The first 100 days: what we change after a Cobalt Glacier acquisition (and what we deliberately don't).

The first 100 days after a software acquisition are the most dangerous stretch in the entire holding period. Not because anything dramatic usually happens, but because the new owner is at peak confidence and minimum context. That's the combination that quietly breaks the businesses you read about in the cautionary case studies.

We spent the early years of Cobalt Glacier learning, mostly through small mistakes, what actually has to change in the first 100 days after we acquire a brand — and what we have to consciously stop ourselves from changing. The list is shorter than most operators expect, in both directions.

Why a permanent-capital integration is a different exercise

A traditional PE integration optimizes for a five-to-seven year exit: cost-out, multiple expansion, a bigger-on-paper number for the next process. A roll-up integration optimizes for platform synergies: consolidation, shared GTM, a single tech stack. We've written about the operating differences between those two models and ours in SaaS roll-ups vs permanent-capital holdcos.

A permanent-capital integration optimizes for one thing: the brand being meaningfully more valuable in year ten than it is in year one, with the same operating team in the building. That single objective function rules out almost every aggressive day-one move — and it sharpens the very small set of moves that actually matter.

Almost every integration mistake is a change made too fast by someone with too little context.

What we change in the first 100 days

Four things — and we keep the list this short on purpose.

1. Banking, billing, and back-office

The unglamorous one, and the one that has to move first. New entity, new bank accounts, new payroll provider, new accounting close cadence, clean intercompany agreements. None of this is visible to the customer, none of it changes the product, and all of it has to happen inside the first 60 days so the next four quarters of financials are clean. We move it once, professionally, and we don't revisit it.

2. Security and compliance baseline

Every brand inherits the same baseline on day one: SSO for the team, secrets management, an auditable access log, and a written incident response runbook. If the brand serves regulated buyers, we sequence SOC 2 (or the equivalent) into the first six months whether or not it was already on the roadmap. This is the one place we are explicitly prescriptive — because the cost of a security incident in year two is asymmetric to anything else we'd save by deferring it.

3. The operating cadence

Not the strategy. The cadence. Weekly metrics review, monthly cohort and NRR review, quarterly board, a single shared dashboard, and a written operating plan for the next four quarters. Most acquired companies don't lack ambition; they lack rhythm. Installing a sane cadence in the first 30 days is the highest-leverage thing we do, and it costs the founder maybe two hours a week.

4. The founder's calendar

We protect the founder's calendar from the noise that an acquisition creates. That means we (the platform team) take over investor communications, capital allocation requests, vendor escalations, legal coordination, and most of the back-office back-and-forth. The founder's job in the first 100 days is the same as it was the morning before close: ship product and grow the business. That continuity is the core of the bet we wrote about in Operator continuity is the real asset.

What we deliberately don't change in the first 100 days

This list is longer, and we hold it tightly. The temptation to start moving on day 30 is real and almost always wrong.

• The product roadmap. We do not adjust the roadmap in the first 100 days. The team that built the business has more context than we will for at least two quarters. Our first contribution is asking better questions, not reordering the backlog.
• Pricing and packaging. Pricing changes in the first 100 days look like opportunism to customers, no matter how well-justified the math is. We sequence pricing work into the second or third quarter, with the operator running it and the platform team pressure-testing the model.
• Brand identity. The brand stays. The domain stays. The voice, the positioning, the marketing site, the social handles — all of it stays. We bought the brand because the brand was working. The portfolio page is four distinct brands on purpose, not one mega-brand with four product lines.
• The operating team. No reorgs in the first 100 days. No "platform-wide" hiring freezes. No imported executives parachuted into the org chart. If the team that built the business to acquisition was good enough to write the check for, they're good enough to keep running it through the first year.
• Customer-facing contracts. No master agreement rewrites, no payment-terms changes, no aggressive renewal repricing. Quiet continuity for existing customers buys us permission to do harder things in year two.

The shape of the first 100 days, week by week

Roughly:

• Weeks 1–2. Close mechanics, banking, payroll, intro emails to the team and top customers, founder's calendar handoff. Almost no decisions made.
• Weeks 3–6. Security baseline rolled out, weekly metrics dashboard up, operating cadence installed, written four-quarter operating plan agreed with the founder.
• Weeks 7–10. Cohort and NRR review running, quarterly board structure set, deeper diligence on what we got wrong in our acquisition model, recalibration where needed.
• Weeks 11–14. First quarterly board, first written retro on the integration, candid conversation with the founder about what hasn't worked and what the platform team should do differently in quarters two through four.

At the end of 100 days, the brand looks almost identical to a customer or a competitor. Internally it has banking, security, a rhythm, and a plan. That's exactly what we're trying to ship.

Why this restraint is the strategy, not a stalling tactic

Permanent capital is the only structure that lets you afford to be this patient. We made the underlying argument in Permanent capital is the only sane structure for software and the timeline-as-moat version in The holding period is the moat. If you're holding a brand for a decade, the cost of moving carefully in the first quarter is rounding error. The cost of moving aggressively and breaking the team that built the business is permanent.

Every brand we've acquired has been quietly more valuable a year later than it was at close — not because we ran a clever post-deal playbook, but because we mostly resisted the urge to run one. The operators who built the business kept building. The platform took the noise off their calendar. Compounding did the rest.

The bottom line

The first 100 days are a window to do four small things well and leave everything else alone. Banking, security, cadence, and the founder's calendar — change those. Roadmap, pricing, brand, and team — leave those. The whole permanent-capital thesis falls apart if we get the second list wrong.

If you're a founder thinking about a permanent-capital home and wondering what the first 100 days would actually feel like inside Cobalt Glacier, the most useful next step is a conversation — start one here. If you're a senior operator who has run integrations like this before and wants to embed alongside a portfolio brand, our Operating Partner program is the door.