Every week I’ll provide updates on the latest trends in cloud software companies. Follow along to stay up to date! Budgets are Bifurcating Last week I showed some data that highlighted the struggles of software companies. Companies are missing guidance at a historically bad rate, and there wasn’t a clear / obvious answer why. Did macro really get worse? Are AI native software companies starting to eat everyone’s lunch? Did a Q4 budget flush pull forward Q1 deals and create Q1 softness? I think the answer is actually something different - budgets are bifurcating.
As one who worked in the office of the CIO of a Fortune 10 company for 14 years before I retired...there is a certain way in which IT budgets are managed at large enterprises. It is a bit more complex than just funds being diverted to AI and away from legacy software. There are two more dimensions to the decision 1. Time for both spend and ROI tied to the annual business cycle of the company. 2. Dependence on the legacy software for running the core business.
This year, 2024, AI projects are a mix of trial balloons led by the innovation team and new strategic investments led by the app dev teams. The CFO does not expect much from the former...perhaps accepting of a 90% failure rate. The latter have say two business cycles to prove themselves before being shut down.
Legacy platforms will continue to receive funding if they are core to running the business.
Most industries and companies have a peak sales season/month/quarter and all their large/strategic tech projects are usually planned in advance leading up to that peak business period. e.g. Retailers have the Holiday season, Back to School, Easter etc. Healthcare has the Medicare and commercial enrollment seasos...insurers are focused on enrollment, hospitals negotiate reimbursement rates about 6 months in advance etc. Consumer tech has the Holiday season and Back to School.
So most of today's AI projects will be focused on achieving some strategic business goal(s) leading up to their business peak season in 2024 and 2025. This gives the CIO about 2 years to show ROI from AI. i.e. 2 business cycles. They wont shutdown a project in one cycle.
Now layer on top of this the annual budgeting calendar. 2025 IT budgets are fairly well decided by May 2024, then they get refined quarterly in Aug and Dec...this lines up with the quarterly earnings report timeline and forward guidance / revenue targets / sales targets / bonus plans / hiring plans / expense plans etc...on a rolling 18-24 months basis. So companies will publicly guide for 2024, but internally they know what they expect to do in revenues in 2025 and they work their budgets backwards from this 2025 revenue outlook.
The point being that software sales decisions are not necessarily being made today based only on today's information. CIOs knew last year where they were going to spend their IT budgets. This should not have been a surprise to the software vendors if their sales teams were paying attention.
The second point is that AI spend has a little more runway for now. The question we are all trying to decipher is when CIOs will be asked to cut AI projects if they are not producing results.
Do you factor in stock based comp when doing rule of 40 or your FCF calculations in order to see how companies really stack up? It would be a great future piece.
This is a great follow! Hats off for your work!! "The Cloud Cast" by Brian Gracely turned me on to this substack ... which is a must follow as well!
Question - What are your thoughts about Consulting Companies selling Gen AI Transformations into the Fortune 1000? Take a Accenture that has said they trained all of their consultants Gen AI (something like 80hrs) to lead transformations ... then the CEO called for basically flat revenues and the stock sold off 10%. It seems like it's a matter of time but not now? ... also maybe like software there's only so much $ to go around.
This is an insightful roadmap of how enterprises are choosing one software vendor over another, the vendors that are flying the AI flag prominently winning and the vendors that are not waving the AI flag getting put on hold. And the purchasing analysis by the enterprise has as much to do with perception (CIO looking good for AI purchases and bad for non-AI/legacy purchases) as it has to do with actual AI functionality. As an investor it would be great to have a metric that shows how much each software vendor spends on AI, such as the portion of their data center/cloud spend that occurs on advanced accelerator hardware vs conventional non-AI hardware. Or compare Cost of Revenue growth vs revenue growth. Spend for advanced accelerators is probably costlier than conventional data center hardware so if Cost of Revenue is growing faster than revenue then the difference can be attributed to AI workload.
As one who worked in the office of the CIO of a Fortune 10 company for 14 years before I retired...there is a certain way in which IT budgets are managed at large enterprises. It is a bit more complex than just funds being diverted to AI and away from legacy software. There are two more dimensions to the decision 1. Time for both spend and ROI tied to the annual business cycle of the company. 2. Dependence on the legacy software for running the core business.
This year, 2024, AI projects are a mix of trial balloons led by the innovation team and new strategic investments led by the app dev teams. The CFO does not expect much from the former...perhaps accepting of a 90% failure rate. The latter have say two business cycles to prove themselves before being shut down.
Legacy platforms will continue to receive funding if they are core to running the business.
Cheers!
Great insight. When you say business cycles do you imply quarterly horizons or annual horizons?
Most industries and companies have a peak sales season/month/quarter and all their large/strategic tech projects are usually planned in advance leading up to that peak business period. e.g. Retailers have the Holiday season, Back to School, Easter etc. Healthcare has the Medicare and commercial enrollment seasos...insurers are focused on enrollment, hospitals negotiate reimbursement rates about 6 months in advance etc. Consumer tech has the Holiday season and Back to School.
So most of today's AI projects will be focused on achieving some strategic business goal(s) leading up to their business peak season in 2024 and 2025. This gives the CIO about 2 years to show ROI from AI. i.e. 2 business cycles. They wont shutdown a project in one cycle.
Now layer on top of this the annual budgeting calendar. 2025 IT budgets are fairly well decided by May 2024, then they get refined quarterly in Aug and Dec...this lines up with the quarterly earnings report timeline and forward guidance / revenue targets / sales targets / bonus plans / hiring plans / expense plans etc...on a rolling 18-24 months basis. So companies will publicly guide for 2024, but internally they know what they expect to do in revenues in 2025 and they work their budgets backwards from this 2025 revenue outlook.
The point being that software sales decisions are not necessarily being made today based only on today's information. CIOs knew last year where they were going to spend their IT budgets. This should not have been a surprise to the software vendors if their sales teams were paying attention.
The second point is that AI spend has a little more runway for now. The question we are all trying to decipher is when CIOs will be asked to cut AI projects if they are not producing results.
Do you factor in stock based comp when doing rule of 40 or your FCF calculations in order to see how companies really stack up? It would be a great future piece.
All numbers that I see here appear to be non-GAAP numbers which imply, SBC is part of FCF and other bottom line metrics that you see here.
This article has a good chart on the difference between reported and adjusted for SBC.
https://open.substack.com/pub/nextbigteng/p/when-fcf-margins-dont-tell-the-full-story?utm_source=share&utm_medium=android&r=nmh0
the 9.7x revenue multiple for high growth companies on the chart doesn't match up to the table - it refers to LTM vs NTM for mid/low growth
This is a great follow! Hats off for your work!! "The Cloud Cast" by Brian Gracely turned me on to this substack ... which is a must follow as well!
Question - What are your thoughts about Consulting Companies selling Gen AI Transformations into the Fortune 1000? Take a Accenture that has said they trained all of their consultants Gen AI (something like 80hrs) to lead transformations ... then the CEO called for basically flat revenues and the stock sold off 10%. It seems like it's a matter of time but not now? ... also maybe like software there's only so much $ to go around.
This is an insightful roadmap of how enterprises are choosing one software vendor over another, the vendors that are flying the AI flag prominently winning and the vendors that are not waving the AI flag getting put on hold. And the purchasing analysis by the enterprise has as much to do with perception (CIO looking good for AI purchases and bad for non-AI/legacy purchases) as it has to do with actual AI functionality. As an investor it would be great to have a metric that shows how much each software vendor spends on AI, such as the portion of their data center/cloud spend that occurs on advanced accelerator hardware vs conventional non-AI hardware. Or compare Cost of Revenue growth vs revenue growth. Spend for advanced accelerators is probably costlier than conventional data center hardware so if Cost of Revenue is growing faster than revenue then the difference can be attributed to AI workload.
The question remains why the market is septic about software companies.
- they Are not convinced that customers are buying their AI and it’s promised business impact.
- they believe a shift to build vs. buy due to reduced development cost with AI.
Friday's must-read from Jamin as always...