Latest / Best of LinkedIn: Digital Transformation & Tech / Best of LinkedIn: Digital Products & Services CW 06/ 07

Transcript

Best of LinkedIn: Digital Products & Services CW 06/ 07 · Best of LinkedIn: Digital Transformation & Tech

• 0:00
  Speaker 0: This episode is provided by Thomas Allgaier and Frennus, based on the most relevant LinkedIn posts about digital products and services in calendar weeks six-and seven.
• 0:09
  Speaker 0: Frenness is a B to P market research company helping enterprises gain the market customer and competitive insights needed to drive customer centric and cost efficient product development.
• 0:19
  Speaker 1: Alright ready to get into it because looking at distinct feeling that the party is over.
• 0:29
  Speaker 0: The Party Is Over?
• 0:30
  Speaker 0: That sounds a little bleak for A Start to a Deep Dive.
• 0:33
  Speaker 1: No, not like that.
• 0:34
  Speaker 1: I mean the hype-party you know...that phase where everyone's just yelling AI at every single problem and hoping something sticks.
• 0:41
  Speaker 0: Oh okay i getcha!
• 0:42
  Speaker 1: The sources we're looking at today feel different.
• 0:44
  Speaker 1: they feel sober.
• 0:46
  Speaker 1: We're seeing this major shift from hey look at this cool tech too.
• 0:49
  Speaker 1: how do we actually run a company with this stuff
• 0:51
  Speaker 0: without going broke or burning Everyone out in the process?
• 0:54
  Speaker 1: Exactly
• 0:55
  Speaker 0: Yeah, I see what you mean.
• 0:56
  Speaker 0: We've come through all the noise on LinkedIn and The signal is definitely about well maturity.
• 1:02
  Speaker 0: we've grouped the insights into three big buckets.
• 1:04
  Speaker 0: first the AI product shift which Is all About moving from simple chatbots to actual infrastructure.
• 1:11
  Speaker 0: second Product strategy And specifically how funding in roadmaps are frankly failing us?
• 1:18
  Speaker 0: And finally ways of working because the data were seeing suggests that humans running these systems or just Drowning
• 1:25
  Speaker 1: high functioning.
• 1:26
  Speaker 1: drowning was the term I saw and we'll get to that.
• 1:28
  Speaker 1: But we have to start with that AI shift.
• 1:30
  Speaker 1: Let's
• 1:30
  Speaker 0: do it.
• 1:31
  Speaker 0: for The last year, the playbook was pretty obvious right.
• 1:35
  Speaker 0: take your app slap a chat With AI button on it in boom you're an AI company.
• 1:38
  Speaker 1: tell your investors You're innovating but
• 1:40
  Speaker 0: looking at posts from people like Jana Basto That whole approach is already getting a wall.
• 1:46
  Speaker 1: It is because it's just a feature, not philosophy.
• 1:49
  Speaker 1: Bass does the whole argument that we need to stop treating AI as thing you add into product and start treating this an operating model question?
• 1:57
  Speaker 0: An operating model question, what does that mean exactly?
• 2:00
  Speaker 0: It
• 2:00
  Speaker 1: means the real value isn't some text summary feature.
• 2:04
  Speaker 1: The value is using AI to change the metabolic rate of your entire organization.
• 2:11
  Speaker 1: Yeah
• 2:11
  Speaker 0: Like how do you synthesize feedback?
• 2:13
  Speaker 0: How do make decisions If just bolt a chatbot onto slow bureaucratic process?
• 2:19
  Speaker 0: You haven't solved a thing.
• 2:21
  Speaker 0: You just have a faster way to generate bad ideas.
• 2:24
  Speaker 0: That
• 2:24
  Speaker 1: really lines up with, well... A counterintuitive point from Ali Mortesapour.
• 2:28
  Speaker 1: He wrote this piece that I think made a lot of product managers uncomfortable.
• 2:32
  Speaker 1: Oh
• 2:32
  Speaker 0: i loved that one!
• 2:33
  Speaker 1: he basically said Stop trying to build these complex machine learning models From day
• 2:39
  Speaker 1: One!!
• 2:39
  Speaker 0: I found it so refreshing.
• 2:41
  Speaker 0: In world where everyone's chasing the biggest LLM He is out here advocating for dumb rules
• 2:46
  Speaker 1: Dumb as in like Symbolic AI, the old school.
• 2:49
  Speaker 1: if this then that logic.
• 2:51
  Speaker 0: Exactly and he's right If you're a startup You have a cold start problem?
• 3:15
  Speaker 1: This is a hill I am willing to die on.
• 3:18
  Speaker 1: We look at chat GPT and think, okay that's the interface of the future but it is terrible UX for most specific tasks.
• 3:26
  Speaker 0: users get blank page syndrome.
• 3:28
  Speaker 0: they just stare at It.
• 3:29
  Speaker 1: They don't know what to ask.
• 3:30
  Speaker 1: Internet freedom Is a burden.
• 3:32
  Speaker 1: you have To guide The user
• 3:33
  Speaker 0: And Sid Aurora saw this with the students.
• 3:35
  Speaker 0: He runs an AIPM accelerator and he says users are just ruthless, they do not care if you're using GPT-IV or a ham sandwich to power your back end.
• 3:43
  Speaker 0: They just care about relevance.
• 3:45
  Speaker 0: If they type in that box and don't get good answer immediately... ...they bounce!
• 3:49
  Speaker 1: They're gone.
• 3:50
  Speaker 1: We obsess over the model.
• 3:51
  Speaker 1: The user obsesses over the outcome.
• 3:53
  Speaker 1: But okay let's say You DO GET THE MODEL RUNNING.
• 3:56
  Speaker 1: How Do YOU KNOW IF IT'S WORKING?
• 3:58
  Speaker 1: This is where Harsha's Rovatsu introduced a term.
• 4:00
  Speaker 1: I think it gonna be everywhere soon, hallucination sense.
• 4:04
  Speaker 0: I really like this.
• 4:05
  Speaker 0: we always talk about product sense that gut feeling.
• 4:08
  Speaker 0: now Shravatsa says We need hallucinations and
• 4:10
  Speaker 1: it's deeper than just spotting a lie.
• 4:12
  Speaker 1: anyone can fact-check an answer.
• 4:14
  Speaker 1: hallucination Sense is the intuition to look at a workflow And predict where?
• 4:19
  Speaker 1: It's going to fail.
• 4:20
  Speaker 0: so it's not.
• 4:21
  Speaker 0: Just Is This Answer Wrong but is this whole question likely To produce A wrong answer?
• 4:27
  Speaker 1: Precisely It's a kind of risk assessment.
• 4:29
  Speaker 1: Is this a retrieval failure?
• 4:31
  Speaker 1: Is it a logic failure, you can't just slather air architecture on everything and pray that solves the truthfulness problem.
• 4:37
  Speaker 1: You have to understand The texture of the models limitations
• 4:41
  Speaker 0: and what's interesting is that Teresa Torres who?
• 4:43
  Speaker 0: Is you know, the queen of product discovery?
• 4:45
  Speaker 0: She pointed out that the best way to build this intuition.
• 4:48
  Speaker 0: It was just using it yourself Mm-hmm.
• 4:50
  Speaker 0: she was using Claude code And realized how she has to engineer context for herself breaking down big tasks providing examples.
• 4:57
  Speaker 0: It's the exact same architecture You need to build a project for others.
• 5:00
  Speaker 1: it blurs the line between user and builder.
• 5:02
  Speaker 0: yeah
• 5:03
  Speaker 1: if you can't manage Context For Yourself There's no Way you Can Build A system That Manages it for a Customer.
• 5:08
  Speaker 0: So the AI shift is clear.
• 5:10
  Speaker 0: Move from features to infrastructure, use dumb rules first and develop an intuition for failure but you can have the best AI in the world And if your business model's broken it just doesn't matter.
• 5:21
  Speaker 1: Which brings us to strategy?
• 5:23
  Speaker 1: Honestly this section was a bit of a wake-up call For me.
• 5:26
  Speaker 0: It started with Igor Voth.
• 5:28
  Speaker 0: He had this super provocative statement When I join a company i ignore The product.
• 5:33
  Speaker 1: It sounds like negligence.
• 5:35
  Speaker 1: It's really disciplined.
• 5:36
  Speaker 0: He says he ignores the roadmap, ignores The backlog...he just looks at the P&L.
• 5:41
  Speaker 0: Who are our customers?
• 5:42
  Speaker 0: What are our costs?
• 5:43
  Speaker 0: How do we actually make money?
• 5:45
  Speaker 0: Both
• 5:46
  Speaker 1: calls himself THE BRIDGE between the CEO and the product team And I think hes identifying this massive gap.
• 5:53
  Speaker 1: We have way too many PMs who were just feature factory managers.
• 5:56
  Speaker 0: Yeah they know how to ship code
• 5:57
  Speaker 1: But don't that code translates into margin.
• 6:00
  Speaker 0: If your road map items Don't map back To customer pain That drives revenue.
• 6:05
  Speaker 0: You're just running a very extensive hobby.
• 6:07
  Speaker 1: So why does that happen?
• 6:08
  Speaker 1: Why are so many product teams so disconnected from the business?
• 6:11
  Speaker 1: Fray H. Finnerty thinks it's because of how we fund them.
• 6:14
  Speaker 0: The whole project versus product debate.
• 6:17
  Speaker 1: It is the root of so much dysfunction!
• 6:19
  Speaker 1: Finnerdy points out.
• 6:20
  Speaker 1: companies say their products lead, but they fund their team like projects
• 6:25
  Speaker 0: Meaning you get an annual budget, start date and end date.
• 6:28
  Speaker 0: And successes defined as did you finish on time?
• 6:32
  Speaker 1: Right!
• 6:33
  Speaker 1: And ON TIME tells absolutely nothing about value...and the worst part is what happens to a team In project funding.
• 6:40
  Speaker 1: after launch The team is often dismantled.
• 6:43
  Speaker 0: All that knowledge just evaporates
• 6:45
  Speaker 1: Gone.
• 6:46
  Speaker 1: Real product funding means funding of persistent teams To solve problems over time And you measure them on outcomes, like did retention go up?
• 6:55
  Speaker 1: Not whether they burned down the JIRA backlog by Friday.
• 6:59
  Speaker 0: Speaking of alignment... Multi-Schultz shared this tactic from Lucid that I thought was brilliant.
• 7:03
  Speaker 0: We talk about alignment like it's some magic spell, but he showed how to force it visually...
• 7:07
  Speaker 1: A roadmap review session?
• 7:08
  Speaker 1: Yeah!
• 7:09
  Speaker 0: So picture a board on the left side.
• 7:10
  Speaker 0: you have The Why, The Strategy, The Vision all of business goals.
• 7:14
  Speaker 0: On right we've got Roadmap and Features.
• 7:17
  Speaker 0: You cannot look at the Right Side without the Left Side being visible.
• 7:21
  Speaker 0: It forces your brain to connect the dots.
• 7:24
  Speaker 0: If there is a feature in the right that doesn't clearly link with Goal on the Left just stands out as a sore thumb.
• 7:30
  Speaker 1: They also do this terrifying thing, where leadership leaves sticky notes live on the board during the presentation.
• 7:37
  Speaker 1: That
• 7:37
  Speaker 0: sounds incredibly stressful!
• 7:39
  Speaker 1: It is but it cuts through all the politeness...it solves that.
• 7:42
  Speaker 1: why are we building this question in real time?
• 7:46
  Speaker 1: Because without that kind of rigor companies just stop inventing.
• 7:50
  Speaker 1: they just iterate.
• 7:51
  Speaker 0: That leads perfectly into Ibrahim Bashir's framework.
• 7:54
  Speaker 0: He says companies get stuck in increment mode.
• 7:57
  Speaker 1: Yeah, he has these three buckets Increment evolve and invent.
• 8:00
  Speaker 1: increment is you know polishing what you have Evolve as adapting it Invent Is that real zero to one work?
• 8:07
  Speaker 0: And let me guess ninety percent of the budget It's just gonna increment.
• 8:10
  Speaker 1: Of course because increment feels safe.
• 8:13
  Speaker 1: You can put on a timeline Predict The ROI.
• 8:16
  Speaker 1: Invent is messy.
• 8:19
  Speaker 1: Bashir's warning is that if you treat indent as a luxury,
• 8:49
  Speaker 0: Well, just Tuesday.
• 8:50
  Speaker 1: High performance while drowning.
• 8:52
  Speaker 0: That phrase really hit home.
• 8:54
  Speaker 0: She describes a state where you haven't missed the deadline.
• 8:57
  Speaker 0: You're handling twenty seven slacks an hour.
• 9:00
  Speaker 0: You've got thirty-seven browser tabs open But you are completely empty!
• 9:04
  Speaker 1: You're router for information not product manager.
• 9:07
  Speaker 0: Exactly it's silence before crash.
• 9:09
  Speaker 0: Your so busy communicating about work that can do deep thinking required to solve actual problems.
• 9:15
  Speaker 1: So how did we stop drowning?
• 9:17
  Speaker 1: David Pereira came in with a machete.
• 9:20
  Speaker 1: He published the list of twenty-one things to do without...
• 9:22
  Speaker 0: I love A Good Stop Doing List!
• 9:23
  Speaker 0: What was on The Chopping Block?
• 9:25
  Speaker 1: Oh, Velocity Tracking, Estimation Poker, Complex Jiro Workflows.
• 9:29
  Speaker 1: he calls it all just noise
• 9:31
  Speaker 0: That is going to trigger so many middle managers?
• 9:33
  Speaker 1: oh absolutely.
• 9:34
  Speaker 1: but his point is simple does knowing your velocity actually help you make a better product decision or Does It Just Give Management a False Sense Of Control?
• 9:43
  Speaker 0: He argues you have to cut that fluff, To make room for green items.
• 9:47
  Speaker 0: Deep work problem understanding.
• 9:49
  Speaker 0: Actually talking users
• 9:51
  Speaker 1: The fluff isn't just process though.
• 9:53
  Speaker 1: Sometimes the fluff is doing someone else's job.
• 9:56
  Speaker 0: Jamie Walsh shared some brutal data on what happens when product manager Is trying to be a product marketer.
• 10:02
  Speaker 1: PM vs PMM trap.
• 10:04
  Speaker 1: It s so common.
• 10:06
  Speaker 0: We assume.
• 10:06
  Speaker 0: since the PM knows products zest They should write launch stuff.
• 10:10
  Speaker 0: Walshes Data says Absolutely not.
• 10:13
  Speaker 1: He found that when PMs do PMM work, launches take forty percent longer.
• 10:18
  Speaker 1: Forty percent?
• 10:19
  Speaker 1: That's a massive delay!
• 10:20
  Speaker 0: And the results are worse because PMs write specs.
• 10:23
  Speaker 0: they're right.
• 10:24
  Speaker 0: this feature uses a vector database to optimize latency.
• 10:27
  Speaker 0: A PMM writes... This feature saves you ten hours per week.
• 10:30
  Speaker 1: Buyers buy outcomes and not specs
• 10:31
  Speaker 0: Right.
• 10:32
  Speaker 0: so When the PM rights sales deck The pipeline is just weak.
• 10:35
  Speaker 0: It's hugely costly mismatch.
• 10:37
  Speaker 0: So We need to stop doing marketing's job.
• 10:41
  Speaker 0: Stop obsessing over velocity.
• 10:43
  Speaker 0: Is there any tech that can actually help here or is AI just adding to all the noise?
• 10:48
  Speaker 1: There was a glimmer of hope Akash Gupta and Caitlyn Sullivan Highlighted how AI is reshaping discovery work, and this is a place where it genuinely saves time.
• 10:58
  Speaker 0: This about analyzing user interviews.
• 11:00
  Speaker 0: right exact
• 11:01
  Speaker 1: I mean traditionally if you do five hours interviews You spend ten hours rewatching Tagging transcripts synthesizing everything.
• 11:07
  Speaker 1: It's such a drag mm-hmm.
• 11:09
  Speaker 1: They showed how these new AI tools can pull out value anchors and retention drivers from raw audio in minutes
• 11:16
  Speaker 0: That the dream automate this synthesis so you can spend your energy on the actual strategy.
• 11:20
  Speaker 1: But, there is always a catch.
• 11:22
  Speaker 1: This way of working just raises the bar for talent.
• 11:25
  Speaker 1: Red Hema Karana looked at hiring standards that places like OpenAI.
• 11:30
  Speaker 1: The game has changed.
• 11:31
  Speaker 1: It's
• 11:31
  Speaker 0: not about memorizing frameworks anymore, the
• 11:33
  Speaker 1: days of let me walk you through the circles method are.
• 11:36
  Speaker 1: they're fading.
• 11:37
  Speaker 1: she says.
• 11:38
  Speaker 1: the new interview bar is AI product sense
• 11:40
  Speaker 0: which goes right back to what we talked about with hallucination since
• 11:43
  Speaker 1: it does.
• 11:48
  Speaker 1: These aren't just policy questions anymore.
• 11:52
  Speaker 1: They are core technical constraints.
• 11:54
  Speaker 1: Wow, if you can't reason through the ethical implications of an autonomous agent in an interview You're not getting that job.
• 12:01
  Speaker 1: That's
• 12:01
  Speaker 0: a huge shift We've gone from.
• 12:04
  Speaker 0: Can you prioritize?
• 12:05
  Speaker 0: A backlog mm-hmm to?
• 12:07
  Speaker 0: can you design and ethical framework for a non deterministic system?
• 12:11
  Speaker 0: The
• 12:11
  Speaker 1: industry is growing up.
• 12:12
  Speaker 1: yeah fast.
• 12:13
  Speaker 0: so If we pull all these signals together From symbolic AI to business-first strategy, to AI enabled discovery.
• 12:21
  Speaker 0: What's the headline?
• 12:22
  Speaker 0: for someone listening?
• 12:23
  Speaker 1: The headline is that we are moving from the hype phase... ...to the operations face.
• 12:27
  Speaker 1: Explain it!
• 12:27
  Speaker 1: The shiny toy era is over.
• 12:29
  Speaker 1: The winners of this next cycle aren't going be ones with most AI features.
• 12:33
  Speaker 1: They're those who have integrated AI into their organizational brain Those
• 12:36
  Speaker 0: who've aligned road maps and actual revenue models
• 12:39
  Speaker 1: And structured teams to avoid burning out.
• 12:42
  Speaker 0: It all about integration.
• 12:43
  Speaker 0: Janobasto's operating model, Igor Voths' business focus, Rahima Karana new talent bar.
• 12:49
  Speaker 0: It is all pointing to the same thing.
• 12:51
  Speaker 0: Sustainability.
• 12:52
  Speaker 1: Exactly!
• 12:53
  Speaker 1: Building cool stuff as easy.
• 12:54
  Speaker 1: Building a system that can build cool stuff repeatedly without collapsing... That s hard part and thats where the focus now.
• 13:01
  Speaker 1: Thats
• 13:01
  Speaker 0: a powerful place.
• 13:02
  Speaker 0: to leave it We are done playing with toys.
• 13:04
  Speaker 0: Now we have actually built the factory.
• 13:07
  Speaker 0: If you enjoyed this episode New episodes drop every two weeks.
• 13:10
  Speaker 0: Also check out our other editions on ICT & Tech artificial intelligence, cloud sustainability and green ICT defense tech.
• 13:17
  Speaker 0: And health tech.
• 13:18
  Speaker 0: thanks for listening.
• 13:19
  Speaker 0: in keep diving deep don't forget to subscribe.