The weekly top 10 for B2B tech operators · Every Friday
ASP, or Average Sales Price. ASP tracks the average price new customers pay at sign-up. Unlike ARPA, which includes renewals and expansions, ASP spotlights initial deal size - a key signal for sales performance and pricing strategy. Track it by region, plan, or channel to optimize revenue growth.
SaaS P&Ls are structured in quite specific ways, as defining what goes into the cost of goods sold (COGS) section is important. So take a good read of what the SaaS CFO recommends to include for COGS, as well as how to structure a SaaS P&L. According to the author - 90% of them are structured incorrectly, 5% are close, and 5% are correct.
Another term for your AI Tech Dictionaries, outcome-based pricing. It's a pricing strategy tied directly to the value your product delivers to customers, very much a pricing model driven by the growth of AI.
After a rough October in the tech/Sec world, two AI bigwigs take on the topic. OpenAI's model found 948 Linux kernel exploits in ten days, and its advice is to treat any low-level account on an unpatched box as good as root for a well-resourced attacker. Anthropic's answer is proof instead of testing: Claude formally verified 123 open-source libraries and turned up hundreds of bugs on the way.
Fast follow to last week's four layers of AI (#3) - so who actually keeps the money? Tunguz follows the gross profit from the power plant all the way up to the app, and on his model the top three layers keep 92% of it: apps 40%, neoclouds 28% (before paying for their GPUs) and inference 24%. Power and data centers split the last 8%.
Raising soon? The median is about six months from the first investor opening your deck to announcing the round, per Papermark's data on 24,541 decks. Decks that closed got twice the reading time (36 minutes vs 18), and 24% of decks are still unopened three days after you hit send.
New acronym for our Tech dictionaries - B2A, business-to-agent. We've all spent years keeping bots off our sites, and now they're picking products on their user's behalf. Agents already drive nearly 40% of Stripe's docs traffic, and about 40% of GitHub PRs now start as an agent draft (up from 17% six months earlier). a16z on what changes when the agent is your customer.
It's a whole new world out there - our team completely replaced monday.com with a better, easy-to-maintain tool last month - which leads to the burning question of how can you be defensible when anyone can clone your products - check more on a perspective on that here.
Speaking of #8 - three of these 5 GTM plays are teams building their own tools instead of buying them. Clay's head of growth built deal attribution in about two hours, Vercel's marketing lead built an events agent that took them from 1-2 to 3-6 events a month, and incident.io's CMO swapped his ad agency for a skill he built in a few days.
This one is super useful (at least I'm well aware of it) if you ever find that you need to "warm up" a domain for bulk email or marketing purposes - Resend shares how they warmed up a brand-new domain for a product launch without landing in spam. Great tactics to pay attention to here, including double opt-in, splitting sends across subdomains, and a six-week volume ramp.
Since 2018 - top articles, tips, metrics and research, for B2B tech operators to put to good use. Every Friday.
Done - check your inbox to confirm.
No spam, no fluff. Unsubscribe anytime.
Get the top 10, every Friday
Curated SaaS and tech insight from around the web, repackaged for people to put to good use — free.