evernotes, email tweets, etc. Why not create a tool that aggregates all that extra functionality, allowing the user to manage their account easily, instead of having to randomly reply to tweets? Finally, after the edit option was launched for everyone on Twitter, how is this special or worth paying for?
The debacle of launching a subscription service where the biggest value it offers is pegged to an asset with massive diminishing returns (checkmarks) cannot be overstated and it stems from not thinking thru the product in a disciplined way.
Metrics
In my opinion, another terrible product decision is adding metrics to tweets. You can now see how many times a post was viewed and saved. My question is… Why?
There is no inherent dopamine hit from views, as other social media sites have realized, because it does not come directly from an action of users. In Netflix, views matter because users decided to click to view that content and then decided to stay. In Twitter, users tend to think of views as a result of semingly-random algorithmic decisions. Brands might care because they pay for these eyeballs, but there is a reason why no other social media site features views prominently.
Instead, the virality and social gravitas in Twitter is pegged to the retweet, as it implies your thought is provoking, funny or resonates with others. It is a validation that there is another person who decided to agree with your point of view (”retweets are not endorsements”, yeah right).
You might wonder, if it doesn’t add anything, surely it doesn’t hurt, right? Wrong. More numbers adds cognitive load on users. Even if it’s imperceptible, this minor additional load has the potential to impact time spent on the app. Lower time spent, a metric notoriously hard to move, by just 1% and you have a massive hit to ad profitability. Not to mention other (granted, minor) drawbacks, such as confusion for users, confusion for advertisers and tech debt to maintain the thing (a real-time counter is probably a massive undertaking for a service the size of Twitter).
Edit button
This one is controversial. Users had asked for an edit button for years, and I could even argue it was one of the reasons Elon thought he could do a better job with Twitter.
Allow me to push-back on this train of thought. An edit button, like view counts or italic/bold options is a user-centric improvement, but we do not know about the repercussions to the network as a whole. The obvious problem is mis-information, where you can tweet something viral and then edit it to say something inflammatory. This is solved by tweet history but you are still solving something that needed no solve.
However, a less obvious issue is that it diminishes some of the distinctive character of the product. Take the “covfefe” meme a few years ago. TLDR; Trump tweeted, at 3 a.m. something with the word “covfefe”, which led to millions of tweets mocking his inability to express himself and trying to figure out what he meant. The unscripted errors made Twitter a more “human” and fun place to explore.
It’s arguable that mass edit’s make the product more like Instagram, where posts are polished, filtered and thought-out in detail. If that is where you want to take the product, that is fine (instagram is very profitable), but you do need to accept that it’s not unique anymore.
A product that looks and feels like other social media sites is a big problem if you’re selling ads and you have a sub-scale network. Reddit has understood this perfectly. It just feels different to be on Reddit and that attracts advertisers, even if the amount of users is substantially lower.
When I try to explain to my mom what I do as a Product Manager, it can go either of two ways: she ends up more confused or she says “so, you coordinate stuff?”. It’s not just her. For example, the amount of LinkedIn requests from “normal” (non-tech) companies nudging me to interview in a completely unrelated position, such as procurement (e.g., “you manage the purchase of products, no?”) is staggering.
That the world has no idea what a Product Manager does makes sense. Only a handful of companies really need professionals to think thru product experiences. Let’s be honest: the founder of a startup can do product and for companies where technology is not the actual product, UX designers and engineers are a good proxy.
However, it turns out Twitter is one of those rare companies where Product Managers, for all the blame we take on slowing things down, could have averted some disasters.
I guess Musk’s confusion was similar to my moms when he fired troves of PMs as he took over the company, but in hindsight, it seems kind of a bad idea…