Today, The Content Technologist begins its sixth year of publication. Cheers! Thank you for reading!
In-person workshop with MIMA August 22
Organic content performance for word people: How to use content analytics for editorial decision-making
I'm super stoked for this in-person workshop next month. If you've ever read this newsletter and thought, "Damn. I want this woman to talk with me about content measurement," please join us! It'll be practical, hands on, and entertaining.
Presented with the Minnesota Interactive Marketing Association (MIMA) in-person at Br8kthru Consulting in Minneapolis at 8AM
Get tickets: $65 for MIMA members / $95 for non-members
Workshop description
Do you know how well your digital content has really performed this year? Organic digital content performance analytics—for your website, social media channels, email newsletters, and everything else—are widely misunderstood, misinterpreted, and dismissed as less valuable than paid advertising.
But organic content converts at far higher rates and for longer periods than digital paid media. And when audiences connect to authoritative content, they’re more likely to become brand advocates. Too often, brands and agencies leave money on the table by misattributing performance and not giving their organic content the credit and attention it deserves.
In this workshop, we’ll cover how to measure the true business performance of content. We’ll explore what content metrics mean, how to create workable benchmarks, and how to use what you’ve learned from content analytics to inform future planning.

Five years of newslettering: Prescience, plans, and the pursuit of what's next
Soundtrack: "Five Years" by David Bowie and "5 Years" by Björk.
Five years ago today I sent the first Content Technologist newsletter to a list of around 40 former clients and colleagues. I wanted to explore all the new software the content marketing and strategy sectors were offering, while still maintaining a strategic consulting practice.
My initial goals for publishing the newsletter:
- to write about the types of complex, strategic projects I liked to work on, so that I would find consulting clients who thought similarly;
- to flex my writing muscle regularly, assume the stance of a semi-professional writer, and prove I could write about content strategy and not run out of original things to say;
- to explore and evaluate the growing market of software created for those publishing online;
- to create the smart content I wanted to read, in hopes others would do the same.
As a personal exercise, the newsletter has absolutely succeeded. I've averaged writing/publishing 44 issues per year for five years running, reaching a broader audience each year. I don't really have to invest in paid marketing or uncomfortable sales calls to grow my business (although I would grow much faster if I did). I can now write like a motherfucker about nearly any content strategy topic, pretty much on command. I've succeeded at feeding the beast.
And as a professional tool, The Content Technologist has been wildly successful at connecting with really smart people who care about high-quality content more than they care about rapid growth. I've been able to make a decent living from the consulting leads generated from this newsletter, which means I've found a sustainable way to be a writer, kinda sorta.
But my audience is modest, compared with the go-getters. (Quality, not quantity!) I don't spend much time or money promoting the brand beyond organic discovery, and even that's been done poorly. I haven't found a satisfactory business model for the newsletter. I think in ambitious, enterprise content strategies, but don't have capacity or resources to coordinate them for my own business.
It is what it is. And I'm proud of the work I've done.
I. Organizing past content for future consumption
As a website and resource itself, I've fallen into the trap of many companies investing in publishing regular newsletter, blog, or social content: after five years of publishing consistently, I don't really know what I have in aggregate. My website needs more organization and optimization. Readers and friends say, "I know you've written about this topic, but I can't find it on your site."
The theory I've been spouting recently is that good websites should function as reference books. But I only have an inkling of what The Content Technologist, as a reference book, is actually about, and it needs some editorial love. Structure. A user experience zhuzh.
I've also been recently aware of repeating myself, and while repetition is good for newsletters and social, a canonical resource structured for discovery is better for web content.
One of the cardinal rules of website strategy is not to let content get stale and old; I regularly advise clients to update their hits, connect and consolidating regularly to create the best possible resource for people actively seeking information.
For the remainder of this calendar year, I'm revisiting my content and updating, consolidating, and connecting the archive into a far more navigable resource. Instead of rapidly producing, I'm working on synthesizing and organizing, in hopes of landing on a consolidated messaging strategy.
And I'm not writing new 2,000-word essays each week, at least until the end of the year.
What does this mean for The Content Technologist newsletter? Once a month I'll be republishing updated classics and topical collections like this one, which aggregate related concepts into a more curated guide. If you've been reading for a while, it might seem familiar — old essays with tweaks and updates. If you're relatively new, you'll get a tour through the archives. And the links will still be new, most weeks.
II. Revisiting our original mission of exploring content technology
But wait! There's more: a return to software reviews!
For the first two years, each Content Technologist issue contained a short essay and a review of software. I created a rubric, and my spouse illustrated my evaluation heuristics. At my two previous agency roles, I'd been in charge of evaluating aspects of my employers' tech stacks, and I wanted to explore the possibilities of new personalization, artificial intelligence, and automation tech.
Reviewing software immediately proved to be a bigger challenge than I'd expected. Some of the many issues I faced in the first two years:
- Press-averse tech companies didn't always want me to review their software. I had several enterprise-level software reps discourage me from publishing a review after a demo and free trial because they wanted to control the sales loop.
- Even companies who wanted press didn't know who I was in the early days.
- Continuously updating software development cycles meant that reviews were dated rather quickly. I'd complain about the lack of a certain feature in the review and boom! it would be added within a year.
- Software companies went out of business or were acquired.
- Reviews or explorations of software tech were among my least popular content.
- After a while I was tired of looking at the same email automation builders and productivity managers...
- ...and I had more client work than I had time to review new tools.
And then the pandemic arrived, followed by extremely close-to-home social unrest. From 2020 to 2022, I reviewed software... but I didn't care about it nearly as much as I had at the start. Like many of us, I was steeped in anxiety about the world at large, as well as the environment just outside my door. I feared for the future, and software didn't seem that important. I wanted to get my work done, write my newsletter, and zone out.
Eventually, gradually, we emerged from the events of '20-21. The helicopters stopped circling my neighborhood. Things got better, but my passion for software flagged. The software evaluation rubric got stale at the edges. I wanted to focus on big, strategic ideas rather than reviews, so I brought on editors and collaborators. We published our last review in June 2022.
But in the meantime, investment in what was now called the Creator Economy caught on. Then, OpenAI launched ChatGPT. Suddenly everyone cared about software again. As I indicated in last week's tech stack, I'm excited by what I'm seeing across creative software, AI or otherwise.
And, so! I'm aiming for one new and one updated software review each month for the rest of the year. I'm also compiling guides to understanding similar tools, a la Wirecutter and the lists of "best X software" that SaaS content marketers regularly produce... but with actual editorial insight and tips for adoption.
III. You heard it here first: Revisiting old ideas in the context of new conversations
More than one content leader I respect has said The Content Technologist is "ahead of [its] time." After a recent revisitation of issues from the first two years, I'm inclined to believe them.
Lo and behold, some ideas I initially lobbed a few years ago have been since picked up by more viral influencer types and/or mainstream media outlets. For example:
- "In my experience executives never ask for your ideas. They read some non-idea sentence online like 'Humans prefer video to photos to text' and that becomes the foundation for a 'pivot to video' strategy, and commercials are sold against it as they always have been, and we have the same content but made shittier, again!" - The Content Technologist, July 2020.
Cory Doctorow popularized a similar concept of "enshittification" in January 2023. - "But my most recent aggravation: either when researching a vertical for a client, searching for new software, or just looking up recipes or home topics, Google results are a giant pile of sameness. Every organic search result reads like the same marketing brochure, reiterating the same half-facts, the same pile of words with very little original information." - The Content Technologist, September 2020.
- Charlie Warzel published a comparable critique of SERPs in The Atlantic in June 2022, followed by Amanda Chicago Lewis' SEO hit piece in The Verge last winter.
- "Content data — literally, looking at what audiences like, search and share and figuring out why — helps us create better stories." - The Content Technologist, August 2019...
...coupled with the below meme and its accompanying post, May 2021
...plus, almost every measurement post we've ever published echoes the same sentiment: that click-level attribution of content marketing is a fool's errand (but there are plenty of available measurement systems that attribute content to business growth).
Software entrepreneur and SEO thought leader Rand Fishkin posted on LinkedIn that "Clicks are dying. Attribution is dying." literally this week. (Has anyone told him that nigh-unmeasurable but effective programmatic display is a $157 billion industry? No one has ever clicked on programmatic banner ads and yet they work.)

It's not that I think any of these influential voices are bad, that they copied anyone's ideas, or even read them in the first place. I've always conceived that these concepts are in the water or in some kind of Jungian collective consciousness. Curious people with similar educational and professional backgrounds arrive at the same conclusions when they start teasing out and thinking through the systems we're embedded in.
Credit where it's due: My measurement and attribution approaches are hardly my own. I worked though them with extremely savvy media planners, enterprise analysts, sales reps, buyers, and publishers over the past 10 years. Especially in measurement, I owe immense credit to my colleagues, clients, professors, and mentors. I am often writing through a hybrid of others' approaches.
It's also likely I've absorbed many of the ideas from other newsletters, conversations, social networks, and more. I write about and publish them, uncredited, unknowing. That's the nature of internet content.
But the pattern of high-profile writers, reporters, and influencers arriving same ideas as I have several years later fosters confidence in this newsletter project, in its many iterations.
Publishing a newsletter consistently for five years was always The Content Technologist's end goal. Five years = satisfaction, I thought in 2019. And I've done that, with lots of help—from Will, from previous editors and collaborators, from clients and mentors and readers and friends. (In next week's issue: a long-overdue full list of influences.)
The editorial tl;dr of it all is:
- No new essays for the rest of 2024.
- I'm revising and reorganizing most of the archive, publishing revised versions of old essays in the newsletter.
- New software reviews and collections are on the horizon.
- At the end of the year, I'll evaluate whatever progress has been made and... decide on what's next!
Thank you ever so much for reading and caring and sharing what you've enjoyed about this newsletter. Five years is a long time. It's not likely I'll publish five more (I have other writing plans), but I'm not stopping just yet.
And so we beat on, content against the current, borne back ceaselessly into the digital void.
–DC
Content tech links of the week
- You likely heard Google Chrome will no longer deprecate cookies, after waffling for almost the entire lifespan of this newsletter. I read a couple of takes that insisted cookies would die because more people would opt out, but the fact remains that when users opt out of cookie tracking, they see the bottom-of-the-barrel ads. The opt-in experience for cookies isn't great, but the opt-out experience is a chumtastic slop fest.
That's why I'm glad publishers are still exploring higher quality cookieless targeting solutions, per Digiday. My recommended strategy for independent publications: Just run weekly premium ad buys and don't bother with programmatic. You control the user experience! It's easier to manage and faster! Better for everyone! - Ernie Smith's history of collaboration software, written for tech association IEEE Spectrum, is a lovely read unto itself without the video embeds. Raise your hand if you ever worked with Lotus Notes!
But about three quarters of the way down the page, Smith shares a fascinating piece of 1990s corporate pop culture embedded in its entirety: in 1992 Microsoft presented an hour-long Broadway musical about software (?); it had a Blues Brothers leitmotif throughout (?!), and Bill Gates was on stage with a major role (?!?) Obviously, it's cringe internet gold. - The Washington Post has built a climate research chatbot based on its own reporting, via NiemanLab. The best part: if the chatbot doesn't know the answer to a question, it says "I don't know" in most cases.
- Google is now the only search engine that can crawl Reddit and display its results because it has paid Reddit, per 404. Bing and DuckDuckGo and all the other alternative search engines no longer show Reddit results, while Google shows them far too prominently.
- The Air Travel Design guide is my new favorite best-in-class website, per Steve Bryant's excellent Delightful newsletter.
The Content Technologist is a newsletter and consultancy based in Minneapolis, working with clients and collaborators around the world. The entire newsletter is written and edited by Deborah Carver, independent content strategy consultant, speaker, and educator.
Advertise with us | Manage your subscription
Affiliate referrals: Ghost publishing system | Bonsai contract/invoicing | The Sample newsletter exchange referral | Writer AI Writing Assistant
Cultural recommendations / personal social: Spotify | Instagram | Letterboxd | PI.FYI
Did you read? is the assorted content at the very bottom of the email. Cultural recommendations, off-kilter thoughts, and quotes from foundational works of media theory we first read in college—all fair game for this section.
My new favorite time-wasting activity on ChatGPT: Ask the chatbot about famous scenes from movies. Unless the bot has been fully briefed on the movie, it likely has no idea what is in a particular scene — especially if it's a mostly visual scene. (Remember: chatbots can only process language and transcripts, not visuals.) In this interaction, I asked the chatbot to describe my favorite scene in my favorite John Hughes movie:

ChatGPT's response is fully invented and, let's be honest, would have been a terrible addition to the Pretty in Pink script. Duckie is in neither the computer scene nor the house party scene, you erroneous robot!