I’ll be heading out later today for RSA Security in SF and 🔥 up to see what’s in store from customer, vendors, and investors for the rest of 2023. As far as what I’m looking at heading into the conference, let’s start with the most recent Goldman Sachs IT Spending Survey from April 16, 2023.

You guessed it - “Security tops spending priorities again” 

Security remains the top spending priority; other top priorities include SaaS Applications, BI/Analytics, Public Cloud, and Cloud-based Financial Applications. SaaS applications moved up to #2 from #3 and business intelligence/analytics moved down from #2 to #3. Public cloud spending moved up to the third spot from fourth and cloud-based financial applications moved up several ranks from #8 to #5. Public cloud infrastructure saw record adoption relative to prior surveys: Our CIO survey suggest that ~31% of workloads are now on the public cloud (vs 24% in June 2022) with ~51% of workloads expected to be on the public cloud in 3 years (~42% in June 2022). This surpassed the previous all-time high of 43% in December 2019. Azure continues to lead cloud workflow adoption as well as the top position in PaaS adoption expectations. 

And within security, the cloud is on 🔥.

Here’s another look at how the mix has changed but but the Cloud has jumped from #4 to number 1 in just a few quarters.

When it comes to Emerging Security Vendors, it’s great to see portfolio cos Snyk and Security Scorecard prominent here. More thoughts here on what to expect at RSA 2023 from Silicon Angle with data to support.

As for me, I’m 🔥 up to to dive into all things AI and security. This is one of biggest tailwinds we've seen in history and while many large enterprises are excited to leverage LLMs/GPT, they are also incredibly scared. Enterprises need to understand who is accessing data, create policies for how, manage authorizations, create basic classification to block, and in some cases leverage secure computing. We have a number of @Boldstartvc portfolio cos solving some of these problems including @capeprivacy, @ProtectAICorp, @bigidsecure, and @recolabs_ai.

Here’s an example of how to use LLMs like GPT4 on confidential data from Cape Privacy, a portfolio co.

Other ideas I keep thinking about is that while GPT and other models will scan code for vulnerabilities impacting revenue at some startups, on flip side, there will be more code than ever before and if these foundational models are trained on crappy code, this could be a massive shot in the arm for many SAST cos. In addition, now anyone who can write a prompt can sling code which means business users will have to have their code scanned as well - what tools will be available to protect that code from creating new security risks at enterprises. AI will also augment the security analyst helping find that needle in a haystack faster.

As always, 🙏🏼 for reading and please share with you friends and colleagues.

Scaling Startups

1. Risk/Reward - earlier you go the more equity you get in data from 2022 Carta

2. OpenView 2023 State of SaaS Talent Market for download here

   • 8 in 10 candidates are scrutinizing company stability before accepting a role.

   • 89% of C-Suite talent rate financial performance as a top criteria when evaluating an offer.

3. Navigating those pesky CFOs even after your champion says we are ready to go - great advice from Fabio

   OnlyCFO's Software World

   CFOs Are Killing My Deals!

   The good old days of growth-at-all-costs have been dead for a while and a lot of CFOs became a lot more powerful at the same time. Even the best, most tenured CFOs got beat up a bit trying to control costs in 2021 when it seemed like SaaS companies could do no wrong in investors eyes. If you weren’t spending tons of money to grow a little bit quicker then you weren’t playing the game on the field…

   Read more

   2 days ago · 16 likes · 15 comments · OnlyCFO

Enterprise Tech

1. 🤯 the amount of compute!

2. What’s after GPT-4? Hyena? (ZDNet)

   In a paper published in March, artificial intelligence (AI) scientists at Stanford University and Canada's MILA institute for AI proposed a technology that could be far more efficient than GPT-4 -- or anything like it -- at gobbling vast amounts of data and transforming it into an answer. 

   Known as Hyena, the technology is able to achieve equivalent accuracy on benchmark tests, such as question answering, while using a fraction of the computing power. In some instances, the Hyena code is able to handle amounts of text that make GPT-style technology simply run out of memory and fail.

3. get those SBOMs ready - a rare supply chain attack orchestrated by North Koreans

   The massive 3CX supply chain attack, in which hackers hid malware in the VoIP firm's application, began when 3CX was breached with malware hidden in a legit trading app—perhaps the first confirmed case of one software supply chain attack causing another. (Andy Greenberg)

4. On what the new normal could be for Net $ Retention

   Ed Sim 5d Now we have the data from @Jamin Ball - no surprises here as median Net Retention down to 116%, down from 120% - IMO, expect it to go down a lot more… “As you plan your sales models and forecasts for 2023, I’d take a harder look at what NDR numbers you’re benchmarking against as it seems to me that while Datadog hit it’s 130% NDR…” — Ed Sim 1

   with more from my Feb 18 post on what best in class NDR will look like this year

   What's Hot in Enterprise IT/VC

   What's 🔥 in Enterprise IT/VC #329

   Datadog is one of the best executing software infrastructure companies having grown 63% YoY to $1.68B of Revenue in 2022. Q4 was also strong clocking in at $469.4M of Revenue but a YoY growth rate of 44%. While these numbers are fantastic, the decelerating growth in Q4 is certainly an area of concern. Such that management guided down its 2023 growth for…

   Read more

   2 months ago · 16 likes · 2 comments · Ed Sim

5. Google Bard now codes - code generation, debugging and explanation (Google Blog)

6. Stack Overflow Will Charge AI Giants for Training Data (Wired) - joins Reddit in charging for access to data to train algos

7. Get started with AutoGPT in seconds…

Markets

1. Impact of expensing software expenses versus amortizing over 5 years - “Software firms across U.S. facing massive tax bills that threaten tech startup world survival” (CNBC)

2. India’s population to surpass China by mid-2023 according to UN (Guardian)