A senior peer, thinking out loud.
Plainly written notes on brand, Employer Brand, and where they meet. The canonical home of the writing — where it ranks, and how a reader who's never heard of Ponte first arrives. No listicles, no posturing.
Latest
Everyone's AI ambition is nobody's position
The most widely shared CMO goal of 2026, becoming an AI leader, is a category entry ticket and not a brand position. With budgets flat, the AI line should be judged by what it makes the brand different at, because the machine-mediated discovery it is bought for rewards difference, and fluency is now cheap.
Google's brand is worth $433bn. Or $1.48tn.
The widening spread between brand valuation houses is neither proof that valuation is worthless nor proof that the industry is healthy. It shows that each method answers a different question. The houses disagree most about the brands whose value depends on AI expectations, so the range tells you more than any single number, and nobody should quote a brand value without the method beside it.
The line every employer brand leaves out
Universum headlines regional divergence; its own frequencies contain a convergence it never wrote up - competitive salary sits at 4/4/6% and ranks around twentieth in all three regions. The 'pay is not a differentiator' explanation fails on the same data, because differentiation is plainly not the selection filter: learning and development appears in 47% of European EVPs, shared by half the market. The filter that does fit is falsifiability - every common EVP attribute (development, collaboration, meaning, balance) is uncheckable by a new hire, and a salary band is the only one that is not. So the EVP attribute list is not what employers offer; it is what they can say without anyone able to come back on it. That is brand-sold versus brand-lived at the level of what gets written down at all, and the timing sharpens it: Most Attractive Employers 2026 is being scored market by market right now against exactly these documents.
Employer Brand Has No Second Draft
Where the five lenses disagree is the angle. The leader hears 'we kept you' as the good news; the employee is holding a contract that changed shape without a conversation; the practitioner has nothing to say to either, because the entire apparatus — careers page, EVP, interview loop, offer — is addressed to someone at the moment of choosing. The distinctive claim, and the brand-sold-vs-brand-lived tension underneath it: employer brand is an acquisition instrument, so the promise is made once to a stranger and is amended rather than re-made thereafter. The blind-spot pass is what makes it new — the people who stayed are simultaneously the population the promise is now actually being tested on and the only population no standard instrument isolates. They sit inside Gallup's engagement data, inside the Good Work Index and outside the attractiveness rankings, and are separated out by none of them. Distinguished from the 2026-09-07 balance-promise piece deliberately: that argued a promise was bought as showable artifacts while the delivery layer declined; this argues a promise has no mechanism for re-issue at all, and locates the gap in measurement rather than in delivery.
The career promise nobody can leave over
The longest-dated promise an employer can make has just entered the top five reasons people choose one — in the same year employers de-committed from fixed plans and the manager layer that delivers it thinned to its lowest engagement on record. The freeze then removes the only instrument most employers use to detect a broken career promise (regretted attrition), and replaces it with a flattering number that a kept promise and a broken one now both produce. The measure that would settle it — internal movement — is the one almost nobody publishes.
The money comes back faster than the craft
The money is rotating back to brand faster than the capability to spend it well can be rebuilt, because the profession builds capability the slowest way available (in-house, unchanged since 2020) and the fastest-rising line is the most craft-dense one. Employer brand is the sharpest instance, because it is brand work assigned to a function never resourced as a brand capability at all — and unlike an overstated product campaign, an overstated employer promise is made to an audience already holding the counter-evidence.
The AI reads your awards. It can't read your reviews.
The machine now assembling employer reputation is fed the employer-authored promotional surface and starved of the lived one — not by any editorial judgment about truth, but because a review corpus is proprietary property and an awards page is marketing. The brand-sold/brand-lived gap has acquired a machine on the sold side and a robots.txt on the lived side, which makes the gap an infrastructural default rather than a choice anyone made — and makes arbitraging it self-defeating, because the candidate still reads both.
The gap between AI productivity people feel individually and the organizational change they can't see — and what it does to an employer's public promise
Speed as the declared strategy makes the AI-transformation claim publishable long before the transformation is real; employer brand is where that claim becomes specific enough to be falsified, and the people best positioned to falsify it already work there.
Your people trust you. That is not the same as trusting each other.
Employees now trust their employer above every other institution, and they expect it to hold the organization together. That expectation asks for a kind of trust an employer brand does not necessarily have.
AI transformation ROI and the collapsing manager layer
Companies buy AI as a technology decision and sell it as an employer-brand story, but the return is gated by the manager layer — the exact tier that has eroded to a five-year engagement low; the AI-forward brand is being lived by managers too depleted to make AI real, so the failure that reads as a software gap is a management gap wearing a software budget.
The 2026 gap between record brand value and record-low employee engagement
Brand-as-asset hit an all-time high in the same year brand-as-lived (engagement) hit a multi-year low — the brand-sold-vs-brand-lived gap turned measurable in Tier-1 data, and AI now makes the outward surface easier to inflate exactly as the interior gets harder to read honestly; the record value is borrowed against a thinning inside.
Content arrives with the practice. The structure and voice are set here.