Tabroom Blue™, or, The Death of the Text Blast

Elon Musk loves to inspire headlines. He’s also good at it, whch is nice because he doesn’t appear to be good at much else. Take, for instance, running a busness. It’s a skill you’d think a self styled god-tier entrepeneur might have. But just last year, Musk signed a $42 billion agreement to buy Twitter without apparently understanding it. Perhaps he didn’t even read it. But Twitter’s lawyers sure did, an when Musk got cold feet and tried to back out of overpaying for the company, they took him to court and lined up a sufficiently rapid beatdown that he surrendered and bought the company anyway. He then strolled into the company, spent barely a week in the office, and fired 3/4 of its workforce without bothering to understand what they do.

Next time you want to light $42 billion on fire, Elon, feel free to invest in speech & debate tabulation software instead.

That only started the headlines; they’ve been near nonstop since. Each article is a paragraph in a wider story. Musk’s original goal, the spark that spurred him to sign that awful contract in the first place, was to find an online home for unimpeded right-wing discourse. He found it unacceptable that Twitter and other social networks drew boundaries around Trump and his fiercest supporters. So Musk bought the place. Now the only people who get banned from Twitter are people who annoy him, people he deems too ‘woke,’ or sometimes folks who point out facts about his decisions.

But that policy shift created some serious problems for Twitter. The forces pushing to remove far right discourse from social media wasn’t a diabolical progressive conspiracy in Twitter’s management, funded by George Soros and connected to the Clintons somehow. Musk’s real problem is that as long as social media is paid for by advertisers, alt right speech cannot be welcome.

The dirty secret of the rightmost wing of the Republican party is that they’re a very small part of the overall population. They exploit loopholes and broken rules in the constitution to produce an oversize political impact in the US. Their demographics are skewed older and whiter, and such people turn out to vote in large numbers; they’re more rural than urban, and rural areas are over-represented in Congress and the electoral college. The GOP has spent a few decades cheerfully narrowing voting rights protections so as to produce election wins from an ever smaller base of support. But, that base is small, and when it comes to business, you can’t built much of a market on just that small faction, unless you’re running a scam.

Businesses need to reach a large mainstream audience. And modern far right rhetoric is polarizing: its few fans want to hear nothing but, but everyone else is allergic to it. So engagement and audience sizes on Twitter have been dropping off ever since Musk designated it as a safe haven for his brand of free speech.

Apropos to nothing, you can now find me on Mastodon at https://tech.lgbt/@mildconcern.

Naturally, as customers leave a platform, so too do advertisers. They always chase the largest audience. But it’s not just a matter of how many ads get seen, but what type of ads they are. Before Musk, Twitter’s revenue was about 90% brand ads. Brand ads are not designed to generate clicks and direct sales, the way most ads on Amazon work. Brand ads intend to create positive associations, so you remember the company the next time you need to buy whatever it is they sell. A car maker is a classic brand advertiser. The chances someone clicks an Instagram post and puts a Camry in their cart are very low. But if readers associate Toyota with “reliable” and “comfortable” and “safe” over time, through repeated exposure to brand ads, they are more likely to buy a Camry a year from now when their old car dies and is too expensive to fix.

If you’re running brand ads, you take care that the only associations you build are positive ones. Twitter is a news platform, and when the news ran bad, brands would pause their ads. Nobody wanted their ads to be associated with wars, earthquakes, or mass shootings. As discussion of the weekly tragedy would dampen out, brand ads would come back.

Now, enter Elon. Racism, sexism, or anti-LGBTQ rhetoric once got you banned; now it gets you retweeted by the company’s owner. Almost immediately, that rhetoric was everywhere. Advertisers reacted. Various estimates say Twitter’s brand ad revenue has dropped anywhere from thirty to seventy-five percent. That free speech sure has cost Musk a lot. Note that this process does not require advertisers themselves to personally object to the speech in question. They only care that most customers hate it. The companies and their management could be wildly racist, sexist and homophobic, but if they answer to investors, they’re still going to yank those ads. The bottom line is their only true morality.

So, a conundrum for Musk. Almost all of Twitter’s revenue was brand ads, and brand ads cannot coexist with the type of speech he spent $42 billion to liberate. Perhaps the genius Elon was supposed to be could untie this knot, but our real world Musk has no answer. His main attempt is Twitter Blue: selling blue checkmarks for eight bucks a month. It’s brought in nothing comparable to the revenues he’s lost. And Twitter has almost nothing else it can offer to a paying customer.

So Musk is now lashing out: he’s threatened his advertisers, but then gave ads away — and neither brought them back. His company is drunkenly lurching towards bankruptcy, perhaps as early as this summer. Musk, without any real idea of how to fix it, has crawled into the last refuge of the rich guy: cheaping out on everything.

He’s cut most of the staff, and cut perks, space and benefits for those who’ve stayed on. He stopped paying rent on his offices. Twitter keeps having technical failures because he shuts off servers at random to prove they’re not needed. Hundreds of vendors have overdue invoices sitting on his desk. He’s cutting pennies because he can’t find dollars. It won’t save Twitter, but it’s all he has left.

A few weeks ago, the tech press was baffled when Twitter announced that 2-factor authentication via text message will only be available to those checkmark buyers. Two-factor authentication, or 2FA, means you need more than a password to get access to an account. In its simplest form, a random code is emailed or texted to a user, and then asked as a confirmation of your identity. That should prove you have the phone listed in the account, which gives the site further proof that you’re you. Musk now said text messages only were for paying customers.

The more complicated method, 2FA by means of an authenticator app, was still free. Such apps work by taking a sequence of random numbers, called a shared secret. The site generates the secret, and you copy it into the app on your phone, usually by means of a QR code. After that, it’s never sent over the internet again. Instead, the app combines it with the current date & time, and runs that in turn through some math that produces a six-digit code. The site you’re logging into runs the same math on the shared secret stored in your account. If its result matches the six-digit code you typed in, you’re allowed to login.

Because they rely on timestamps, the correct code changes every minute, making them hard to snoop usefully — by the time you’ve cracked one, it’s changed. And, for Math Reasons, you cannot easily guess the original shared secret from the codes it outputs. The math works only forward, not backward. One-direction codes like that are called hashes, and among other things, it’s how your Tabroom password is stored in our database. When you log in, we can’t tell what your original password is, but we can tell that you typed it in correctly, or else the hashes would not match.

Shared-secret apps are a much more secure method of 2FA than using text messages. Texts aren’t sent encrypted, and it’s proven relatively simple to hijack cell phone numbers, especially if you’re targeting someone in particular. Phone company customer service reps are relatively easy to fool, and have often give account and phone access to hackers. With apps, the shared-secret is never sent to a phone company; it’s never trusted to anyone at all besides you and the service you’re logging into.

That’s why the tech press was confused. Paying customers are being given exclusive access to a less secure system. Most outlets assumed Elon was just being arbitrary and bizarre. Again. Given how much of the last year Musk has spent playing the role of “clueless clown on fire,” they can be forgiven for falling back on that explanation. But I knew right away what it’s all about, because it’s related to a slow crisis that Tabroom has been facing for most of the last year.

You probably think sending a text message is free. They have been free for people with most US phone plans for at least a decade. As soon as smart phones became the norm, and data became the bulk of cell traffic, phone plans stopped charging for talk minutes and texts. Data was where the money was. Who even talks on their phone anymore, anyway? My own ringer has been on silent for about four years, now.

So texts between phones are free now. But sending a text from a non-phone, such as a web system that manages speech & debate tournaments, to pull a random example out of nowhere, it’s more complicated. For Tabroom, I have two basic choices. I could use an SMS/MMS gateway service, which talks directly to the cell providers and sends messages. Or, I can use email. An email sent to 5555551234@messaging.att.net will deliver as a text, as long as that line is on AT&T.

So why not use a gateway? They charge a small amount per text you send. But if you’re Twitter, those charges can add up. Rumors say that Twitter spent $60 million or so per year on these texts, which seems staggering but not when you weigh it against a company that once had $5.4 billion in annual review. That is, until a business leader with no idea how to save his sinking ship takes over and starts attacking every little expense he can find. So, I figured he saw that bill and decided to cut it.

I hope this revelation doesn’t shock you, but Tabroom has never had $5.4 billion in annual revenue. So, we use the free method. But much as Musk’s free speech had costs, so too does free text messaging. Email to text gateways are much less reliable. Messages get tagged as spam, or sometimes just don’t deliver, all the time. I also have to keep track of your cell provider, because each one has a different email address I have to send messages to. When Tabroom users switch carriers their text blasts stop working until they update their account. And of course, when folks who judge tournaments do switch carriers, their Tabroom blasts aren’t usually their first concern, or their tenth. Few judges are constantly yearning to be harassed into hitting their start round button.

Those downsides are all annoyances. However, now Tabroom also faces a legitimate danger: those email gateways are not really intended for bulk messaging. And, spammers sometimes use them to bother phone customers. So the carriers are growing more restrictive. AT&T drops any message with a web address in it. That’s why Tabroom doesn’t send map URLs anymore. And then in late 2022, T-Mobile began capping how many messages per hour a single sender can send them. Which means, if you haven’t noticed, that T-Mobile users don’t get text blasts during the middle of the day most Saturdays; we hit that cap all the time when Tabroom is busy.

I can’t complain to T-Mobile, because T-Mobile’s answer will be “you should be paying to use the SMS gateway.” So, Tabroom hosted 192 tournaments this weekend, with 2356 schools, 21,312 entries and 6,133 judges. Over the course of this Saturday, those 27,445 folks were sent 144,350 text blasts. At that rate, we’d run up a $75,000 annual bill. That’s a lot of money for the right to yell at debaters to be on time.

Is it worth it to increase Tabroom fees a substantial amount for that? Should we try to track who texts a lot and bill them extra — but therefore increase our own overhead tracking every time Menick harasses that judge who never hits start? What happens if Verizon or AT&T follow suit and start limiting us, too? AT&T is already quick to delay deliveries sometimes too, when it decides we’re sending spam. Do we rely on emails and forget texts, when most people don’t get audio notifications of new email anymore? Do I have to write an app for Tabroom just so people get notifications through it — and then pray new judges install it basically so they can be bothered?

I don’t know the answer. But I hope it doesn’t become a crisis. We’ve come to rely on the text message as a prime form of communication, especially during the pandemic and online debates. We need to rethink that a lot in the coming months or years. It’s already tumbling around my brain, for what it’s worth. I’m trying to make sure Tabroom survives longer than Twitter will, at least.

The way Musk is going, that shouldn’t be too difficult.