Felix Markman

They argue that technology now favours small operators: large organisations route decisions through many human approvers and move more slowly.

• A solo operator with the right tools can outrun a team of five, and a pair can outrun ten or twenty.
• Markman points to economic data showing that the productivity boom predicted by early generative-AI hype has not fully arrived.

Instinct, a personal-assistant startup with fewer than ten people, reached a reported $2.5 billion valuation before its app was generally available. Markman cites Wellington Management 2026, which found that macroeconomic productivity growth has shown only modest improvement since generative AI became available.

Read original

They predict that a year later, well under 20% of users will be able to avoid logging into Salesforce, and that the pendulum will swing back within two or three years towards features that work only if users log in.

• Their reasoning: demos look finished, but adoption is the real barrier, and the hardest business problems require KPIs, clean data and permissions—not a new product line.
• First-mover advantage is real in markets, they argue, but rarely in unpriced, half-available platform bets.
• Their advice is to let others take the first patch, study the results, then move decisively.

Markman names Claudeforce, Slackforce and Coworker as tools that are early, impressive and sticky—but not yet proven in production. They put the cost of adding an unproven surface to a production CRM at a year of cleanup, a loss of trust with the business and a board question no keynote clip can answer.

A tension runs through Markman's posts: AI is both a tool they use to compile a broad view before judging an improvement and a category they tell buyers not to rush into. Their caution targets unproven platform bets, not AI as a working tool.

Read original · Original 2

They identify six leaks to look for: idle seats, two tools in the same category run by different owners, untracked auto-renew notice windows, brittle integrations held together by someone's weekly CSV cleanup, duplicate or stale CRM and marketing-automation data, and AI copilots with no owner or ROI map. They say fixing these issues requires no new tool or big project—just looking in the right place first.

Markman cites Zylo's 2026 SaaS Management Index, which found that an average of 36% of SaaS licences go unused, and notes that idle seats alone account for 10 to 20% of licences billed with zero activity. On an illustrative $40,000/month stack, they put the exposure at roughly $237,000 a year and point to a Revenue Stack Savings Guide covering six leaks, the savings math line by line, and a 30/60/90 plan.

Read original

They use AI to compile a helicopter view of the landscape first, arguing that it makes little sense to study one tree when you can see the forest. Their posts start with a market-level question and work down to a specific action, rather than starting with a single tool.

In their ROI post, Markman focuses on the gap between early generative-AI hype and measured macroeconomic productivity, then turns to their remit: defining, measuring and producing ROI for revenue operations.

Read original

Their central mission is delivering ROI in revenue operations and questioning software orthodoxy. It lays out the core thesis before turning to specific tactics.

Markman argues that AI gives small teams an advantage over large, slow-moving companies.

• A solo operator with the right tools can outrun a team of five; a dynamic duo can outrun ten or twenty.
• They point to Instinct, a personal assistant startup with fewer than 10 people, which reached a $2.5 billion valuation before general availability.
• They note that macroeconomic productivity growth has been modest despite generative AI hype and ask why.
• Their mission is to right the wrongs of software past and produce measurable ROI for revenue operations.

For brands planning a campaign

For a B2B SaaS company focused on cost reduction and efficiency in revenue operations, this post positions Felix Markman as ROI-focused, skeptical of hype, and oriented toward measurable outcomes. They suit briefs that call for a practical, numbers-driven approach to RevOps.

It shows their skepticism of hype and focus on practical, production-ready solutions, central to their revenue operations philosophy.

After Dreamforce, Markman challenges the claim that AI replaces the user interface.

• They predict that in one year, fewer than 20% of users will be independent of logging into Salesforce, and that within two or three years Salesforce will sell features that require logging in.
• The barrier is not the demo—demos always look finished—but adoption.
• Tools like Claudeforce, Slackforce, and Coworker are early: impressive and sticky, but not proven in production.
• The hardest business problems yield to KPIs, clean data, and permissions, not to an “AIforce wand.”
• First-mover advantage is real in markets but rarely in unpriced, half-GA platform bets.
• The cost of putting an unproven surface on production CRM is a year of cleanup, a trust hit, and an unanswerable board question.
• The advice: let others take the first patch, study the results, then pounce.

For brands planning a campaign

For a brand evaluating Felix Markman for AI adoption or Salesforce optimization, this post demonstrates cautious, evidence-driven thinking.

• They prioritize production readiness, governance, and user trust over keynote promises.
• It suits briefs about responsible AI implementation or avoiding costly platform bets.

It demonstrates their practical approach to revenue operations, showing how to uncover hidden costs in a SaaS stack. It’s a useful starting point for readers focused on savings.

Markman cites Zylo's 2026 SaaS Management Index: 36% of SaaS licenses go unused.

• They list six common leaks: idle seats (10–20% of licenses billed with zero activity), tool overlap, auto-renew cliffs, brittle syncs, bad CRM/MAP data, and shadow AI.
• On a ~$40K/month stack, that's roughly $237K a year in exposure (illustrative).
• They note that most teams look for savings in the wrong places—pushing for 5% off at renewal or arguing over a $30/month tool—while the real money leaks quietly.
• They offer a Revenue Stack Savings Guide with the six leaks, line-by-line savings math, and a 30/60/90 plan.

For brands planning a campaign

For a B2B SaaS company focused on cost reduction, this post shows Felix Markman’s hands-on expertise in identifying and quantifying SaaS waste.

• They can help a brand find savings in its existing stack without new tools or big projects.
• It suits briefs about SaaS stack optimization and RevOps efficiency.

They cite Zylo’s 2026 SaaS Management Index, which puts the average share of unused licenses at 36%, then name six specific leaks a company can look for: idle seats (10–20% of licenses billed with zero activity), two tools in the same category run by different owners, auto-renew notice windows nobody tracked, brittle integrations held together by someone’s weekly CSV cleanup, duplicate and stale CRM or marketing-automation data, and shadow AI copilots with no owner and no ROI map.

• Their argument is that most teams hunt for savings in the wrong places — pushing vendors for 5% off at renewal or arguing over a $30/month tool — while larger sums disappear quietly.
• They run the math on an illustrative ~$40K/month stack, estimating roughly $237K a year in exposure, and say the fix needs no new tool or big project: just look in the right place first.

Opens with a provocative question, ties it to a named third-party index, then breaks the problem into labelled categories, each with a cost, and adds a simple arithmetic example.

For brands planning a campaign

This is the closest match to a campaign aimed at cost reduction.

• Markman’s contribution is the diagnostic checklist — idle seats, category overlap, auto-renew cliffs, brittle syncs, dirty CRM/MAP data and unowned AI tools — and the argument that savings come from stack hygiene, not vendor negotiation.
• A brand could ask them to apply that lens to its own stack in a sponsored piece or frame a renewal-season savings message.

Examples: Source 1

They argue that technology now gives small operators an advantage over large incumbents because big organizations make decisions through many human approvers and are therefore slower.

• They claim a solo operator with the right tools can outrun a team of five, and a pair can outrun ten or twenty, pointing to the personal-assistant startup Instinct reaching a reported $2.5 billion valuation with fewer than ten people before the app was generally available.
• They then note that macroeconomic productivity growth has shown only modest improvement since generative AI arrived, compared with the larger boom early hype predicted, cite Wellington Management 2026, and say the gap bothers them.
• Their stated mission is to define, measure and produce ROI for revenue operations.
• They describe their method as using AI to compile a helicopter view of the landscape before judging any single improvement.

Opens with a personal claim about what Markman delivers, supports it with a company example and a cited research note, then turns the gap between expected and measured results into their own remit.

For brands planning a campaign

For a campaign aimed at driving ROI, this post shows how Markman frames the problem: define and measure returns before adding tools.

• It also signals the kind of argument they are comfortable making on a brand’s behalf — that smaller, faster operating models can outperform larger ones when tooling is used well.
• The Instinct and Wellington references are theirs.

Examples: Source 1

They bet that a year later, well under 20% of users will be independent of logging into Salesforce, and predict the pendulum will swing back, with Salesforce selling features two or three years out that only work if you log in.

• Their reasoning: demos always look finished, the real barrier is adoption, and tools such as Claudeforce, Slackforce and Coworker are impressive and sticky but not the same as proven in production.
• They argue that the hardest business problems yield to KPIs, clean data and permissions you would bet your job on, not to a wave of an AI product line, and that first-mover advantage is real in markets but rarely in unpriced, half-general-availability platform bets.
• The cost of putting an unproven surface on a production CRM, they say, is a year of cleanup, a hit to trust with the business and a board question no keynote clip answers.
• Their advice is to let others take the first patch, study the results and then move decisively.

Structured as an explicit hot take: name the vendor claim, make a falsifiable prediction, list the obstacles in order (adoption, data quality, governance, permissions), then state the recommended buying posture.

Two of the four posts make the same argument in near-identical form, one with emoji and hashtag formatting and one without. They represent one position, not two.

For brands planning a campaign

This is a wait-and-see buying argument, useful to a brand whose audience is being sold new AI surfaces and needs permission to delay.

• Markman can carry a message about adoption, data readiness and governance rather than feature novelty.
• A vendor selling an unproven AI capability should expect this framing to work against it, not for it.

Examples: Source 1 · Source 2

They react to industry moments — a Dreamforce keynote, a SaaS management index — rather than publish evergreen explainers.

• They anchor arguments in named external data (Zylo’s 2026 SaaS Management Index, Wellington Management 2026) and company examples such as Instinct, writing in short, declarative paragraphs with a clear stance.
• They also use a practical-resource format: the SaaS cost post ends by describing a Revenue Stack Savings Guide covering the six leaks, the savings math line by line, and a 30/60/90 plan to capture savings before the next renewal, offered via a link in the comments.

Combines opinion-led commentary with a downloadable or gated working resource, using the post to make the argument and the guide as the handoff.

For brands planning a campaign

A brand gets two usable formats from this creator: a sharp, data-anchored point of view tied to a news moment, and a practical checklist or plan readers can act on.

• The checklist format particularly suits a cost-reduction campaign where the value lies in what readers do, not just what they read.
• The guide sits behind a link in the comments, so a sponsored version would need its own delivery path.

Examples: Source 1 · Source 2 · Source 3

Markman writes about Salesforce as an outside observer; the posts do not state any sponsorship, employment or partnership with the company.

View example

The post presents Zylo as a data source, not a sponsor or client.

View example

His analysis of ServiceNow's acquisition of Sweep argues that the value sits in a metadata graph showing agents the objects, fields, permissions, code and dependencies inside a Salesforce org. That echoes Markman's point that the hardest business problems yield to clean data, KPIs and permissions, not a new interface.

Keenan writes from the vendor and DevOps-architecture side, covering acquisition strategy, delivery architecture and release automation, and sends readers to longer written analyses. Markman writes from the buyer side about where a company's SaaS budget quietly leaks and whether a purchase will pay for itself.

Read original · Original 2

His post after his tenth Dreamforce reports tired teams, messy data and change management as the line item most often cut.

• He names a 'Manual Tax': when an agent is placed on a broken process, it automates the tax instead of removing it.
• That is close to Markman's insistence that adoption, data and governance decide whether an AI feature works.

Coutts frames the problem as operational cost and delay, describing quote-to-cash across an ERP, a CRM the sales team works around, and roughly 40 spreadsheets.

• He writes for operators inside industrial companies.
• Markman's lens is the software stack itself: licences, overlapping tools, auto-renew dates and idle seats.

Read original · Original 2

Those are the same barriers Markman points to when arguing that AI features earn their place in production, not in a keynote. Both read the Dreamforce messaging as a shift in where the product lives rather than a finished capability.

Goel writes from the CIO and IT-leadership seat, covering agent oversight boards, 'agentic AI manager' roles and citizen-developer governance across banks, telecom and technology companies. Markman focuses on revenue operations and SaaS spend, with remedies based on checklists and savings math rather than organisational design.

Read original · Original 2

Their posts analyse ServiceNow's acquisition of Sweep as a “Delivery Digital Twin” — a metadata graph that shows AI agents the objects, fields, automations, permissions, code, and dependencies inside a Salesforce org — and argue that ServiceNow is pursuing “control without data gravity.”

• A separate post reads Salesforce's AIforce launch as “a control story wearing a UI costume,” with the governed context underneath as the real product and a trust harness not expected until 2027.
• For buyers weighing platform consolidation, governing agents from outside the CRM, or where release automation and metadata visibility fit into a SaaS cost problem, that analytical frame can be more useful than a savings list.
• The post includes a five-act timeline and a delivery scorecard for testing a vendor claim.

The tradeoff is scope.

• Vernon Keenan's material is analyst-style and acquisition-centred, and assumes readers already care about enterprise platform strategy.
• It offers no hands-on audit method, per-leak cost model, or renewal-season plan, and does not address unused SaaS licences, tool overlap, or procurement tactics.
• Buyers looking to act on their own stack next quarter will need to translate the governance argument into an internal business case.

Examples: Source 1 · Source 2

He writes about companies that model upfront cost, maintenance, labour, interest, and headcount but never put a number on waiting.

• The pattern shows up in quote turnaround, forecasts nobody trusts enough to plan production against, and deals that quietly stop closing.
• One example: a manufacturer running quote-to-cash across an ERP, a CRM the sales team works around, and about 40 spreadsheets.
• Coutts argues that every quarter of delay brings unmeasured revenue leakage and gives a competitor time to get faster; the bill often comes due during diligence or a value-creation review, when there is no runway left.
• A separate post on Dreamforce says the first AI use case has to be small and measurable, and that placing an agent on top of manual work does not remove the “Manual Tax” — it automates it.
• This is a strong fit when buyers need to argue “fix the process now” rather than “cut a licence.”

The tradeoff is that Mike Coutts offers no tool-level stack inventory, licence-utilisation benchmark, or 30/60/90 savings plan.

• His focus is the cost of not changing the commercial engine, not the mechanics of reducing SaaS spend.
• Buyers looking to find idle seats, overlapping tools, or auto-renew traps won't get a line-item method from his posts.
• His analysis is observational, not a published data study; the arithmetic is described as a recurring pattern, not a published model.

Examples: Source 1 · Source 2

Goel relays lessons from CIOs at large banks, telecom, and technology companies.

• Recurring points include the ungoverned tools and AI-risk ownership questions raised by the age of the citizen developer; some CIOs' proposal for an independent “Agentic AI Oversight Board”; the possibility that a future “Agentic AI Manager” might run a team of 20 agents; lost knowledge and poor documentation that limit what agents can do, prompting incentives to add knowledge to the company wiki; and the hard roadblock posed by integration between systems such as CRM and ERP, which requires sales, enterprise apps, Salesforce developers, and many others to align.
• A separate Dreamforce post argues that AI adoption is uneven, Salesforce is rebranding as an AI company, and the product is no longer the UI, leaving open whether underlying data, integration, agent management, and security problems will be fixed.
• For buyers blocked by cross-system alignment, agent oversight, or knowledge loss, this is closer to the problem than a licence-savings playbook.

The tradeoff is that Shubham Goel focuses on IT leadership in very large organisations.

• The posts are not a SaaS cost study and do not discuss renewals, overlapping subscriptions, or idle seats, or provide a cost model buyers can apply to their own stack.
• The decision they support — how to make AI and integration safe and governable at enterprise scale — may be adjacent to a cost need, but it is a different task.
• Buyers seeking procurement or licence-optimisation advice should ask for that specifically.

Examples: Source 1 · Source 2