You have probably heard of the SaaSpocalypse by now.

The (now popular) thesis goes like this: AI coding tools have gotten so good that companies will stop buying software and start building it themselves instead. Every subscription you pay for is something you could have built yourself over a long weekend (what did YOU do this long weekend?).

“Software, as a business, is walking dead.”

This has become an incredibly popular take and the public equities market has begun to price it in. Software multiples compressed hard this year, including at companies that were compounding revenue and expanding margins the entire time.

The market hates not knowing whether the performance metrics it (for now) so prizes will still matter in three years.

I am not going to argue with the market. The market does what the market wants.

I am going to argue with the thesis, particularly for the beauty industry, because it is wrong here in a specific and expensive way.

I have had this conversation with enough brands and manufacturers this year to know it is no longer just a Silicon Valley thing.

Someone internal, usually sharp, usually right about plenty of other things, asks the reasonable question:

“Why are we paying for this? We have an engineer. We have Claude.”

Fair enough. Let me show you the bill.💸

Strip the acronym off SaaS and it means one thing: You pay someone else to take care of a problem so that nobody on your team has to think about it (and waste valuable time outside of their domain expertise to solve it).

When you put your storefront on Shopify, you are not buying a website. You are buying the fact that sales tax across fifty states, VAT in the EU, PCI compliance, chargebacks, and fraud are somebody else's job. Thousands of people at that company think about checkout so your team can think about the thing you actually sell. For $39 a month. That’s a pretty good deal.

We at IRI-Sys think about it the same way but with different business functions.

When you run a formula through a compliance screen, you are not just buying the regulatory report. The report is the easy part. You are buying the fact that somebody read Omnibus VIII. And Omnibus Ingredient II. And the IECIC revision that quietly changed the maximum permitted use levels. And Taiwan's registration phase-in. And the different U.S. States’ PFAS bans.

Claude can definitely code you the interface in a weekend.

But the knowledge, expertise, and vigilance amount to a multi-person team that’s on payroll.

And that’s expensive.

You know I cannot resist some good back-of-the-napkin math, so here it is.

Everyone who pitches an internal build estimates the BUILD line item (see below). Almost nobody estimates the other two.

BUILD: one engineer for roughly four to six months (~$70k), plus regulatory reviewer time to validate the outputs (~$25K), plus the data itself, licensed or assembled by hand (~$10-30k).

MAINTAIN: that same engineer, forever, at some fraction (~$25k/yr). Hosting (~$5k/yr), monitoring, error alerting, backups, bug fixes, developing new requests from colleagues. AI token spend (~$5-10k/yr). Rework of a chunk of the system every single time a regulation changes.

MISS: one relabeling run. One delayed launch. One warning letter, in the worst case. What’s the true cost of that? $50k? $150k?

Kishan wrote about that AI token spend line item in July, when Uber burned an entire annual AI budget in four months and the CEO could not justify the spend.

Uber.

A company with more engineers than most of our industry has employees.

But the line item that actually decides this is MISS because nobody can put a price tag on it before it happens.

If the build case were right, you would expect me, of all people, to have acted on it.

I run a software company. We have a lot of engineers. Our bias toward building is about as strong as a bias can get.

However, we still pay for our CRM, our accounting software, our e-signature tool, our automation layer, and the platform this newsletter is sitting in right now.

We pay more for all of it than we did a year ago, not less.

Not because we could not build them.

Because building them would be the worst possible use of the one thing we are constrained on: engineering time.

Because our industry has a certain asymmetry that makes the build case worse here than almost anywhere else.

If your internal CRM is wrong, a salesperson calls the wrong person on the wrong day. If your internal compliance tool is wrong, your product contains not permitted ingredients in a market you are already shipping into.

That asymmetry is why we have people at IRI-Sys who track regulatory changes across 70+ markets and update the underlying data weekly.

That is not a feature we ship on top of the product.

It is the product.

The software is just how it reaches you.

Which brings me to the question that should be asked:

Across every market we sell into, what are the odds my team catches every change, on time, forever, with nobody assigned to it full-time?

So, to the doomsdayers prophesying the SaaSpocalypse, no, I do not think the regulatory director at a mid-size manufacturer, already responsible for a dozen launches, is going to cancel their subscription, spend a quarter building a worse version with the help of an outsourced engineer, and then own it for the rest of her career.

They are going to keep buying the bread at the supermarket instead of baking it themselves.

Tell me if you agree or disagree: [email protected]

Wella filed to go public, and that is the biggest thing to happen to beauty in the public markets all year. 📈

In the meantime, the strategics stayed busy. Nestlé handed $1B of vitamins to Yellow Wood Partners, Essity bought Kenvue's Brazilian feminine care business, and Symrise offloaded its terpenes unit to Mutares.

The sponsors were busy too. Regent put Avon back together, KKR bought a Japanese platform off L Catterton, and CVC wrote a $220M check into K-beauty distribution.

Let's dive in:

KKR filed The Wella Company for a New York Stock Exchange listing under the ticker $WELA, in an offering Renaissance Capital estimates could raise more than $500M.

The seven-brand haircare platform generated $2.94B of revenue, $522M of adjusted EBITDA and $62M of net income for the year ended June 30, 2026. KKR bought 60% of the business out of Coty (NYSE: COTY) in 2020. Earlier this year, KKR acquired the rest of Wella that it did not yet own.

Yellow Wood Partners agreed to acquire Nestlé's (SWX: NESN) Holistic Health platform of mainstream vitamins, minerals, and supplements brands for $1B. Learn more.

Regent acquired Avon North America, reuniting it with Avon International under common ownership for the first time since 2016; terms were not disclosed. Closing was expected by September 1.

Lisa Siders, COO of Avon International and Regent's operating partner for Avon, becomes CEO of the combined group, succeeding Kristof Neirynck. Learn more.

Essity (STO: ESSITY-B) agreed to acquire the Brazilian feminine care business of Kenvue (NYSE: KVUE) for $284M. Learn more.

CVC Capital Partners invested roughly $220M in Silicon2, a Korean distributor of K-beauty products, to fund international expansion. Learn more.

KKR agreed to acquire Ci Flavors, a Japanese beauty and lifestyle platform, from L Catterton; terms were not disclosed. Learn more.

Skyline Beauty Group acquired men's grooming brands Lumin and Meridian from Pangaea Holdings; terms were not disclosed.

Skyline CEO Joe Chen puts the pair at roughly $35M of trailing sales. Pangaea had raised $87M by 2021, including a $53M Eurazeo Brands check inside a $68M round, with Eurazeo the largest holder going into the transaction. Skyline is on record targeting 3x to 4x EBITDA multiples for acquisition valuations. Back-of-the-napkin math easily shows us that this was a tough sale. Learn more.

Wipro Consumer Care acquired a 60% stake in Dermatouch, an Indian direct-to-consumer skincare brand, at an enterprise value of about $40.5M. Learn more.

Mutares SE & Co. KGaA agreed to acquire the Terpene Ingredients business of Symrise (ETR: SY1); terms were not disclosed. Symrise expects a mid-double-digit million-euro non-cash loss on the sale. Learn more.

CAVU Consumer Partners led an $11M Series A in Dolce Glow, a self-tanning brand, in its first institutional round.

The check extends CAVU's push out of food and beverage into beauty and personal care, alongside portfolio names including Topicals, Nécessaire and 4AM Skin. Learn more.

Goodai Global took a minority stake in Torriden, a Korean skincare brand, in its first outside round; terms were not disclosed.

Goodai is funding the growth while leaving Torriden independent rather than acquiring outright, per KED Global.

APS Innovation (KOSDAQ: 079810) acquired Irae, a Korean maker of cosmetics filling and packaging automation equipment, for $17.7M. Learn more.

Jason Fiedler launched Hilltop, an early-stage consumer fund with a $125M Fund I, writing $5M to $10M lead checks into consumer products and services businesses. Four investments are already done.

Fiedler spent seven years at Left Lane Capital before starting the fund. He is joined by Ellis Fried, who led the Series A in Grüns and worked with that team through its $1B-plus exit to Unilever (LSE: ULVR). Learn more.

River Associates Investments acquired Diamond Wipes, a California-based wet wipe manufacturer; terms were not disclosed. Learn more.

Sauce led a $2.2M seed extension in Be Clinical, an Indian clinical skincare brand, with V3 Ventures participating. Learn more.

Three things I can give you that do not fit in a newsletter.

Beauty Bytes has grown almost entirely by word of mouth, so this is how we say thank you. Share your link, and here is what unlocks:

No purchase and no catch. Just send it to people who would actually read it.

Four sentences. If any of them describe your last two weeks, we should talk.

That is what IRI-Sys is for: international regulatory compliance, formula management and report generation, marketing claims and competitor analysis, and labeling and artwork review. Twenty minutes and I will show you how it works on your own data.