How Far Back Should Your Resume Go

Ask where the "only go back 10 to 15 years" rule came from and nobody has a source. Not a hiring manager who explained what they do with the roles before that cutoff, not a study of which resumes got callbacks. It's the same kind of inherited number as the one-page rule, repeated because it sounds like a standard rather than because anyone checked it against what actually gets you an interview. Treated as a fixed cutoff, it throws away exactly the roles that would sometimes be the strongest match on the page.

What the rule is actually trying to solve

The concern underneath the rule is real, even if the number isn't. A resume that lists every job back to a first internship in 1998 makes a reader wade through two decades of roles to find the two or three that matter for this posting. Early-career jobs are usually the least relevant thing on the page — junior scope, tools nobody uses anymore, responsibilities that don't map to where you are now. Cutting them is usually the right call. The rule gets the usual case right and then applies it as if it were a rule about time, when it's actually a rule about relevance that happens to correlate with time most of the time.

That distinction matters because "usually" isn't "always," and a fixed cutoff has no way to notice when it isn't.

Where a fixed cutoff throws out the wrong thing

Say a posting wants five years of experience with a specific piece of manufacturing equipment, or a regulatory framework, or a niche language that had a run of popularity and then fell out of fashion. If the role where you actually did that work sits at year fourteen, a 10-to-15-year rule keeps it by luck. If it sits at year seventeen, the same rule cuts the single best match on your entire resume to satisfy a convention nobody can source, in favor of keeping a more recent role that's a weaker fit for this specific posting.

This is the same mistake as the one-page rule for the same reason: a length constraint and a relevance constraint get treated as the same instruction, and a rule built for the common case fires even in the case where it's wrong. The fix is the same too. Age is a reasonable proxy for irrelevance. It's not the thing itself, and when a posting tells you directly what it needs, the posting should overrule the proxy.

The one place age is doing real work

None of this means an eighteen-year-old role belongs on the page just because it happens to be relevant to something. There's a separate, legitimate reason people trim old work that has nothing to do with relevance: an early graduation year or a work history stretching back three decades tells a reader your age before they've read a word about what you can do, and that's information a resume doesn't need to volunteer. This is worth being honest about rather than pretending it's purely an editorial choice. If a role is genuinely the strongest match for a posting, keep it — but you can keep the substance without keeping every date marker that comes with it. A role from 2004 can appear without needing the graduation year from 1998 to sit next to it in the education section.

What to do instead of picking a number

Go role by role and ask the same question the rest of a tailored resume already runs on: does this bear on what this specific posting is asking for. Not "is this recent" — does it matter here. An old role that's a strong match earns real space, the same as any recent one. An old role that's a weak match gets compressed to a line or dropped, the same treatment a weak recent role would get. The cutoff isn't a year. It's whether the role is still doing work for this application.

In practice this means keeping more of your history around than any single resume will use, so the choice is available when a posting calls for it. A role you cut for one job might be exactly the one a different posting needs eighteen months from now, once you're two jobs further from it, if you've kept the material anywhere at all. This is the same logic as building a fuller career profile to tailor from: the one-page or two-page version you send out is a selection, not the whole record, and a selection made once and reused everywhere throws away the option to select differently next time.

What this looks like on the page

A few practical patterns follow from treating this as a relevance decision rather than a date cutoff:

• A relevant old role gets a real entry, with dates, title, and the one or two bullets that speak to this posting — the same as a recent role would get, not a shortened version because of its age.
• An irrelevant old role gets compressed to a line or a short list — title, company, years, no bullets — if it's worth mentioning at all for continuity, or dropped entirely if it isn't.
• Education dates and graduation years can be handled independently of whether the work history includes early roles. Omitting a graduation year doesn't require omitting the job that happened to start the same year.
• A string of early, unrelated roles doesn't need individual treatment. "Various retail and hospitality roles, 2001-2006" is an honest, complete answer for a stretch of work history that isn't the point of the resume, without pretending it didn't happen.

The actual question

"How far back should a resume go" doesn't have a single right answer because it's not really a question about years. It's a question about which parts of a career history are still doing work for the job in front of you, and that changes with every posting, not with the calendar. A tool that tailors from your full history rather than your last resume can make that call posting by posting instead of applying one cutoff to every application you send — which is the same reason it's worth keeping the early, no-longer-relevant-yet roles written down somewhere at all, rather than trimming them out of the record the first time they stop mattering.