TWIL (This Week I've Learned) #1
I’ve decided to try a new format - summarising the things I’ve been exposed to during the week. Hope this format will stick.
Rob Henderson - Raising the Cost of Social Support.
Activists use two tactics to cancel their adversaries and get people fired. Sometimes I’m happy I didn’t pursue my PhD 20 years ago and focused on the commercial sector instead.
Time and money
Raising the cost of social support. Making life hell for the people around the target – calling, attacking, intimidating them purely because of their association with the person.
Most colleagues (usually they’re colleagues but can be friends or club members) are afraid to stand up against this abuse because they can become the next target. There are tangible reputational consequences of doing so, which is an unbearable loss for the people.
Fighting activists is time-sucking; defending the target is hard because of it. Time is money, and the longer it takes to come up with a line of defence – the more expensive it is to the defenders. Raising the opportunity cost of defending the target, unfortunately, is a strategy that works well – so far.
Problematic opinions
People are labelled in the “Adjective – Profession – Name” format. “A controversial professor John Smith” is a good example. Read the paragraph about headlines in “Calling Bullshit, Pt. 1” and you’ll get a good idea how powerful and damaging false summaries can be.
Once a person is labelled “provocative” / “problematic” (or a synonym of these words), their activities and outputs (articles, publications, opinions) are marked similarly with obvious negative social consequences.
It immediately becomes risky to get associated with such person by supporting / agreeing with them or expressing similar views.
Richard Waters - Venture capital’s silent crash: when the tech boom met reality.
So far investors and founders are resisting the full implications of a market downturn unless the runway of their business is unacceptably short. In such cases companies have to raise at any valuation (read – down rounds) to survive.
FOMO in VCs caused investing at almost any price as well as performing inadequate due diligence. [MK: when my old company raised Round B in July 2022, the due diligence process took several months, but it was a non-standard case.]
The investors’ assumption was that VC as a class outperformed public stock investment in the long run, hence the right strategy was always to hold and double down. Each new private round allowed VCs to mark up their holding in their companies demonstrating excellent paper returns to their LPs. [MK: It wasn’t uncommon for new rounds to happen solely to mark up the price.]
This created a perverse incentive to keep companies private for as long as possible while not asking for profitability.
The lack of suitable investment options caused capital to flow into initiatives with high uncertainty and very long payback periods (potentially longer than the 7-8 years’ VC fund life). Examples are quantum computing, driverless cars, commercial space and space exploration. The risks are obvious: once LPs become less risk-seeking and stop writing checks, these “moon shots” won’t be financially sustainable.
The biggest disappointments (at least in the US) are ultrafast delivery, fintechs, blockchain ventures. But the clock is ticking for most unprofitable companies, and since even profitable businesses have seen their valuations severely slashed, no one will be unaffected.
Timing for VC is everything, and newly raised funds are likely to show unprecedented losses instead of the super profits they had anticipated.
Profitable companies and those with enough cash on their balance sheet obviously have better prospects of weathering this turmoil. However, expensive and clearly unprofitable initiatives for customer acquisition and market share grabbing will have to be put on hold until VCs develop a new FOMO. Thus, no 3x YoY growth for SaaS startups any time soon.
Evan Armstrong - Snapchat is Possibly Screwed
Snapchat claims to be a "camera" company, but they rely on phone cameras, and their own hardware experiments have all failed (Spectacle, drone). Poor execution and a deliberate distraction for retail investors to obtain the innovation premium.
A "Decade of Innovation" haven't led to meaningful differentiated user growth. Their growth is mostly driven by the release of their Android app in 2018 and the expansion in the "Rest of the World" meaning mostly India again thanks to their Android app. User growth lags TikTok and even Meta.
Their user monetisation substantially lags Meta and Twitter with no clear path to at least match them. Here's why they are unlikely to improving their ad revenue in a meaningful way:
Ad performance (how well the average ad does on the platform and how measurable are those results). It's determined by targeting and attribution. After ATT (Apple's App Transparency Tracking) Snapchat's targeting has deteriorated. It was mediocre to start with (13-34 year-olds make for a not so profitable user bunch), and also the timing of ad serving (between clips) isn't good for e-commerce and transactional advertising. Claiming in Oct 2021 that Apple's changes haven't affected the business, while going 180 degrees 6 months later is an indication that Snapchat doesn't understand the business they're in. Note: bringing attribution in-house may ease the pain.
Ad inventory (what ad placements are available and what the volumes are). Snapchat offers lots of video formats (including AR), but most of them are not scalable: ad copy production should be easy to create. Video is harder to produce than photos and images. Also, video ads also require iterations upon iterations to work; it's simply way more expensive to do it with video. They're pushing only AR ads, which limits their customer base to large companies. SMBs won't bother. But as Snapchat's audience is smaller than Meta's, large advertisers will run out of addressable audience much more quickly. Snap's traffic can be cheaper than Meta's, but the absolute volumes are way lower.
Ad price. As with most real-time auctions, the more bidders there are - the higher the ad prices. Snap doesn't use auction for its AR ads thus leaving money on the table. Brand (non-transactional) ads are broken by the macro conditions.