Authority Building Through Funnel Content
Why I plotted my own posts
I’d made a version of this chart once before β a simple bar of post-volume by year, titled “Authority Building Through Funnel Content.” It showed the bars going up and I felt good about it.
Now operating in a very different lane β Salesforce Agentforce R&D, partnering with system integrators β I exported my full X archive and decided to redo the chart properly. Not to count posts. To see whether the kind of posting had changed in any meaningful way.
It had. And the shape of the change was the actual story.
The funnel rubric
“Funnel content” can mean a thousand things. For this analysis I used a deliberately narrow rubric β three buckets that map to the journey from learning a thing to selling a thing:
TOFU Β· Learning the ABCs
Public exploration. Curiosity. Observations. Questions to the audience. Broad takes on the topic. Signal: discovery in motion.
MOFU Β· Synthesis & craft
Frameworks. How-tos. Opinions with a stance. Teaching threads (e.g., my #BrokenFunnel series). Signal: authority being built.
BOFU Β· Revenue-oriented motion
Posts that name an offer, describe who I help, share results, or call for partnership. The cash hasn’t always followed β but the intent has. Signal: behaving like a business.
A note on BOFU. Strictly defined, BOFU is “revenue in the bank.” For me, that number is still effectively zero in this lane. So I’m using a softer definition: posts where I’m moving toward revenue β naming offers, building case studies, pointing at a service. The rigor of acting like a business is the BOFU signal here, not the bank balance.
The chart
The story the chart tells
Volume goes up. That part isn’t surprising β I post more now than I did three years ago. The interesting variable is the mix:
Read that as: the work I’m doing on X has become more deliberate. Less “look at this interesting thing I noticed,” more “here’s what I help with, here’s how, here’s who it’s for.” The orange line on the right axis tracks that shift directly. It’s the rigor signal β the part of the chart I now care about most.
The inflection point
There’s a single date where the rigor pivot becomes visible in the data: March 4, 2025. That morning I added “Salesforce Agentforce Builder” to my Upwork profile and posted this:
Before that post I was a generalist who occasionally talked about AI agents and funnels. After it, I was someone with a stated lane. The mix didn’t change overnight β but every quarter since has been more deliberately focused than the one before. Naming the lane is what unlocks BOFU posting. You can’t sell something you haven’t named.
A few earlier moments matter too, in retrospect. November 2022 β the first time I publicly said “I’m offering my services” (it was a CleverTap post; tiny, awkward, but the seed was planted). February 2024 β first “AI agent” mention, the topic enters my lane. April 2024 β I started the #BrokenFunnel series, which became the workhorse of my MOFU output. None of those individually felt like turning points at the time. The chart is what reveals them.
What each bucket actually looks like
The classifier is keyword-based, which means it’s wrong sometimes. Below are the top-engagement sample posts in each bucket per year so you can see β and spot-check β what the algorithm thought was TOFU vs. MOFU vs. BOFU. Open a year to expand.
Apply the rubric to yourself
Take your 30 most recent posts on the platform you care about. Tag each one as TOFU / MOFU / BOFU using the rubric above. Plug the counts in below β see your mix and how it compares to mine.
Your mix
The deeper story is in the newsletter
I write a newsletter where I work through this stuff in depth β the Agentforce deployments, the funnel rigor, the things that don’t go in 280 characters. It’s where the BOFU work actually compounds.
Salesforce SI partner reading this?
If you run a Salesforce SI practice and Agentforce capacity is on your mind, the rigor on this page is the rigor I bring to client work. Funnels-driven, guardrail-first, deployment-ready. Let’s see if there’s a fit.
How I built this (methodology)
Source: Full X archive export (May 2026). 34,893 total tweet records; 7,228 originals from 2022 onward after dropping retweets and replies.
Filtering: A two-stage classifier. Stage 1 drops obvious non-funnel noise (URL-only posts, sub-40-character chatter, posts whose only hashtags are sports/cultural/political). Stage 2 buckets the remaining candidates into TOFU / MOFU / BOFU based on keyword anchors aligned with my professional lane (Agentforce, Salesforce, AI agents, funnels, product strategy, growth, etc.) plus structural cues (length, line breaks, declarative tone). Posts that pass stage 1 but fail stage 2’s domain check are marked “off-topic” and excluded from the chart.
Honest about the limits: Keyword classifiers miss nuance. Some posts in TOFU are arguably MOFU; some BOFU posts are really aspirational MOFU dressed up. The numbers should be read as directional, not exact. The shape of the trend is robust; the absolute counts have a margin of error probably in the Β±15% range. This page is a snapshot of my May 2026 archive; I plan to re-run it quarterly.
Last updated: May 2026 Β· @hov8a Β· nothingelsematterz.com