The IT talent market has cooled noticeably from its 2022-23 peak. The juniors with eighteen months of experience aren’t getting the offers they were two years ago. The people who matter — the senior engineers, the architects, the platform leads who actually keep the lights on — are still in high demand. The retention conversation has shifted accordingly.

I’ve been having this conversation with peers across financial services, retail, and professional services for the last few months. The patterns of what’s working are clearer than they were even a year ago.

Money is necessary but not sufficient

The market correction has not been a free pass on compensation. Senior engineers still command meaningful premiums. Pretending otherwise is a fast way to lose them. But the differentiation between organisations that retain well and organisations that don’t isn’t primarily on compensation — it’s on the things that matter to senior people once compensation is in the right ballpark.

The organisations that have leaned hardest on compensation as a retention strategy have generally not gotten the outcomes they expected. The senior engineer who leaves for $30K more is usually leaving for reasons that have been brewing for months. The money is the trigger, not the cause.

The interesting work question

The single largest theme in what works for retention in 2026 is the question of whether the work is interesting. Senior engineers want to work on problems that stretch them. They want to ship things they’re proud of. They want to feel that the work they’re doing is meaningful and that the organisation is using their time well.

The organisations that retain well are running engineering programs where senior people get to work on the harder problems. AI integration. Platform modernisation. Genuinely complex architectural work. Senior engineers stuck in five years of incremental backlog will leave, and the only question is when.

The organisations that have struggled with retention often have a senior engineering layer that’s been turned into senior project management. The engineers spend their time in meetings, status updates, and steering ceremonies. The actual engineering goes to the juniors and the contractors. The seniors notice this and act on it.

Career paths that don’t require management

Most organisations have improved their non-management career paths since 2020 but the implementation is uneven. The retention question that comes up consistently is whether the principal-engineer track actually exists in practice or only on the org chart.

Senior engineers who don’t want to manage people but want to keep growing need a path that gives them more interesting work, broader scope, and compensation progression without forcing them into management. The organisations that have built this path properly retain better. The organisations that have built it on paper but in practice push everyone toward management retain worse.

The test is simple. Are there genuinely senior individual contributors at the same level as engineering managers, with comparable compensation, comparable visibility, and comparable strategic influence? If the answer is no, the IC track is decorative.

Manager quality matters more than ever

The conventional wisdom that “people leave managers, not companies” is generally true. The IT talent market in 2026 is making it more true. Senior engineers have options. They will not stay through bad management for long.

The manager-quality investment that’s working is multi-faceted. Hiring managers who actually have engineering credibility. Investing in management development. Removing managers who can’t do the job. Holding managers accountable for the engagement and retention of their teams, not just for delivery.

The organisations that have done this seriously over the last three years have noticeably better retention. The organisations that promote based on tenure and avoid the hard conversations about manager quality are losing senior people steadily.

Working arrangements remain a real lever

Hybrid work is the established pattern in most organisations now but the implementation continues to be a retention factor. The organisations that have settled on a clear, considered hybrid model — typically two or three days in office per week, with flexibility for the work that genuinely benefits from co-location — are largely fine.

The organisations that have flip-flopped, that mandate office attendance without a credible reason, or that micromanage the office days are paying a retention cost. Senior engineers who lived through 2020-23 and proved they could deliver remotely don’t take kindly to being treated like they need supervision.

The organisations that have dropped office attendance entirely have a different but real challenge — onboarding juniors and building team cohesion is harder fully remote. They’re aware of this and most are working on it.

Manager-IC ratio and bureaucratic overhead

The ratio of managers to individual contributors has crept upward in many organisations over the last decade. The corollary is more meetings, more approvals, more process. Senior engineers feel the bureaucratic overhead more than juniors because they’re more often the people whose work is gated by it.

The organisations doing well on this are reducing meetings, removing approval gates that don’t serve a real purpose, and giving senior engineers actual autonomy to make architectural and design decisions within their scope. The organisations doing badly are layering more process in response to incidents, audit findings, and risk concerns.

This is a hard area to fix. Process exists for reasons. But process that’s grown beyond its original purpose is one of the things senior engineers will leave to escape.

What new joiners look at

The senior engineers I’ve talked to about why they took their current role say similar things. The reputation of the engineering team. The chance to work with people they respect. The technical scope. The integrity of the engineering leadership. Compensation and benefits matter but mostly as table stakes.

The organisations that recruit senior engineers well in 2026 are the ones that can credibly demonstrate these things. Engineering blog content that shows real depth. Talks at conferences by people who are doing the work. Reputation in the relevant communities. These are not marketing exercises — they’re signals that the organisation is genuinely a good place for serious engineers.

The retention conversation that doesn’t help

The retention conversation that doesn’t help is the one focused on perks. Free lunches, branded merchandise, occasional offsites. None of this moves the needle for senior people. The organisations that lean hardest on perks tend to be substituting them for the things that actually matter.

The conversation that does help is direct. Asking senior engineers what’s working and what isn’t. Acting on the answers. Following up. Demonstrating that the feedback was heard. The organisations that take this seriously have meaningfully better retention than the ones that run engagement surveys and file the results.

Where this goes

The IT talent market for senior engineers is going to remain tight even as the broader market softens. The structural demand for AI engineering, platform engineering, and security expertise is not going to ease. The organisations that retain well will be the ones that take the long view on what makes the work environment good for senior people.

The organisations that don’t will keep losing the people they most need to keep, and will keep being surprised about it.