Software bankers and PE investors: what’s the mood like?
With everything happening in the market what are you all seeing? More deal flow as companies try to figure this out?
Trending Content - Private Equity Forum
| +104 | Future of PE | 27 | 2h |
| +26 | Hardest time I have ever seen to be a GP | 3 | 4d |
| +20 | How to Get on Career Track / Stay Post ASO years | 6 | 3d |
| +14 | KKR comp for Principal | 20 | 1d |
| +12 | MBA and Private Equity | 5 | 19h |
| +10 | Undergrad summer PE analyst positions | 3 | 3h |
| +9 | 2028 Oncycle PE | 10 | 53s |
| +9 | Buyside in London with nice WLB and pay | 0 | 6h |
| +9 | LMM/MM PE London | 5 | 3d |
| +8 | Lindsay Goldberg FT 27 | 8 | 1d |
Career Resources
Not a software banker, and curious if anyone in the industry disagrees, but thinking this is a bit overblown.
I see stocks like Uber and DoorDash getting hit which doesn’t make sense to me. I don’t think it’s the actual software that drives value, but the scale of the platform and amount of users. I’m sure plenty of people even today could code an uber like app, but it’s worthless without the infrastructure and driver network.
Companies like salesforce and Oracle are so deeply integrated into businesses and they themselves will be able to use AI to cheaply develop new features. I doubt vibe coding a salesforce replacement will be feasible for most companies, and even if it is, it’s a risk vs using an established and trusted company.
That’s not to mention the regulatory moats that a lot these companies have. Software players in Fintech, cyber security, etc. are probably insulated for a while
Everyone is stressed right now.
You make a fair point above but people still need to disconnect themselves a bit from the over correction that’s happening in the public markets with software in privates to an extent.
And by that I mean what you’re seeing in the public markets is essentially a rerating of these software names but the correction is drastic. The quiet part that most haven’t said out loud yet is that this same thing will happen in the private markets. Most of these assets, not all, were unjustifiably trading at pretty absurd valuations because of the growth profile and recurring revenue / perceived stickiness of said revenues (even if unprofitable). You’re not mark to market, so it’ll play out slower, but I think most believe that the valuations these assets were purchased at won’t hold. Maybe people hold for longer and hope that the rate of change slows, which it might, but even if it’s only incrementally slower the perception at which software is now viewed and will be viewed going forward in the face of AI is fundamentally different.
Now for my personal view:
I don’t think software just gets wiped out overnight, but the structural bear case is compelling. My main fear is disruption to the core economics of seat-based models that are prevalent today. Most of these companies fundamentally are built around maintaining per-user/seat pricing, predictable ARR growth, and high gross margins. If AI makes each employee materially more productive, seat growth slows or even reverses. In theory companies can shift to usage or outcome-based pricing, but that transition is harder than it sounds. So then you introduce revenue volatility, potential cannibalization, and real variable compute costs that pressure margins that really haven’t been experienced to this extent. So I don’t think we’re at a mass extinction event, but you’ll likely see weaker net retention, weaker pricing power, and margin compression, and ultimately multiple compression from legacy software players.
I agree with your view on regulatory moats but I’d argue that’s realistically a small number of companies in any fund’s portfolio at a given time. You’d be surprised how much niche vertical software exists out there. With all that said, I still tend to lean in the direction that AI impact is overblown but two things can be true in this case.
As others have pointed out, ‘vibe coding’ was a red herring and never the real risk to software companies.
This is a very poor take overall...
Software investor (actively fleeing after spending too much time on AI not to). I disagree with 2/3 of your position.
Nicely done on Uber and Doordash. Pretty aligned there and you’re thinking about their moat correctly. Good names can be hit unfairly during these flash swings. COVID as a great example. Look at FedEx and SPG stock during this time. You basically had to believe luxury mall traffic would get cut by ~70% forever to not buy it at its trough and SPG maintained their dividend that quarter anyway. FedEx trading down that low was actually just brain dead. Prolonged shutdown is probably gonna help a company that delivers to your door, right? Bought both of these names at the bottom, nbd.
I think Uber and DoorDash have more AI risks than you think. I’m so long winded on this topic so will try to tighten up. Open to any follow-ups.
Their risks are mostly longer-term and more second derivative type threats than a SaaS biz. Think the Citrini article… well some of it. I’ve been grappling some of the ideas in there for awhile.
On the traditional SaaS names, it’s bleaker. The risk isn’t somebody just vibe coding it. It’s the rapid progression of coding tools in combination w/ advancements in coordinating agents that are scary and hit term value. I don’t think a solo l33t vibe coder is going to make a product and start beating them in RFPs consistently. Most of the damage will come from companies switching to internally developed SW that previously were too expensive and complicated to build without AI / ‘teams’ of agents. It’ll also allow for better and faster true customization for enterprises. F100 companies are getting whatever features they ask for Salesforce but every customer has a priority ranking and there are only so many engineers to go around right now.
Natural counter is well Salesforce can just have their agents do it for the customer. Totally fair and I’m sure they will BUT it’ll still inevitably cause downward pricing pressure at a minimum.
This is why illiquid software PE investments are in a rough spot. Especially in non-ERP / CRM SW biz. $50k ACV Martech biz that sells well bcs of UX/UI but has relatively simply workflows… I think people or agents are going to be able to replicate these pretty soon. Maybe they’ll only get 95% of the functionality but they can sell it at $5-10k ACV because they have low overhead.
On regulations point, just disagree and there’s a lot going on in these subsectors. Ironically, especially in fintech and cyber. I gotta wrap this up so I’ll just end with “it’s happening almost everywhere if you look” — in a nice way not a look harder way
You must be my MD. Hit on essentially all of the same points simultaneously
lol I just refreshed and saw another long post… “shit did I double post”
I gotta read yours and report back tomorrow. Not an MD… just a mid-level stuck under some guys that are pretending the sky isn’t falling after making shiite investments that would’ve been bad in any cycle. I’m spreading my wings, brother.
This isn’t too dissimilar to how we’re framing software internally at a top BB. Genuinely open to pushback here: I’m broadly familiar but not a deep product expert, as you might expect from an analyst 1 in tech banking.
On Salesforce / systems of record, I feel like the risk might not be dissapperance, but more around revenue growth rates. I feel like the risk is the non-core modules you described getting pulled in-house as AI lowers build costs and enterprises use agents to customize workflows instead of paying for incremental SKUs. However, feel like a system of record is still needed for canonical data, governance, integrations, compliance., edge cases, etc. I wonder if the real question is really about economic capture of agents: does Salesforce monetize the agent layer and deepen embedment, or just become the database agents sit on top of while pricing compresses?
Also wondering do you think growth/VC are more safe from AI-disruption than PE. At least from what I am seeing in terms of tracking companies in banking, a lot more growth and VC firms are investing in AI-native companies / the disruptors than traditiona PE. I think I like the sector as a whole, and largely want to treat the buyside as a 2 year stint before going into tech operating roles, so would love to get a sense of where I can get the best buyside experience to be prepared for that.
Warning — some blunt pushback inbound:
Banks make money from transactions happening… not from owning businesses. Candidly, you could be at a 3-person bank 45 minutes outside of Reno or “a top BB”… either way the buy-side is going to heavily discount your opinion / internal framing because you have ZERO incentive to say anything negative.
Overall your internal framing reads to me like the most positive way to spin a dire situation. Why would a senior tech banker tell his juniors (or his clients) the truth if it would likely lead to low morale or less fees?
Salesforce / SOR point. Similar to my points but again mixed in positivity whenever possible. I disagree with the positive points from from “would I underwrite this working for next 5-7 years.” 99.9999% of us arent buying Salesforce so that scale benefit that still exists isn’t worth pointing to as good for tech investing going forward. All else equal, bigger companies are more stable / die slower than smaller companies… hard to argue with that… your boss is just using a lot of words to say it.
On your buyside question - thanks for asking. The activity in VC/GE is because the AI companies that are “hits” scale very quickly and are hard to capture for many PE funds. The really good ones get acquired by hyperscalers faster than we can get our hands on them (or need more equity than we can justify).
If I were you… I’d try to go to Corp dev at big name tech firm straight from banking given your long term goals. You aren’t even playing for carry if you want to leave in two years anyway. Outside of that I’d target GE to get some modeling skills, followed by VC. Don’t get me wrong VC is great option, it’s just #3.
Somebody unplug this M4 Mac Mini with 16GB of RAM
Traditional softwares ( except those addictive social media & short videos ) are as good as dead.This is the beginning of the end for traditional SaaS companies. Now many SaaS still have decent revenue because of inertia and prior sunken cost -- enterprise customers have an inertia of using old tool even if it has been proven that the old tool is a complete waste of money -- they would still keep using it for another 2-3 years before it's gradually demised. Also those CTO who approved the budget for buying SaaS licenses dont wanna admit that they've made bad decision to waste money on useless stuff so to protect their own political image, they are delaying the execution of demise. However this inertia and bad-decision-cover-up will be short-lived. Inertia will not last forever.
It's funny that these SaaS companies tackle this inevitable demise by pretending to be an AI company, but I've seen first-hand how AI projects are done at these SaaS nonsense -- they are all a total mess ! instead of reducing work and friction, it adds more of those. These companies dont wanna the automation brought by AI, they wanna use AI to increase its product usage and sales, that require them to add extra integration step and extra maintenance overhead to an otherwise simple and low-maintenance workflow. The AI attracts investors because of its automousity and simplicity of interaction ( natural language ) but companies like serviceNow uses AI in ways that make the product less autonomous and make the interaction protocol more convoluted and elusive !!!
The motivation for doing these fake AI projects is that these companies' revenue is directly tied to the number of license they sell -- the more people required to manually maintain its product, the more licenses they sell. , Therefore it is in their best interest to make their product less automated, harder to maintain, and some SaaS companies even create a bunch of extra manual work out of thin air that didn't exist before
Do you see deal flow dying down since people are more worried about software as a whole?
Illo repellat quia aliquid sed delectus. Repudiandae ea quae iste eaque repudiandae dignissimos deserunt. Et aut est atque enim. Neque inventore odio voluptatum quam maiores quia sit. Minima sit sit voluptatem quaerat enim culpa. Excepturi qui molestiae quo voluptatibus assumenda.
Soluta aut omnis ea ut nam. Qui veritatis non mollitia est nisi id omnis.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...