Technical consulting has a deserved image problem: the 200-slide deck, the framework with a trademark, the recommendation to begin a transformation whose first phase is more consulting. Companies pay for clarity and receive theatre — which is a shame, because the genuine article is one of the highest-leverage purchases in software.

Good consulting is compact: a person who has actually built the thing before, reading your real code and your real constraints, telling you the uncomfortable truth while it is still cheap to act on. The deliverable is a decision you can defend, not a document you can weigh.

The moments it pays for itself

Before a major build: an architecture review that catches the scaling wall or the wrong-tool choice pays for itself hundreds of times over — the cheapest fix in software is the one made before the code exists. During due diligence: for acquirers and investors, an honest read of a codebase, its debt and its team risk turns a leap of faith into a priced decision. At the crossroads: rebuild or refactor, monolith or services, buy or build — decisions expensive to reverse deserve an experienced outside eye with no stake in the answer.

And in the fractional gap: companies too small for a full-time CTO but too technical to lack one entirely borrow the judgement — hiring bars, roadmap sanity, vendor evaluation — at a fraction of the cost.

How to tell the real thing

Real consultants look at artefacts, not just slides about artefacts — the code, the pipeline, the incident history. They give recommendations specific enough to be wrong, with trade-offs stated rather than hidden. They will tell you not to do the project that would have paid them. And they are comfortable pairing with your engineers on the actual fix, because advice that cannot survive contact with the codebase was not advice.

The wrong kind is identifiable too: deliverables measured in pages, urgency toward long engagements, and a strange allergy to looking at the code itself.

Why it matters more now

The decision space has widened — AI capabilities, build-versus-buy, platform bets — and the cost of a wrong technical direction compounds faster than ever. A few days of genuinely experienced judgement at the right moment is the cheapest risk reduction available to a technical business.

Consulting is not a deliverable. It is borrowed scar tissue. Buy it from people who earned theirs building.