Leaving corporate IT does not increase freedom. It redistributes risk. In employment, volatility is absorbed by the organisation. In consulting, volatility is absorbed by you. That sounds abstract. It isn’t.
Salary vs Consulting: The Math People Avoid
Take a realistic scenario. Senior backend engineer in Germany: €85,000 gross salary, paid holidays, sick leave, employer contributions, no sales effort required. Independent consultant at €800/day — not unusual in DACH markets for senior technical roles. The headline day rate looks great. The denominator is where it collapses.
The theoretical annual gross at €800/day across 220 working days is €176,000. In practice, realistic billable days for a solo consultant land between 140 and 170 per year — 65 to 75% utilisation. That gives you €112,000 to €136,000 in gross revenue, before you subtract health insurance, pension contributions, accountant fees, liability insurance, unpaid gaps between contracts, late invoices, and tax prepayments. The delta between that and a €85,000 salary with benefits included shrinks faster than most engineers expect when they first run the numbers.
Utilisation Is the Hidden Variable
Consulting income is not about day rate. It’s about utilisation. Established consulting firms may target 75–85% billable time across their bench. Solo consultants rarely hit that consistently in the first 12–18 months. Reality includes proposal writing, sales calls, scope management, contract negotiation, admin, networking, and discovery time. None of that is billable. Consulting is 30–50% engineering. The rest is business management. That ratio surprises people more than almost anything else.
The UK and Germany Context
In the UK, IT contracting has become less predictable under IR35 classification risk. Many companies that used to engage contractors directly now route through umbrella companies, which changes the economics meaningfully. In Germany, freelancers must handle VAT accounting, quarterly tax prepayments, and insurance directly — administrative load that is not optional and that takes time regardless of how much you’d prefer to be billing. Neither market is hostile to consulting, but both have complexity that isn’t visible from the outside.
The First-Year Spike Trap
Year one often looks good. Former employers become clients. Networks produce warm leads. Revenue spikes. Year two tests sustainability. When pipeline depends only on repeat clients or reputation, volatility increases. Most engineers are not trained in pipeline management — the skill of maintaining multiple leads at different stages of conversation simultaneously, so that contract gaps don’t become income gaps. Most discover this gap in year two, after they’ve already committed to the overhead of the transition.
Psychological Volatility
Corporate stress comes from politics and bureaucracy — the slow meeting, the approval that requires three sign-offs, the colleague who needs to be managed around. Consulting stress is different in texture. Empty-calendar weeks when a contract ends and the next one hasn’t started. A client who delays payment thirty days beyond terms while your tax prepayment is due. A scope dispute where you’ve delivered what was agreed but the client believes the agreement meant something else, and you have to decide whether to absorb it or fight for a change order in a relationship you’re still trying to build. The income variance itself becomes a background anxiety that doesn’t disappear even when things are going well — I’ve spoken with consultants earning substantially more than their corporate peers who described the constant mental overhead of pipeline management as something they hadn’t adequately anticipated. Decision fatigue is real, and it compounds when every decision is also a business decision.
Positioning Is Harder Than Skill
“Senior cloud engineer” is not a proposition. “Reduce AWS cost by 25% in legacy systems within 90 days” is closer. Consulting requires narrowing. Engineers resist narrowing because it feels limiting — if you position as an AWS cost optimisation specialist, what about all the other things you can do? The answer is that without narrowing, pricing power stays average. Average pricing combined with low utilisation produces disappointment. A digital strategy agency operates on exactly the same principle. The narrower the positioning, the more specific the buyer, the shorter the sales cycle, and the harder it is for the client to compare you on rate alone.
Corporate Safety Is Not Absolute — But It’s Structured
Corporate roles are not immune to layoffs. But risk is pooled across the organisation. When downturns hit, discretionary consulting spend is usually cut earlier than core headcount. The consulting business model is structurally exposed to exactly the budget scrutiny that increases during the periods when you most need the income to be reliable.
When Consulting Works
It works when your expertise is scarce and specific enough to justify your rate without a long explanation. It works when you have a savings buffer covering six to twelve months before you need the pipeline to perform. It works when you’ve built a pipeline strategy — not just a network — before you resign, so that the first contract isn’t also the first sales conversation you’ve ever had. It works when your pricing reflects real differentiation rather than “what seems reasonable.” And it requires administrative discipline that many engineers underestimate until they’re doing it.
It rarely works long-term as “general senior developer available for projects.” That market is crowded and price-sensitive in ways that make sustainable margins difficult.
The Actual Trade
Most engineers calculate the upside of consulting. Fewer calculate their own downside tolerance — the specific question of how many consecutive months of low utilisation they could absorb financially and psychologically before it starts affecting decisions. That tolerance is the determining variable, not the day rate and not the theoretical annual gross. The engineers I’ve seen make consulting work sustainably are the ones who answered that question honestly before they resigned, not the ones who discovered the answer eighteen months in.