The Careers Page Test

Ten minutes, two tabs, and the most honest employer brand audit you’ll ever run Every employer brand audit I’ve ever seen commissioned came back as a deck. Sixty pages, a competitor analysis, a perception survey, a set of recommendations, and an invoice with a comma in it.

And almost all of them dance around a piece of evidence that’s free, public, permanently updated, and sitting one browser tab away from the asset being audited.

So this week’s perspective is a test. It takes ten minutes. It requires no budget, no vendor and no approval. And I’d argue it tells you more about your real employer brand than most of those decks.

Open two tabs.

Tab one: your careers page. The one that went through brand review, legal, and three rounds of stakeholder feedback. The one with the photography.

Tab two: your reviews. Glassdoor, SEEK, Indeed, wherever your people go to say what they actually thought. The tab nobody approved.

Read them side by side. Slowly. The distance between them is your real employer brand.

The misconception: the brand is what we publish

Underneath the discomfort this test produces sits a belief so widespread it’s practically the industry’s founding assumption: that an employer brand is an asset the organisation creates and controls. Build the pillars, craft the narrative, publish the page, and the brand exists.

I’ve spent a large part of my career making exactly those assets, so I say this with affection:

that belief is wrong, and everyone in the industry half-knows it.

A brand, employer or otherwise, is not what an organisation publishes. It’s the composite impression that forms in someone’s mind, assembled from every signal available to them.

The published page is one signal. The reviews are hundreds of signals. The stories told at dinner tables, which we covered two weeks ago, are hundreds more. Your brand is the weighted average, and you don’t control the weights.

And the weights are brutal. The careers page was written by the organisation, and every reader knows it, so it gets read the way people read advertising: with suspicion, at a discount. The reviews were written by people who lived the experience and have nothing to sell, so they get read the way people read a friend’s warning: with trust, at a premium. Study after study of candidate behaviour lands in the same place: people trust current and former employees’ accounts of a workplace far more than the employer’s own voice.

One promise, discounted. Hundreds of testimonies, at a premium. That’s the arithmetic your beautiful page is up against.

The three kinds of distance When I run this test with organisations, the gap between the tabs usually sorts into three types, and each one is a different diagnosis.

The vocabulary gap. The page speaks a language no one inside the organisation uses. “Empowerment.” “Journey.” “Family.” Meanwhile the reviews speak in concrete nouns:

rosters, managers, parking, pay cycles, workload. When the two tabs don’t even share a vocabulary, it usually means the promise was written far from the work: by an agency, or a head office function, without a single operational voice in the room. The fix here is the cheapest of the three: rewrite the page in the words your people actually use. Their vocabulary is on tab two, free of charge.

The theme gap. The page is silent on the exact subject the reviews shout about. Every review mentions workload; the page doesn’t contain the word. Every second review mentions “depends on your manager”; the page describes a uniform culture that manifestly isn’t. A theme gap is more serious than a vocabulary gap, because silence on a known issue doesn’t read as neutral. To a candidate who has read both tabs, it reads as evasion, and it discounts everything else the page says. The candidate’s logic is simple: if they won’t mention the thing everyone mentions, what else are they smoothing over?

The evidence gap. The page makes claims; the reviews supply counterexamples. “Flexibility that works for your life” versus “flexible until you actually use it” isn’t a difference of tone. It’s a claim meeting a witness. This is the most expensive gap of the three, because it’s not a communications problem at all. It’s the Promise Gap itself, surfacing in public: a commitment being made in one tab and reported broken in the other.

Vocabulary gaps embarrass you. Theme gaps discount you. Evidence gaps accuse you.

Know which one you’re looking at before you decide what to do about it.

But the reviews are biased Let me take the counterargument seriously, because it’s the first thing said in every room where this test gets run, and it’s not wrong.

Yes, reviews skew. People who leave angry write more than people who stay content. A handful of reviews from one bad team can colour the whole picture. Some reviews are unfair, some are outdated, and occasionally one is simply untrue. If you read tab two as an accurate census of your culture, you’ll over-correct toward your loudest leavers.

But notice what this objection does and doesn’t change.

It doesn’t change what candidates do. They read the reviews anyway, weight them anyway, and decide anyway. The bias objection is an argument about epistemology being deployed against a question of behaviour. Whether the reviews are fair is a separate matter from whether they’re operating, and they are operating, on every candidate you want, right now.

And the test never asked you to treat tab two as truth. It asked you to measure the distance between the tabs. That’s the elegant part: even biased reviews produce a useful reading, because you’re not scoring the reviews, you’re scoring the gap. An organisation with an honest page and skewed reviews shows a modest, explainable distance. An organisation describing a workplace its own people don’t recognise shows a chasm, and no amount of “reviews are biased” explains a chasm.

There’s also a newer reason the bias objection has lost its comfort. When a candidate asks an AI tool what it’s like to work at your organisation, and they increasingly do, the answer is synthesised from everything public: your page, and all of tab two. Your organisation gets one voice in that synthesis. Your reviewers get the chorus. The composite you used to be able to ignore is now being assembled automatically, on demand, for anyone who asks. More on that in the final essay of this series, because it changes the game more than most people have registered.

The uncomfortable direction of the fix Here’s where the test earns its keep, because it forces a conclusion most employer brand programs are structured to avoid.

You cannot close the gap in tab one.

The instinct, when the distance is exposed, is to fix the page: soften the boldest claims, add a “we’re on a journey” paragraph, refresh the photography. And a more honest page genuinely helps, we’ll get to that. But rewriting the promise while the experience stands still doesn’t close the gap. It relocates it. The reviews keep arriving, one per departure, each one a fresh data point from the lived side of the ledger.

The gap closes in tab two, and tab two can’t be edited. It can only be out-lived. Every kept promise eventually becomes a review, a story, a data point on the other side. It’s slow, it’s operational, and it belongs to leaders far beyond the employer brand team, which is exactly why the fast fix keeps winning and the gap keeps reopening.

That said, tab one isn’t innocent either, and here’s the version of “fix the page” that actually works: make the page honest about what’s hard. The strongest careers pages I’ve seen include some version of “this work is demanding, here’s what we ask of you and here’s what we give back.” That sentence terrifies brand guardians and does two remarkable things. It converts fewer of the wrong candidates, the ones who would have arrived, met reality, and left a review. And it makes every other claim on the page more believable, because a page that admits something costs is a page that might be telling the truth about the rest. Honesty on tab one doesn’t close the gap, but it stops the page from widening it.

What to do with this on Monday

Run the test yourself, before anyone else does it to you. Two tabs, thirty minutes, notes. Sort what you find into vocabulary, theme and evidence gaps. That one page of notes is a sharper audit than most commissioned decks.

Run it with your leadership team, live. Don’t present findings; make them read both tabs in the room. The silence that follows an evidence gap read aloud by the executive who owns that area is worth a year of persuasion.

Mine tab two quarterly, as strategy input. Not for reputation management, for theme detection. Reviews are free longitudinal listening data from your least defended source.

Treat the recurring themes as a standing agenda item, next to the engagement scores we talked about last week.

Add one honest sentence to your page. Just one, about something the work genuinely costs.

Watch what it does to the quality, not quantity, of applications.

Respond to reviews like promises are being audited, because they are. Not the legalistic “we take feedback seriously” template. A named, human reply that engages with the substance.

Candidates read the replies more carefully than the reviews; the reply is the organisation being observed keeping, or dodging, a promise in real time.

The sentence to keep

Your employer brand isn’t what you publish. It’s what survives comparison.

The comparison is already happening, in split screens and AI summaries, run by every candidate you’re trying to reach. The only question is whether the organisation has the nerve to run it on itself first.

Next week: the moment the promise meets its first real audience. A new starter learns your actual culture in their first fortnight, and they don’t learn it from anything they read.

If you run the test on your own organisation this week, I’d genuinely like to hear what the gap looked like: vocabulary, theme, or evidence. Reply or comment; the patterns across industries are becoming an essay of their own.

Shehzad Bhanji writes The Promise Gap, a weekly perspective on the relationship between organisational promises and lived experiences. Across a 25-year international career spanning marketing, customer experience, employer brand, HR technology and people experience, he has worked across Australia, Asia, Europe, the Middle East and Africa.