A mid-stage SaaS company asked me to look at their hiring funnel in March 2026. They were spending $640,000 a year on employer-brand initiatives, including a refreshed careers page, a Glassdoor reputation campaign, a podcast about company culture, and quarterly LinkedIn ads pushing values content. The funnel produced 220 qualified applicants per month for engineering roles. The conversion from “qualified applicant” to “signed offer” was 2.4 percent. Sixty-two percent of the offers were going to candidates who told them in interviews that they had been referred by an existing employee. The employer-brand work was, in measurable terms, not producing offers. The employee network was.
This pattern repeats across companies sized 50 to 5,000. Employer-brand budgets balloon. The activities feel like the right thing to do. The plays come from category-defining decks at companies like Stripe, Shopify, and Notion that pioneered “brand as recruiting moat.” But the actual mechanism that closes senior hires is almost never the polished careers page. It is the referral, the warm intro, the specific blog post a candidate read on a Sunday night that made them decide to look. Most employer-brand strategy invests in the visible activities and starves the actual mechanism. The 8 plays below are the ones that move the conversion rate from qualified applicant to signed offer, based on what senior engineers and operators have told me in exit conversations across roughly 200 hires over the past four years.
Play 1: ship public technical writing from your real engineers, not a marketing pen
The single highest-converting piece of employer-brand content is a deeply technical blog post written by an actual engineer at the company about a real problem they solved, with specifics. Not a “here’s how we think about engineering culture” essay. A “here is the migration we ran from MongoDB to Postgres, here are the four things that broke at scale, and here is the load-test data” post.
Senior engineers read this kind of post and form an immediate hiring opinion about the company. The opinion is binary: this is a place where smart people work and learn from each other, or this is a place where they write LinkedIn-ready essays about engineering values. There is no middle ground. The candidates you actually want to hire are reading the post for the substance. If the substance is real, they apply. If the substance is missing, they leave the tab open for two minutes and move on.
The companies that have made this work, including Stripe, Vercel, Cloudflare, and Sentry, share a structural feature: the technical writing is published under the engineer’s real byline, with no marketing rewrite, on a domain the engineer is proud to point to. Internal review for accuracy is fine. Marketing-team rewrites are poison.
Play 2: make the careers page boring on purpose
The careers page is supposed to be the moment the candidate converts from interested to applied. Most careers pages are designed by the marketing team and look like landing pages for a SaaS product. Hero image of a smiling diverse team, three values stated in a font weight that suggests gravity, a band of “what we believe” tiles, then twelve roles listed at the bottom in a small font that the candidate has to scroll to find.

The careers page that converts is structurally boring. Three sentences at the top about what the company does, what kind of person thrives there, and how interviews are run. A clean list of open roles with the team, location, and seniority visible without a click. A link to the engineering blog. A link to a behind-the-scenes Loom video from the hiring manager. No values band. No mission statement in a serif font. The page exists to get the candidate from “interested” to “applied” in 60 seconds. Anything that does not serve that conversion gets cut.
Play 3: write the hiring-manager Loom for each open role
The hiring-manager Loom is the highest-impact 15 minutes of employer brand work the company can do. The hiring manager records a six-to-eight-minute unscripted video that covers four things: what the team is working on right now, what the person in this role will own in the first six months, what makes this team distinct from the same role at any other company, and what kind of background tends to thrive here.
The Loom does not need production value. A phone propped on a stack of books and a quiet room is enough. The reason it works is that it gives the candidate a real, unmediated read on the hiring manager. Candidates can tell within sixty seconds whether they would want to work for this person. The Loom answers the question before the candidate has to apply, which removes the highest-cost friction in the funnel: an interview process with someone they have decided they do not want to work for. A great Loom doubles your apply-to-interview conversion. A great Loom paired with a real engineer’s blog post on the team’s work triples it.
Play 4: stand up a real referral program that includes process, not just bonus
Most referral programs are a $5,000 bonus, an email blast every quarter, and a Slack channel where employees post job links. That program produces three to five referrals per quarter at a mid-stage company. The program that produces 20 to 40 quality referrals per quarter has different mechanics.
The high-performing referral program runs a 30-minute “referral office hours” every other week, where the hiring team sits with a different department and walks through which roles are open, what the bar is, and what kind of person would thrive. Employees leave with specific names they can think to refer, not just a generic “anyone hiring.” The program also handles the referral hand-off with care: the referrer gets a personal thank-you note from the hiring manager regardless of outcome, and a weekly status update on candidates they referred so they know to nudge their friend. Treating referrals as a routine social transaction is what kills most programs. Treating referrals as a small but real act of care from the company toward the referrer is what makes them compound.
Play 5: write the offer narrative, then deliver it consistently
The offer is the moment the candidate decides to sign or walk. Most companies treat the offer as a number transmission. Send the offer letter, talk through compensation, answer questions, wait. The offer that closes is the offer that arrives with a narrative attached.
The narrative is a 120-word email from the hiring manager that says specifically what the company saw in the candidate, what the first six months will look like, who they will work with most closely, and one specific reason this role is the right one at this moment in their career. The email lands two hours after the verbal offer call. It compounds every signal the candidate has received through the process into a single piece of writing they can re-read, forward to their partner, and use as the artifact that makes the decision feel right.

Companies that started writing offer narratives moved their offer-acceptance rate up by 8 to 14 percentage points within a quarter. The cost is 20 minutes per offer. The return is one or two additional senior hires per year, which more than pays for every other employer-brand activity combined.
Play 6: get four real reviews on Glassdoor and stop chasing volume
Glassdoor reviews are the third-most-trafficked surface in any candidate’s pre-application research, behind the company’s careers page and LinkedIn profile. Most companies treat Glassdoor as a reputation-management problem, paying for premium subscriptions, asking employees to leave positive reviews, and gaming the recency algorithm.
The companies that win the Glassdoor surface treat it as four specific reviews instead of a volume game. The first review is a current senior employee writing 250 honest words about what is hard and what is good. The second is a former employee who left on good terms writing a thoughtful exit perspective. The third is a current employee in a non-engineering role giving a different angle. The fourth is a recent hire describing the onboarding experience. Four real reviews of this shape are more useful to a candidate than 80 generic five-star reviews, because candidates have learned to detect manufactured reviews and discount them entirely. The candidate making a real career decision is looking for honest signal. Give them honest signal and the conversion follows.
Play 7: build the candidate experience as a real product
Most companies treat the interview process as a logistics problem owned by talent acquisition. The companies that hire well treat the candidate experience as a product owned by the hiring manager, with a service-level expectation and a feedback loop. Specifically: the candidate hears back within 48 hours of every step, receives an explicit interview agenda 24 hours before each round, gets named introductions to everyone they will meet, and has a debrief call after the final round whether or not the offer comes.
This sounds operational, not branding. It is both. Candidates talk. The senior operator who runs through your process and gets ghosted for nine days tells four peers in their network. The senior operator who runs through your process and feels handled with care also tells four peers. The math of employer brand at the senior level is largely a function of which version of your process gets discussed in private chats. The “candidate experience as product” investment produces compounding inbound referrals over 18 to 24 months that no amount of LinkedIn-ads spend reproduces.
Play 8: choose the three external surfaces you will be excellent at, drop everything else
Most employer-brand programs try to be present on every surface: LinkedIn, Twitter, Glassdoor, podcasts, careers page, conference sponsorships, university recruiting, engineering blog, employee advocacy, video content, employer-brand awards. The result is competent presence everywhere and excellence nowhere. The companies that hire well pick three surfaces, invest enough on each one to be genuinely good, and drop the rest with no guilt.
For a mid-stage engineering-led company, the three surfaces that almost always win are: the engineering blog, the hiring-manager Looms, and a focused referral program. Each one compounds. Each one is hard for competitors to copy because the asset is the people doing the work, not the budget being spent. For a sales-heavy company, the three surfaces shift toward customer-case-study content, podcast presence on operator-audience shows, and the offer narrative. The principle is the same: pick three, get excellent at them, accept that you will be invisible everywhere else, and trust that excellence on the surfaces that matter outperforms competence on twelve surfaces that do not.
The employer brand built this way looks small from the outside. The careers page is plain. The social presence is muted. The blog publishes monthly, not daily. From the inside, the funnel works: qualified applicants convert to offers at 6 to 9 percent instead of 2.4 percent, senior offers close at 70 percent instead of 45, and the cost per hire drops by 40 to 60 percent across an 18-month window. The companies that figure this out spend less and hire better. The companies that keep adding surfaces never close the gap.