Why I Ghosted My Own Blog (A Boomerang Hire Returns)

On April 29, I published a post about candidate fraud and then disappeared for almost six months. Somewhere, a vendor is turning that sequence of events into a case study, I’m sure.

The argument, as I remember it, was that job seekers have always embellished their experience, and recruiters who’ve described a dysfunctional workplace as “fast-paced and entrepreneurial” might want to put down the stones before inspecting anyone else’s glass house.

Then I stopped showing up to my own website. Apparently, my commitment to covering the candidate experience extends to reenacting it. So, yes, I ghosted my own blog. Consider this post the content equivalent of a boomerang hire, complete with a suspicious employment gap and the expectation that onboarding will be abbreviated because I already know where everything is, who to avoid and what to expense. That’s probably an imperfect metaphor, but you get the idea.

I do owe you an explanation, even if you likely didn’t notice. Unfortunately, it involves admitting that I spent six months giving people advice I wasn’t taking. Business, on the plus side, has been better than expected. My time management skills? Not so much.

“I’ll Do It Tomorrow” Is Always A Day Away.

The writing never stopped. The publishing just happened everywhere except here.

There’s MediaBistro’s Weekly Drop, which gives me a standing appointment with media consolidation, layoffs and the question of whether the people whose work trained an AI model might eventually get paid for it. There’s Recruiting Unfiltered, a monthly newsletter from Joveo I spend way too much time putting together. Then there’s Kyle & Co., where the work includes research, newsletters, white papers, go-to-market strategy and enough thought leadership to make the term itself feel like a workplace hazard.

Some of that work carries my name. Most of it is under someone else’s byline (which, if I’d link, would break the first rule of ghostwriting). If you’ve read something smart about recruiting in a vendor newsletter this year, there’s a decent chance I helped write it. If you read something stupid, I appreciate your discretion, and obviously, I had nothing to do with it.

I love the work. But, as much as I’d love to believe that you should just follow your passion, I also discovered that a client deadline attached to an invoice beats a personal publishing ambition attached to absolutely nothing but good vibes and social engagement.

There’s academic backing for this, which is convenient when you’d prefer to describe a character flaw as an economic phenomenon, as one does. Or maybe it’s just me. In their 1999 paper, “Doing It Now or Later,” economists Ted O’Donoghue and Matthew Rabin modeled how operating biases in the present encourage procrastination when the effort comes first, and the payoff, if it ever comes, tends to happen much later. Being naive about your future self-control makes the problem worse, and for me, it’s compounded over the intervening months as the work gets more complex and a lot busier.

Every week, I assumed next week’s version of me would have more time. That guy has been missing deadlines since college.

My Absence Is Available as a PDF.

One of the biggest explanations for my silence came out yesterday.

The State of Interview Intelligence, the benchmark study I’ve spent much of the summer and fall working on at Kyle & Co., is finally available. The September draft ran to 64 pages and roughly 21,000 words, which is an unreasonable amount of writing about conversations most hiring managers struggle to document in three sentences.

We surveyed hundreds of employers and job seekers, using mirrored questions to compare how each side experiences hiring. We also spoke with TA and HR leaders. I’m counting the Talent on Tap happy hours and executive roundtables in Seattle and San Francisco as field work, although the methodology gets less reproducible after the second drink.

The research was really intense, and the writing process even more so. But ultimately, I’m proud of the way the finished product turned out (after copious rounds of editing and design revisions) and am looking forward to launching my next project, a category compass on high volume and hourly hiring.

I also learned more than anyone ever needed to know about interview intelligence, and the thing that struck me while prepping the research was something that I’ve never thought about before, and, turns out, neither have many employers: the job interview produces more data than any other part of the hiring process, but almost no one does anything actionable or predictive with that data.

What’s worse is that no one even does anything prescriptive with it, either. For example, most employers we surveyed thought they kicked ass at interviewing, giving themselves an NPS of around +48. The candidates on the other side of the table, though, disagreed, rating their collective interview experience a +3, or slightly better than your average ATS.

One interesting note is that transparency and communication were the biggest drivers of candidate perception; contrary to popular belief (and my own assumptions, going in), whether or not they ultimately received an offer had the least impact on the interview experience.

These are, obviously, very different groups with very different focuses, but that perception gap is something employers should really pay attention to – particularly because the interview process is the biggest determinant not only of who ultimately gets hired, but more importantly, whether or not those finalists choose to accept an offer.

For all the time we spend focused on the top of the talent funnel, we spend relatively little bandwidth or resources on optimizing for the end of the hiring process – one that’s way, way more important and impactful than building pipelines or talent communities.

The fraud findings were particularly fitting given where I left this blog. Talent leaders rated their interviewers 8.1 out of 10 at detecting live AI assistance. Meanwhile, 39% of candidates said they’d been wrongly suspected of cheating or misrepresenting themselves.

Neither response establishes how accurately interviewers detect fraud. Together, they make a pretty good case for checking before buying another product that promises to help everyone become even more confident.

Then there’s the spending. Eighty-five percent of surveyed employers already use some form of interview intelligence tooling, and 97% are at least piloting it. Yet only 14% reached the report’s higher maturity levels, where organizations move into systematic analysis and calibration.

We’ve gotten very good at preserving the conversation. We’re considerably less disciplined about asking whether it helped us hire someone who could do the job. As I said in the perfunctory press release, and will repeat here because frankly, it’s a good line, the market bought transcription and called it intelligence.

A recording can tell you what a hiring manager said. Someone still has to decide whether the question was useful, whether the evaluation followed the evidence, and whether the eventual hire performed as expected. Those responsibilities don’t disappear because the meeting bot has a friendly name.

The report ends with practical recommendations, including several that don’t require another software purchase. I’m aware this is an unconventional approach to HR technology thought leadership, but somebody had to try it.

You can check out the full report here. I promise reading it takes less time than I took to write this blog post. Although probably not the first draft, which clocked in at a cool 21,000 words, or approximately the length of an early Russian novel.

My Out-of-Office Had Better Distribution

Of course, there were also airports, hotels and my least favorite thing on earth, interacting with real people in real life.

In July, I moderated a RecFest UK panel on the shift from talent acquisition to talent ownership, with practitioners from Ipsos, American Express and Safety-Kleen. Knebworth was a terrific setting, although TA operating models remain a harder sell than Oasis. I followed that up by spending the Fourth of July on a yacht in Majorca, hoping my credit card didn’t get declined.

I was also in Nashville for the US version, where I moderated two panels – one on the gap between HR Technology investments and adoption, featuring leaders from Krafton, HCA and Sage, and another, previewing the findings of the interview intelligence study, featuring panelists from The Gap, LiveRamp, Barracuda Networks and Tricentis.

I also joined GoGig as a strategic advisor – if you’re not familiar with them, you should be. I’ve also continued working with great clients like Dalia, Joveo, Findem and HiringBranch on brand marketing, positioning, content strategy and messaging.

The thing is, at some point, if you’re going to keep telling companies what they ought to do, you should probably accept some responsibility for helping them do it. Otherwise, advisory work starts looking suspiciously like heckling with a LinkedIn announcement.

Next week, I’ll be at UNLEASH World in Paris, presenting “Beyond the Demo: Spotting the HR Innovation Worth Betting On.” I should probably tell HR Tech I can’t make it, but I’ve kinda put that off, too.

The premise draws on roughly two decades of watching software demonstrations and trying to distinguish a meaningful improvement in how work gets done from an existing feature with a new marketing budget.

Joe Worsten, regrettably, will not be joining me.

Owning Beats Renting. But It Doesn’t Pay the Bills.

Here’s the part I find a little harder to justify, and feel the need to explain just a little bit. Back in April, I argued that employers needed to invest in their career sites instead of indefinitely renting candidate attention from job boards and aggregators.

Then I spent six months putting my own work on other people’s properties while neglecting the one with my name in the URL. I even co-hosted a roundtable on generative and answer engine optimization, discussing who gets found and cited when machines increasingly mediate discovery. Then I went back to ignoring my own archive.

Apparently, implementation costs extra. Which raises a very important question: why invest in a blog when getting people to visit one keeps getting harder, and the audience for long form content keeps getting smaller? That’s been rattling around in my interior monologue for months, mainly as a self-justification for my own procrastination.

A recent analysis of browsing data found that traditional search results received clicks on 8% of Google visits with an AI summary, compared with 15% without one. Links inside the summaries received clicks on just 1% of visits. That’s a snapshot based on March 2025 browsing, rather than a controlled experiment or a forecast for every publisher. It still gives anyone who writes for a living something to think about beyond keyword density. Also, apparently using external data citations helps with GEO/SEO. The more you know….

In any event, owning a website won’t force Google to send traffic, make you more discoverable on an LLM you or turn a content archive into a out-licensing business and IP portfolio. Anyone promising otherwise probably has a webinar they’d like you to attend, or starts every conversation by asking if you listen to their podcast (nope, unless you’re Keith Morrison or Ira Glass). But it does give your work a permanent address you control, and a destination that transcends current employer or content output. You can update it, organize it, link to it and give readers somewhere to return without asking a feed for permission.

LinkedIn is (somewhat) useful for content distribution. Client work (and presumably, Kyle & Co.) pays the bills. Neither excuses leaving my own brand and content marketing outputs unattended while advising employers to fix theirs. If you know me, you know I hate hypocrisy. And unfortunately, that’s kind of been the story of my life these past few months.

Return of the Mack

I’d love to end by announcing an ambitious editorial calendar, but we’ve already established that my future self shouldn’t be allowed to make commitments without supervision or a clearly defined SOW.

What I can say is that I’m back, there’s plenty to write about, and this site deserves a place on the calendar before everything else has been delivered. Waiting until I have spare time has produced a fairly conclusive six-month experiment.

There’s also a talent acquisition takeaway here, because of course there is. Career sites, candidate communications and employer content are easy to put off, simply because you always have urgent open reqs with a time to fill that’s never fast enough, having to balance the competing priorities of hiring managers, candidates and colleagues, and, of course, talking about AI ad nauseam. Anything that doesn’t have an immediate, demonstrable impact often falls through the cracks – and no LLM can help with intrinsic motivation and personal priorities. That’s kind of on you.

Like anything else, this blog included, unless you take ownership and accountability and actually protect the time it takes to do some of the incremental, but imperative work, “I’ll get to it” becomes less of a feature than a bug, and ultimately, procrastination is an unsustainable operating model. A few months later, you’ll look up and realize you’ve become more or less irrelevant – kind of like me.

I know this because I give that advice for a living. Hell, one of these days, I might even start taking it.

Anyways, It’s good to be back. I’d appreciate it if you’d waive the drug test and reference check.