Latest / Uphoff on Media Podcast / What Does It Mean When the Company That Grades Everyone Else’s Disruption Gets Disrupted?

Transcript

What Does It Mean When the Company That Grades Everyone Else’s Disruption Gets Disrupted? · Uphoff on Media Podcast

• 0:01
  SPEAKER_00: Hey there, it's Tony.
• 0:02
  SPEAKER_00: Welcome to Uphoff on Media podcast.
• 0:06
  SPEAKER_00: Today's episode takes a unique look at technology disruption.
• 0:10
  SPEAKER_00: Specifically,
• 0:11
  SPEAKER_00: what happens when the firms that grade your risk
• 0:13
  SPEAKER_00: for disruption get disrupted themselves?
• 0:16
  SPEAKER_00: Yeah, talk about irony.
• 0:19
  SPEAKER_00: The technology analyst firms Gardner and Forrester are being disrupted by agentic AI
• 0:24
  SPEAKER_00: and the speed and severity of the disruption creates a case study
• 0:27
  SPEAKER_00: for every B2B leader.
• 0:31
  SPEAKER_00: In addition to listening to the podcast version,
• 0:33
  SPEAKER_00: please make sure to read the post on Uphoff on Media on Substack so you can download
• 0:38
  SPEAKER_00: the business durability chart that you can use to score your business
• 0:41
  SPEAKER_00: and manage appropriately.
• 0:44
  SPEAKER_00: This episode is titled,
• 0:45
  SPEAKER_00: What Does It Mean When the Company
• 0:47
  SPEAKER_00: That Grades Everyone Else's Disruption Gets Disrupted?
• 0:51
  SPEAKER_00: Okay, let's get to work.
• 0:55
  SPEAKER_00: Gardner and Forrester spent decades explaining disruption to everyone else.
• 0:59
  SPEAKER_00: Their own balance sheets are the case study now.
• 1:02
  SPEAKER_00: Gardner stock is down roughly 71% from its November 2024 high.
• 1:08
  SPEAKER_00: Market cap fell from more than $45 billion to $12 billion.
• 1:13
  SPEAKER_00: Forrester is worse off.
• 1:15
  SPEAKER_00: Market cap near $105 million down from roughly 300 million on a full year net loss
• 1:20
  SPEAKER_00: of 119.4 million.
• 1:23
  SPEAKER_00: These are the two firms B2B buyers have paid for decades to tell them
• 1:27
  SPEAKER_00: which companies were about to get disrupted.
• 1:30
  SPEAKER_00: Neither one saw its own name on that list.
• 1:35
  SPEAKER_00: The numbers are blunt about where the pain is.
• 1:38
  SPEAKER_00: Gardner's consulting revenue fell 12.8% in the fourth quarter.
• 1:42
  SPEAKER_00: Forrester's strategy consulting bookings dropped more than 50% in a year and
• 1:47
  SPEAKER_00: the firm is exiting strategy consulting entirely alongside
• 1:50
  SPEAKER_00: an 8% workforce cut announced in February.
• 1:54
  SPEAKER_00: CEO Gene Hall on Gartner's Q4 earnings call put the case plainly, quote,
• 2:01
  SPEAKER_00: executives have responded to this by slowing and deferring everything possible.
• 2:05
  SPEAKER_00: This makes for a much tougher selling environment, end quote.
• 2:09
  SPEAKER_00: While research and consulting revenue fell at both firms,
• 2:12
  SPEAKER_00: Gartner's conferences segment grew 11% for the full year to $645 million.
• 2:19
  SPEAKER_00: Forrester's event business,
• 2:20
  SPEAKER_00: even after cutting event days,
• 2:22
  SPEAKER_00: is what the firm is redesigning around rather than walking away from.
• 2:26
  SPEAKER_00: The written word and the paid opinion are getting commoditized inside
• 2:30
  SPEAKER_00: the very companies built to sell them.
• 2:32
  SPEAKER_00: The room is not.
• 2:35
  SPEAKER_00: Buyers left before it showed up in the balance sheet.
• 2:39
  SPEAKER_00: Only 13% of B2B buyers used an analyst report
• 2:42
  SPEAKER_00: in their most recent purchase decision,
• 2:45
  SPEAKER_00: down 63% since 2022, according to G2's 2026 buyer behavior report.
• 2:53
  SPEAKER_00: Eight in 10 have sourced a software recommendation directly from ChatGPT
• 2:57
  SPEAKER_00: or Google's AI mode in the last two years.
• 3:01
  SPEAKER_00: That's not a preference shift, that's a category in decline.
• 3:05
  SPEAKER_00: The same report found what should worry Gartner and Forrester more than either stock chart.
• 3:10
  SPEAKER_00: Buyers are moving from modeled ROI,
• 3:13
  SPEAKER_00: a number an analyst estimates from the outside,
• 3:16
  SPEAKER_00: to observed ROI, proof pulled from their own usage data.
• 3:20
  SPEAKER_00: The product is starting to make its own business case.
• 3:23
  SPEAKER_00: The messenger is the part getting cut.
• 3:26
  SPEAKER_00: Agentic AI does something more specific here than simply get smarter at answering questions.
• 3:32
  SPEAKER_00: Gardner's own July 2026 research puts $234 billion of enterprise software spend
• 3:39
  SPEAKER_00: at risk from what it calls agentic arbitrage,
• 3:43
  SPEAKER_00: agents completing a task across systems without a human ever opening the app
• 3:48
  SPEAKER_00: that used to charge for access.
• 3:50
  SPEAKER_00: A synthesized report is exactly the kind of task that collapses first under that pressure.
• 3:56
  SPEAKER_00: It exists to turn public and licensed information into a judgment of human reads.
• 4:00
  SPEAKER_00: An agent can now do that synthesis and the time it takes to ask the question at
• 4:04
  SPEAKER_00: a marginal cost close to zero.
• 4:08
  SPEAKER_00: The company that built a model to grade AI disruption never ran it on itself.
• 4:13
  SPEAKER_00: Forrester spent part of 2026 building its own AI disruption model,
• 4:17
  SPEAKER_00: a tool scoring how exposed other companies' business models are
• 4:21
  SPEAKER_00: to AI substitution based on labor intensity,
• 4:24
  SPEAKER_00: data advantages, and switching costs, among other factors.
• 4:28
  SPEAKER_00: As far as I can find, Forrester has never pointed that model at itself.
• 4:33
  SPEAKER_00: It would be the easiest assignment on the team's desk.
• 4:36
  SPEAKER_00: One analyst, one afternoon, the same inputs it asks of everyone else.
• 4:41
  SPEAKER_00: That silence is its own data point.
• 4:43
  SPEAKER_00: Gardner's magic quadrant,
• 4:44
  SPEAKER_00: meanwhile, has carried a credibility tax since long before agentic AI existed.
• 4:50
  SPEAKER_00: Courts have ruled it isn't technically pay to play.
• 4:52
  SPEAKER_00: Gardner won a lawsuit on exactly that claim in 2014,
• 4:56
  SPEAKER_00: but analyst relations practitioners will tell you privately
• 4:59
  SPEAKER_00: that meaningful analyst time requires becoming a paying client,
• 5:03
  SPEAKER_00: which functions the same way, even when it isn't the same thing legally.
• 5:07
  SPEAKER_00: AI didn't create that trust problem.
• 5:09
  SPEAKER_00: It just removed the last reason buyers had to put up with it.
• 5:14
  SPEAKER_00: This is the same argument I made three months ago.
• 5:16
  SPEAKER_00: The numbers just caught up to it.
• 5:19
  SPEAKER_00: In the vertical intelligence company,
• 5:20
  SPEAKER_00: I wrote that the cost structure protecting large research firms like Gartner
• 5:24
  SPEAKER_00: and Forrester had come down and that agentic AI would be brutal
• 5:28
  SPEAKER_00: to commodity information while amplifying the value of what it can't replicate,
• 5:33
  SPEAKER_00: domain depth, proprietary data, and trusted community.
• 5:36
  SPEAKER_00: That was a prediction in June.
• 5:38
  SPEAKER_00: It's on the balance sheet in September.
• 5:41
  SPEAKER_00: Watch what they do, not what they say.
• 5:44
  SPEAKER_00: Both firms are already showing what they believe is worth keeping
• 5:47
  SPEAKER_00: and they're doing it with budget, not press releases.
• 5:50
  SPEAKER_00: Forrester's own AI product line grew unique visitors 55% year-over-year
• 5:57
  SPEAKER_00: and prompt volume 65% funded in part by savings
• 6:00
  SPEAKER_00: from exiting strategy consulting this year.
• 6:03
  SPEAKER_00: That's not a research firm defending its old business.
• 6:06
  SPEAKER_00: That's a research firm using the disruptor's own technology to reinvent itself on
• 6:10
  SPEAKER_00: the bet that owning the interface will be worth more than owning the content
• 6:15
  SPEAKER_00: the interface pulls from.
• 6:18
  SPEAKER_00: Neither company is walking away from the room either.
• 6:21
  SPEAKER_00: Gartner is expanding it.
• 6:22
  SPEAKER_00: Local sea level communities,
• 6:24
  SPEAKER_00: longer inquiry days,
• 6:25
  SPEAKER_00: more destination events on top of a conferences segment that already grew 11%
• 6:30
  SPEAKER_00: while research fell.
• 6:32
  SPEAKER_00: Watching where a company under this much pressure chooses to spend tells you more
• 6:36
  SPEAKER_00: than anything on its earnings call.
• 6:39
  SPEAKER_00: Strategy consulting isn't an isolated casualty.
• 6:42
  SPEAKER_00: It's the leading edge of a category.
• 6:44
  SPEAKER_00: Businesses that sell synthesize judgment as a subscription or a report
• 6:49
  SPEAKER_00: that also includes sell-side equity research,
• 6:53
  SPEAKER_00: credit rating agencies, and most trade press, mine included.
• 6:57
  SPEAKER_00: None of those industries have been tested at Gartner's scale yet.
• 7:00
  SPEAKER_00: All of them are running the same experiment, just on a slower clock.
• 7:05
  SPEAKER_00: I'm not writing this from outside the blast radius.
• 7:08
  SPEAKER_00: Look, I contribute as an analyst for Cloud Wars.
• 7:11
  SPEAKER_00: I run an advisory practice that sells judgment for a fee.
• 7:14
  SPEAKER_00: If the argument here is that commoditized content and standardized opinion are
• 7:18
  SPEAKER_00: the exposed part of an information business, I'm exposed to.
• 7:22
  SPEAKER_00: And it would be a thinner piece if I pretended otherwise.
• 7:26
  SPEAKER_00: The line runs through my own work.
• 7:28
  SPEAKER_00: My opinion on a trend delivered to a broad audience's content and AI commoditizes
• 7:33
  SPEAKER_00: the same way as commoditizing Gartner's research.
• 7:36
  SPEAKER_00: The intelligence layer is what's left.
• 7:39
  SPEAKER_00: Domain expertise,
• 7:40
  SPEAKER_00: decades of pattern recognition,
• 7:42
  SPEAKER_00: context and judgment applied to one operator's actual decision.
• 7:47
  SPEAKER_00: That layer used to reach one client at a time.
• 7:49
  SPEAKER_00: Agentic AI now lets me deliver it like software, personalized and at software margins.
• 7:55
  SPEAKER_00: It's the argument I've made since the post solo scale
• 7:58
  SPEAKER_00: and Gartner's own numbers are drawing the same line
• 8:00
  SPEAKER_00: through their business right now, whether the company says so or not.
• 8:06
  SPEAKER_00: Sort your own information business before the market sorts it for you.
• 8:10
  SPEAKER_00: Every B2B information,
• 8:11
  SPEAKER_00: research,
• 8:12
  SPEAKER_00: or advisory business is really two businesses running under one name,
• 8:17
  SPEAKER_00: a content and opinion layer and a connection and relationship layer.
• 8:21
  SPEAKER_00: The content layer scales cheaply,
• 8:23
  SPEAKER_00: which used to be the entire advantage and is now the entire exposure.
• 8:26
  SPEAKER_00: The connection layer doesn't scale the same way,
• 8:29
  SPEAKER_00: which used to look like a limitation and now looks like the only part of
• 8:32
  SPEAKER_00: the business AI can't touch.
• 8:34
  SPEAKER_00: To make that split concrete, instead of directional, I built the durability test.
• 8:39
  SPEAKER_00: It's six questions,
• 8:41
  SPEAKER_00: each scored from one to five, where five means the hardest for AI to replace.
• 8:48
  SPEAKER_00: Question one, what do you actually deliver?
• 8:51
  SPEAKER_00: A reporter ranking is a one, a live working session is a five.
• 8:56
  SPEAKER_00: Question two, where does the underlying material come from?
• 9:00
  SPEAKER_00: Public information synthesized scores a one.
• 9:03
  SPEAKER_00: Proprietary access nobody else has is a five.
• 9:07
  SPEAKER_00: Question number three, how personalized is the input?
• 9:11
  SPEAKER_00: The same for every buyer, scores a one.
• 9:13
  SPEAKER_00: Built around one buyer's situation, scores a five.
• 9:17
  SPEAKER_00: Question number four, why does the buyer say they're paying you?
• 9:22
  SPEAKER_00: Tell me what's happening, scores a one.
• 9:25
  SPEAKER_00: Be accountable in the room with me, scores a five.
• 9:29
  SPEAKER_00: Question number five, could a well-prompted AI produce 80% of it in a minute?
• 9:34
  SPEAKER_00: Yes, scores a one, not without you, scores a five.
• 9:39
  SPEAKER_00: And question number six, where does the buyer's trust sit?
• 9:42
  SPEAKER_00: In your brand and methodology, scores a one.
• 9:45
  SPEAKER_00: In one person's track record, scores a five.
• 9:49
  SPEAKER_00: Here's where the Uphoff on Media framework can be downloaded and used
• 9:53
  SPEAKER_00: a diagnostic tool in your business by visiting uphoffonmedia at substack.com.
• 10:00
  SPEAKER_00: Add up your six scores and act on the range you land in.
• 10:03
  SPEAKER_00: A score of 22 to 30, protect it.
• 10:06
  SPEAKER_00: Your value sits where the AI raises the bar instead of replacing the seat.
• 10:10
  SPEAKER_00: Use agents to deliver it faster.
• 10:13
  SPEAKER_00: Score 31 to 21, split it.
• 10:15
  SPEAKER_00: You're running two businesses under one name.
• 10:18
  SPEAKER_00: Separate the content lines from the judgment lines this quarter
• 10:21
  SPEAKER_00: and move your investment to the judgment side.
• 10:24
  SPEAKER_00: Score six to 12, it's time to rebuild it.
• 10:27
  SPEAKER_00: You're the content layer of your own business, whatever your brand still says on the door.
• 10:31
  SPEAKER_00: Start with your 10 best accounts and ask what each would pay for that a prompt can't provide.
• 10:39
  SPEAKER_00: Gartner and Forrester spent four decades telling the other companies where they stood.
• 10:43
  SPEAKER_00: Now the market is telling them.
• 10:45
  SPEAKER_00: Run the durability test on your own business this week.
• 10:48
  SPEAKER_00: Gartner and Forrester didn't get to choose when they found out where they stand.
• 10:52
  SPEAKER_00: You still do.
• 10:55
  SPEAKER_00: Hey, well, that's it for this episode.
• 10:57
  SPEAKER_00: If you found this useful, please do one thing.
• 10:59
  SPEAKER_00: Follow the Uphoff on Media podcast on Apple or Spotify and leave a rating.
• 11:04
  SPEAKER_00: It's the single best way to help other operators find this.
• 11:08
  SPEAKER_00: I'm Tony Uphoff, back at it next week.