It took us two months to understand the tariffs: how companies react to trade shocks faster

By the team at Human Ready · Updated July 2026

When new tariffs hit, most mid-market and enterprise companies take one to two months to answer the basic question: what does this mean for our business? The delay is a structural problem. Shock analysis requires assembling data nobody pre-assembled, building scenarios nobody pre-built, and moving a decision through an organisation that runs on a monthly rhythm. Companies that react in days instead of months don't have smarter analysts; they have pre-modeled driver trees, standing scenario infrastructure, and data that can be interrogated the day the news breaks. This page explains why the two-month lag exists, why 2026's trade environment punishes it harder than ever, and what fast reactors do differently.

Why does it take two months to understand what a tariff means for the business?

The analysis has to be built from scratch, inside a planning cycle that was never designed for shocks. The clearest version of this pattern comes from Human Ready's buyer conversations with finance leaders at European mid-market and enterprise companies, where the two-month tariff reaction has now been described independently by seven different companies.

The canonical case: the CFO of a large European industrial exporter told us his team took two months to react to the US tariff announcements. The sequence was the standard one: wait for the monthly cycle, commission an analyst impact analysis, run brainstorming sessions on the results, and only then converge on a response (in their case, coordinated EU industry negotiation). Two months from announcement to position, at a well-run company with good data. His ask afterwards was telling: an automation that "every morning reads the news and tells me: this happened, here's the impact, here are your two or three options."

The two months decompose into three lags, and each is structural:

  1. Data assembly. A tariff question cuts across the P&L in ways no standard report anticipates: revenue by destination country, margin by product line, landed cost by supplier, currency exposure by contract. The data exists, scattered across the ERP, the planning tool, and a dozen spreadsheets, but nobody pre-joined it for this question. Assembling it is days to weeks of analyst and IT work. (This is the same request-queue-extract-build cycle that makes every new finance question take a week; a shock simply asks five new questions at once.)
  2. Scenario rebuild. "What if tariffs go to 15%?" is a model, not a query. If no driver-based model connects tariff rates to prices, volumes, and margin, someone has to build one under time pressure, and every assumption in it gets debated before anyone trusts the output.
  3. Organisational latency. The analysis moves at the speed of the calendar: the monthly close, the monthly business review, the board meeting. A shock that lands the week after the review waits close to a month before leadership formally looks at it.

What makes tariff analysis in 2026 especially hard?

The target moves faster than the analysis. Consider what a US-exposed European exporter had to model in the first seven months of 2026 alone. In early February, the US Supreme Court ruled that the President had exceeded his authority under the 1977 International Emergency Economic Powers Act (IEEPA), invalidating the reciprocal tariffs built on it. The Yale Budget Lab estimated that without IEEPA tariffs the average effective tariff rate would stand at 9.1%, "the highest since 1946 excluding 2025." Replacement tariffs under Section 122 followed within days, then expired in July and were replaced again under Section 301. As of Yale's July 24, 2026 update, the average statutory tariff rate stands at 11.1%, and under current law, which includes several scheduled increases in the coming months, is set to reach 11.8% by the end of the year.

Three legal regimes in seven months. A two-month impact analysis started in January was obsolete twice before it reached the board.

This is not a niche concern. In the Q1 2026 CFO Survey (Duke University, FRB Richmond, FRB Atlanta), tariffs and trade policy were the top concern among survey respondents for the fifth consecutive quarter, and 39% of firms said their price expectations or realizations for 2025–2026 had been affected by tariffs or tariff uncertainty. The shock is permanent enough to be every CFO's top worry, and unstable enough that last quarter's analysis doesn't answer this quarter's question.

Why do trade shocks rarely arrive alone?

A tariff is a first domino, not an isolated event, and the cascade is what actually has to be modeled. A director at a mid-size European textile manufacturer, with more than half its revenue in the US, described what the April 2025 tariff wave did to his business: the baseline tariff jumped from 6% to 15% overnight, the euro appreciated 10–15% against the dollar (eroding roughly $3M of margin on about $30M of US revenue), credit insurers pulled cover on US clients, and logistics descended into chaos, with containers parked in transit ports and ships rerouting. All of it inside the same six-week window. His summary, translated from Portuguese: "everything crashed at once", "aterrou tudo, aterrou tudo."

That compound texture is why static impact memos age so badly. "What does a 15% tariff cost us?" is the smaller question. The real one is "what do the tariff, the currency move, and the credit squeeze do together, and which of our responses survives all three?" Answering that requires a model you can re-run, not a deck you rebuild.

What do companies that react in days do differently?

Three things, none of which can be improvised mid-shock:

  1. Pre-modeled driver trees. Fast reactors maintain a living model of how external variables (tariff rates, FX, input costs, demand) propagate to revenue, margin, and cash. When the shock hits, the question is a parameter change, not a modelling project. A finance leader at a European industrial group, caught without one during a 2026 geopolitical shock, put the gap precisely (translated): "I would have liked to see the slide – this is our budget, this is the impact under scenario A, B, or C – but it was never made."
  2. Standing scenario infrastructure. Scenarios built once, during a crisis, die with the crisis. Fast reactors keep scenario analysis as permanent infrastructure: a base case plus one or two downside paths, refreshed as conditions move, so "tariffs rise again in October" is an update to an existing scenario rather than a new project.
  3. Question-ready data. The slowest step in shock analysis is almost always data assembly. Companies that react quickly have already connected revenue, cost, and exposure data so it can be interrogated ad hoc, which means the first day of a shock is spent on options, not on extracts. Every answer still has to be trusted, which is why the fast ones also insist the numbers be traceable to how they were produced.

None of this requires predicting the shock. Nobody predicted three US tariff regimes in seven months. It requires being structurally ready to analyse whatever arrives, which is an infrastructure decision made in calm weather.

Where Advisor fits

Advisor, built by Human Ready, is an AI-native advisory platform for mid-market and enterprise finance teams that turns shock analysis from a two-month project into a question. It sits above the existing data stack (ERP, warehouse, BI) and runs scenario modelling on the company's actual data, so "what happens to the plan if tariffs rise to 15% and the euro strengthens 10%?" is answered by deterministic analytical engines against your real revenue, cost, and exposure figures, with every number traceable to source. The external-signal use case exists because buyers asked for exactly that morning brief: this happened, here is the impact on your plan, here are your options. If the last shock took your team two months, that gap is what we work on: humanready.io.

Related reading in this series:

Sources

Page maintained by Human Ready. Last reviewed July 2026.