Charlotte Ketelaar is the CEO of Capwave, which they describe as building AI tools for venture capital firms. Their LinkedIn posts focus on AI agents, startup go-to-market strategy, and how investors can use AI to organize information and make decisions.
Charlotte Ketelaar is most useful for practical arguments on AI agents, data readiness, and venture investment decisions.
Best for
A B2B SaaS brand selling data readiness or agent deployment tools for a webinar and article series.
Not for
A mass-market consumer app wanting short-form video with guaranteed views, clicks, or installs.
Key takeaways
The agent is the last thing you build, not the first.
They built 50 AI agents in 10 days as a solo CEO.
Compute is now the constraint that prices every AI business.
Check first: Confirm audience demographics, engagement results, prior webinar outcomes, and commercial terms before committing.
Main ideas
The agent is the last thing you build, not the first · Buyers will pay more for one specialized agent than for a stack with AI bolted on · Test every hire against what an agent can already do
Audience
13,205 LinkedIn followers · 8 posts, 10 April–28 May 2026 Reviewed post window
Charlotte Ketelaar is the CEO of Capwave, which they describe as building AI tools for venture capital firms.
Their LinkedIn writing focuses on what AI changes in company operations and investing: how to prepare data before deploying agents, assess AI businesses and costs, and preserve context across investment decisions.
Follow them for practical, opinionated arguments that connect AI trends to specific operating or investment questions.
Ketelaar turns broad claims about AI into tests readers can apply.
In one post, they argue that agent projects often stall because company data is fragmented, then ask how many places an agent would need to search to answer a real business question.
In another, they lay out scenarios for OpenAI valuations and urge investors to test whether a lower valuation would break their thesis.
They connect those arguments to their own work, describing the creation of 50 agents across Capwave's operations and framing the product as institutional memory for venture firms, not another CRM.
Their writing also challenges technical optimism: they argue that the tech industry can overestimate how quickly the wider world will adopt AI.
The posts are forceful and often use specific market figures.
Ask for sources behind cited research and podcast discussions before relying on those figures.
Ketelaar writes as an operator at Capwave, a company they describe as building AI tools for venture capital firms.
Read more
The same argument runs through nearly every post: AI's constraint is rarely the model.
It is the data underneath an agent, the workflow it is meant to own, the hire it replaces, or the compute bill it inherits.
They anchor claims in named third parties — a podcast guest, a bank's figures, an analyst forecast — then reduce them to a decision readers can act on this week: build the agent, wait, hire, rewrite the one-liner, or run the sensitivity table.
Posts typically close with a direct question.
Recurring priorities include lean headcount, specialized agents over feature bundles, data readiness before deployment, and AI that compounds a firm's judgment rather than adding another dashboard.
The agent is the last thing you build, not the first
Ketelaar argues that the gap between buying agents and running them in production is widest in enterprise software, and that the cause is the data agents receive, not the agents themselves.
Read more
Fewer than 20 percent of companies say their data is ready to be reasoned over.
When context is split between an unupdated CRM, a forty-tab spreadsheet, meeting transcripts, and one person's memory, even a strong agent acts on numbers that stopped being true months earlier.
The practical instruction: consolidate data first, then buy the agent.
The post contrasts near-universal adoption — almost four in five companies — with only one in nine running agents in production, and cites Gartner's expectation that 40 percent of agent projects will fail by 2027. It ends with two questions a team can answer before switching anything on: how many places would an agent have to dig through to answer one real question about the business, and is the data ready to be reasoned with or just stored.
Buyers will pay more for one specialized agent than for a stack with AI bolted on
Ketelaar argues that enterprise buyers will not pay a premium for a product that is 60 percent bundled AI features, but will pay substantially more for an agent that owns a single workflow end to end.
Read more
The 60 percent figure serves as a diligence checkpoint, not a debate: does the buyer pay for a feature suite or an outcome, and can the agent complete the workflow without an IT committee meeting? Products that fail those tests are incumbents in a compressing market.
Ketelaar recounts reviewing a Series A deck showing $1.2M in contracted ARR but only $240K collected in the previous 90 days.
They argue the same company would clear a Series A if it led with collected revenue on paid seats instead of contracted ARR on a SaaS stack.
The post closes by asking readers to share their workflows and whether a procurement committee is trying to block them.
Test every hire against what an agent can already do
Ketelaar argues that 2023 hiring plans — a head of sales, two account executives, a customer success lead at $2M ARR — were defensive and that those reflexes no longer hold.
Read more
Before opening a requisition, ask whether an agent can do the job by the end of the next quarter: yes means build the agent, maybe means wait and reassess, and no means hire and write a precise job description. The underlying priority is capital efficiency over headcount, the credential investors reward.
The hiring post cites a claim that AI-native companies are reaching $10M ARR with five or six employees.
It reports one company moving from $200 per seat per month to $15 per successful agent action, with revenue up 3.2x in 11 weeks.
Ketelaar's own build account supports the case: 50 agents in 10 days covering content strategy, deal triage, market research, portfolio monitoring, and investor communications.
Workflow mapping took about 70 percent of the time, agent construction the rest; output accelerated from roughly one agent a day to three or four by agent 30.
Compute is now the constraint that prices every AI business
Ketelaar argues that compute has moved from a footnote to the center of the venture conversation, with consequences for AI-native cost structures rather than the models themselves.
Read more
The figures point to three implications: third-year gross margin assumptions reset because infrastructure does not commoditize like models; an infrastructure depreciation cliff arrives into late 2027; and leverage shifts to whoever controls power, transformers, fab capacity, and memory rather than whoever ships the wrapper. The post calls the capex thesis consensus, not contrarian, and challenges readers to produce a gross margin model that survives 2027.
The post opens with a stack of figures: $725B of 2026 hyperscaler capex against operating cash flow at the four largest platforms, $462B of Google Cloud committed backlog, $625B in broader committed revenue up 110 percent year over year, and Azure backlog at 2.5 years committed. Ketelaar attributes the framing of demand as a vertical wall with compute as the bottleneck to OpenAI's CFO, and the cash-flow comparison to Goldman and Mark Wilson.
Run the scenarios or you have a prayer, not a portfolio
Writing for investors, Ketelaar argues that every venture-backed AI company is now benchmarked against a company that loses money at scale and has not settled its own future, so marks depend on a path nobody controls.
Read more
They recommend building a sensitivity table on the largest AI position and running all three outcomes, with the middle case as the deciding test. If that scenario breaks the thesis, the position is not a portfolio.
The three scenarios: an OpenAI listing above $800B, where markups hold; a listing at $400–500B, producing a 40–50 percent markdown on comparable positions; and staying private while burning another $14B, with no price discovery. Ketelaar frames the exercise as arithmetic, not a bearish call.
Ketelaar's product argument starts with a comparison: the best chiefs of staff are not assistants but institutional memory, holding why a team passed eight months ago, which founder a partner liked after one dinner, where conviction started rising, and the small detail everyone else forgot.
Read more
Most venture firms, the argument goes, are drowning in meetings, transcripts, inboxes, and founder updates, and almost none of it compounds. Capwave aims to be an intelligence layer that remembers how a firm thinks — prioritizing deals, preparing meetings, tracking conviction, flagging inconsistencies before diligence — rather than another CRM.
A companion post uses a trip to Costa Rica to argue that most firms need leverage rather than a robot philosopher, and that the tech world may overestimate how fast everyone else adopts AI. The post also lists what Capwave is meant to do for investors, including turning fragmented conversations into institutional memory and spotting inconsistencies before diligence.
What does this person build, and how do they work?
Read more
This post offers the clearest account of the hands-on side of their practice.
Ketelaar describes building 50 AI agents alone in 10 days for content strategy, deal flow triage, market research, portfolio monitoring and investor communications, without an engineering team or procurement process.
They list three lessons: workflow design, not the agent, took roughly 70 percent of the time; output accelerated from about one agent a day to three or four by agent 30; and the CEO's full view of operations made the agents better than an outside contractor's would have been.
The post ends by asking readers which manual task an agent could take over this week.
For brands planning a campaign
A useful indication that Ketelaar can draw on hands-on building experience rather than abstract commentary. A webinar or article on agent adoption could start with workflow mapping, which the post says consumed most of the effort.
This post gives the sharpest version of Ketelaar's recurring argument.
The post contrasts adoption with production use: nearly four in five companies have adopted AI agents, but only one in nine run them in production.
Ketelaar attributes the gap to data, citing that fewer than 20 percent of companies say their data is ready for an agent and Gartner's expectation that 40 percent of agent projects will fail by 2027.
They list where context usually sits—inboxes, an unupdated CRM, a spreadsheet with forty tabs, meeting transcripts and one person's head—and end with two questions readers can apply to their own systems.
For brands planning a campaign
A strong anchor for a practical session on data readiness before agent deployment. It offers a named, checkable statistic and a simple diagnostic question for a co-branded article or webinar.
How does this person think about the economics beneath AI companies?
Read more
This post shows the investment-analysis side of their writing.
Ketelaar argues that compute has moved from a footnote to the center of the venture conversation, citing $725B of 2026 hyperscaler capex, a $462B Google Cloud committed backlog, $625B in broader committed revenue, up 110 percent year over year, and Azure backlog at 2.5 years.
They draw three consequences for AI-native cost structures: gross margin assumptions reset, the depreciation cliff arrives into late 2027, and leverage shifts to power, transformers, fab capacity and HBM rather than model wrappers.
The post closes by asking for a gross margin model that survives 2027.
For brands planning a campaign
A useful analytical counterpoint to the agent-adoption posts. A content series could pair this cost-basis argument with the data-readiness argument to cover both the economics and the operating reality of AI deployment.
What does this person look for when evaluating a company?
Read more
This post shows how Ketelaar turns a market view into a specific test.
The post argues enterprise buyers will not pay a premium for a product that is 60 percent bundled AI features, but will pay more for a specialized agent that owns one workflow end to end.
Ketelaar recounts a Series A deck with $1.2M in contracted ARR where only $240K had been collected in 90 days, contrasting the pitch that died with the one-liner that would have worked.
They offer two tests for a company's positioning: feature suite versus outcome, and whether the agent can complete the workflow without an IT committee.
The post calls the 60 percent rule a diligence checkpoint rather than a debate.
For brands planning a campaign
A concrete, memorable framework a partner could reuse in a joint asset. The collected-versus-contracted gap could anchor a co-authored piece for founders preparing to raise.
How does this person expect teams to change as agents take over work?
Read more
This post sets out their operating argument clearly.
Ketelaar argues that five people in the right room outperform a hundred in the wrong one, and that companies reaching $10M ARR with five or six employees show the pattern.
They frame 2023 hiring plans as defensive reflexes—a head of sales, two account executives and a customer success lead at $2M ARR—and cite two podcast discussions on outcome-based pricing and a shift from $200 per seat per month to $15 per successful agent action, with revenue up 3.2x in 11 weeks.
The post ends with a three-way test before opening a requisition: build the agent, wait a quarter, or hire—and be exact about the job description.
For brands planning a campaign
A natural topic for a founder-facing webinar or article on lean team design. The requisition test could structure audience questions or a downloadable checklist.
How does this person talk specifically to investors?
Read more
This post shows their risk-framing style for a venture audience.
Ketelaar sets out three scenarios for AI portfolio valuations depending on OpenAI's path: a listing above $800B, a listing at $400–500B with a 40–50 percent markdown on comparable positions, or staying private while burning another $14B with no price discovery. They frame the exercise as arithmetic rather than a bearish view and ask readers to run the same scenarios on their largest AI position, warning that if the middle scenario breaks the thesis, the portfolio is a prayer rather than a portfolio.
For brands planning a campaign
A risk-analysis register for investors rather than operators, useful for a campaign targeting a venture audience alongside the founder-focused posts.
What product is this person building, and what problem does it address?
Read more
The post explains Capwave's positioning in their own words.
Ketelaar argues that the best chiefs of staff are not assistants but institutional memory.
They retain why the team said no eight months ago, which founder a partner liked after one dinner, what changed between meetings, where conviction increased, and the small detail everyone else forgot.
The post says most venture firms are drowning in meetings, transcripts, inboxes, founder updates, and fragmented context, but almost none of it compounds.
It positions Capwave as an intelligence layer that remembers how a firm thinks rather than another CRM.
For brands planning a campaign
The clearest product-positioning post in the material. It offers a problem statement and metaphor for a partnership covering AI for investors.
How does this person frame the market beyond early adopters?
Read more
The post gives the clearest statement of their views on how fast AI changes ordinary businesses.
Ketelaar contrasts a Costa Rica dinner debating whether AI will wipe out humanity with a small town where almost nobody has heard the word agents.
They argue the future of AI adoption may look different from what people in tech expect.
The post says most firms need leverage rather than a robot philosopher and lists what Capwave helps investors do: prioritize deals, prepare for meetings quickly, track conviction across conversations, spot inconsistencies before diligence, and turn fragmented conversations into institutional memory.
It ends with three questions for readers about adoption speed, useful tools, and whether investors have too many tools or too little intelligence.
For brands planning a campaign
Useful if a campaign wants to reach beyond AI-native buyers. The adoption-gap framing can support content aimed at mainstream or traditional firms, not just early adopters.
Approach: named third-party evidence, then one tactical move7
Venture decision-making as a memory problem3
Agent-first staffing and capital efficiency2
Specialized agents versus bundled AI features1
Data readiness as the real bottleneck for agent deployments1
Compute economics and the AI cost basis1
Agent-first staffing and capital efficiency
Ketelaar argues that hiring plans should be tested against what an agent can already do, citing a claim that AI-native companies are reaching $10M ARR with five or six employees.
Read more
The post recommends asking one question before opening any new role: can an agent do this job by the end of Q3?
Yes means build the agent, maybe means wait a quarter, and no means hire a human and define the job precisely.
A separate post describes building 50 agents alone in 10 days across content strategy, deal triage, market research, portfolio monitoring, and investor communications.
Workflow mapping took roughly 70 percent of the time; agent construction took 30 percent.
A three-way decision test (build the agent, wait a quarter, or hire) paired with mapping the workflow before building the agent.
For brands planning a campaign
A clean hook for a partnership on lean operating models or AI-assisted functions. The build/wait/hire test and workflow-first build sequence translate directly into a worksheet, teardown, or office-hours session where a brand walks a team through its open roles.
Ketelaar argues that enterprise buyers will not pay premium prices for a product that is 60 percent bundled AI features bolted onto a SaaS stack, but will pay substantially more for a specialized agent that owns one workflow end to end.
Read more
The 60 percent figure serves as a diligence checkpoint, not a debate: is the buyer paying for a feature suite or an outcome, and can the agent complete the workflow without an IT committee meeting? The post illustrates the pricing gap with a Series A deck they reviewed, which showed $1.2M in contracted ARR but only $240K collected in 90 days.
The deal figures come from one deck Ketelaar says they reviewed, and serve as an illustration rather than market data.
For brands planning a campaign
Suited to a go-to-market or positioning series. The “rewrite your one-liner this week” prompt is a ready-made interactive exercise, and the contrast between collected revenue and contracted ARR gives a B2B SaaS audience a concrete diligence question to apply to its own pitch.
Data readiness as the real bottleneck for agent deployments
Ketelaar frames the gap between buying agents and getting them to work as the widest in enterprise software.
Read more
They cite figures showing that nearly four in five companies have adopted AI agents while only one in nine run them in production, fewer than 20 percent say their data is ready for an agent to use, and Gartner expects 40 percent of agent projects to fail by 2027. The argument: building the agent should come last, because agents working with fragmented, half-updated data will keep acting on numbers that stopped being true months earlier.
Two pre-launch questions a buyer can answer before switching anything on.
For brands planning a campaign
A natural counterweight to agent-hype content and a strong basis for a partnership around implementation readiness, data hygiene, or pre-deployment audits. The closing prompts—how many places would an agent have to dig through, and is your data ready to be reasoned with?—work as a live diagnostic in a webinar or assessment tool.
Ketelaar argues that compute has moved from a footnote to venture's central constraint.
Read more
The post stacks evidence: $725B of 2026 hyperscaler capex against operating cash flow at the four largest platforms; an OpenAI CFO describing demand as a vertical wall, with compute as the bottleneck; a $462B Google Cloud committed backlog and $625B broader committed revenue, up 110 percent year over year; and Azure backlog at 2.5 years committed. Three implications follow: year-three gross margin assumptions for AI-native companies reset, an infrastructure depreciation cliff lands in late 2027, and leverage sits with whoever controls power, transformers, fab capacity, and memory rather than the model wrapper.
A receipts-then-implications structure that closes with a falsifiable challenge to the reader.
For brands planning a campaign
Fits an infrastructure, cloud, or investor-audience campaign. The sensitivity-table habit in the closing challenge translates well to a co-created cost model or a structured debate about which AI business models survive the depreciation schedule.
Ketelaar describes the best chiefs of staff as institutional memory rather than assistants: people who remember why the team passed eight months ago, which founder a partner liked after one dinner, and where conviction started increasing.
Read more
That framing carries into the product: AI inside a venture firm should look like an AI chief of staff, not automation.
Capwave is described as an intelligence layer that prioritizes the right deals, prepares for meetings, tracks conviction across conversations, flags inconsistencies before diligence, and turns fragmented conversations into institutional memory.
A related post applies the same analytical habit to AI portfolio marks, running three OpenAI valuation scenarios and warning that a portfolio whose thesis breaks in the middle scenario is a prayer, not a portfolio.
Scenario tables with a named middle case as the deciding test, plus a direct challenge to run a firm's largest AI position through all three scenarios.
The OpenAI valuation scenarios are presented as a thought exercise, not a forecast.
For brands planning a campaign
Useful for a venture-tools or fund-operations partnership, with institutional memory as the differentiator. The three-scenario sensitivity exercise gives a fund audience a repeatable, concrete takeaway that a brand can sponsor or co-host.
Approach: named third-party evidence, then one tactical move
Across the posts, the recurring structure is to anchor a claim in a named source or figure—Aaron Harris speaking on Logan Bartlett's show, an Invest Like The Best guest, 20VC, Goldman, Gartner, or OpenAI's CFO—then convert it into one concrete action, such as the build/wait/hire test, rewriting a one-liner, building a sensitivity table, or cleaning data first, and close with a direct question to the reader.
Read more
Posts run long and dense, with numbered or arrowed lists, bolded claims, and a short personal aside or product mention in the closing lines.
Personal asides and Capwave mentions appear near the end of several posts, so a partner should expect a product reference in the closing lines rather than strictly neutral analysis.
For brands planning a campaign
A co-produced piece can follow the same shape: cited premise, one action, direct call for responses. That suits interactive formats such as a live teardown, worksheet, or office-hours webinar better than a flat commissioned article.
Charlotte Ketelaar's posts revolve around Capwave, the venture-capital software company they lead.
Read more
They describe it as building an "intelligence layer" to help investors prioritise deals, prepare for meetings, track conviction, and preserve institutional memory, and say they built 50 internal AI agents there as CEO.
The posts make this the only commercial relationship explicit.
Other companies named — OpenAI, Google Cloud, Azure, Goldman Sachs — appear as examples or third-party data points, while the venture podcasts they quote (Invest Like The Best, 20VC, the Logan Bartlett show) are cited material rather than disclosed partnerships.
For any sponsorship, affiliate, or client relationship, ask Ketelaar directly.
Capwave
Employment
Ketelaar writes that Capwave is what they and the team are building for venture capital firms: an intelligence layer that helps investors prioritise deals, prepare for meetings, track conviction across conversations, and turn fragmented discussions into institutional memory.
Read more
In a post about building 50 AI agents in 10 days, they identify as the CEO behind the build and note that the company still has staff. The posts establish their role but say nothing about a sponsorship offer or product pricing.
Charlotte Ketelaar writes long, evidence-first posts about where AI agents land in a business: which workflows they can own, which hires they replace, and what they cost to run.
Read more
New readers should start with the post arguing that an agent is the last thing a company should build; it lists the concrete data failures that stall deployments.
Next, read the build/wait/hire test for staffing and the compute post for the cost side.
Across the posts, Ketelaar names the source of a claim, reduces it to one decision a reader can make this week, and closes with a direct question.
The recurring subjects are AI agents and go-to-market, enterprise positioning, data readiness, compute economics, hiring, venture valuations, and how AI changes the way a firm remembers its own decisions.
Spence writes from the same operating seat Ketelaar uses for their go-to-market arguments.
Read more
Both reduce a scattered sales or AI question to an ordered sequence and push back on adding headcount before the underlying system works. Spence's warning about stacking more reps on a broken sales model parallels their test of every hire against what an agent can already do.
Spence's sequence is sales infrastructure: go-to-market strategy, strategic account-management standards, forecasting governance, and CRM architecture. His posts do not argue from AI agents, deployment data, or venture economics—the territory where Ketelaar starts.
Both write about how AI changes the way a service business is packaged and sold, then turn that into a concrete offer rather than a trend piece.
Read more
Harrington's tiered SEO lifecycle matrix, sold by stage instead of by monthly hours, makes the same move as Ketelaar's argument that a buyer will pay more for one specialized agent than for a stack with AI bolted on.
Harrington focuses on search, web infrastructure, and hiring economics in Latin American marketing, and his posts close with promotion of his own agency. Ketelaar writes about venture capital, agent deployment, and AI cost structure.
Closest match on approach: both write dense, technically specific LinkedIn posts that treat agentic AI as an organizational change rather than a product category, and each closes with a direct question to the reader.
Read more
Timcke's point that controls act jointly, like a circuit, echoes Ketelaar's argument that an agent is only as capable as the data it can reach.
Timcke writes about audit, risk, and governance: control design, segregation of duties, monitoring agents, and the collapse of legacy IT and finance structures. Ketelaar writes about venture and go-to-market: deal pricing, hiring decisions, portfolio marks, and firm memory.
Charlotte Ketelaar’s LinkedIn posts are aimed at founders, venture investors and operators making decisions about AI adoption, startup growth and investment-firm workflows.
Read more
Their topics make them a plausible editorial fit for a B2B series on AI agents and go-to-market strategy, but their posts do not say who follows or engages with them.
Charlotte Ketelaar’s posts don’t publish a sponsorship policy, rate card, media kit or disclosed paid brand partnership.
Read more
Their commercial content promotes Capwave, the AI tooling company they lead for venture capital firms, and invites readers to engage with it; it describes no third-party collaboration or terms.
A B2B SaaS company considering a collaborative webinar or article series on AI agents and go-to-market strategy would need to negotiate formats, deliverables and compensation directly.
The practical first step is to pitch Ketelaar on LinkedIn, outlining the proposed series, its audience, what the company would supply and what it would expect in return.
Ask directly whether they’re open to sponsorships.
Strong fit on subject and format; commercial terms remain unverified.
Read more
Brief
A B2B SaaS company selling an enterprise data-readiness or agent-deployment platform wants a co-hosted webinar and a two-part article series for founders and operations leaders at software companies on getting AI agents from purchase into production.
Conditions
The posts show no sponsorship policy, rate card or media kit. Negotiate deliverables, usage rights and compensation directly, starting with a LinkedIn message that lays out the proposed series, audience, company contribution and expected return.
Ask Charlotte Ketelaar for a recent results pack covering engagement, impressions and past webinar and article-series results before committing budget.
The posts document no audience roles, seniority, company size or geography. The profile's 13,205 followers are the only measured reach figure; treat that as an upper bound.
The posts are text only, with no evidence of on-camera or live delivery. Treat the first webinar as a format test.
Several posts close with content about Capwave, the venture tooling company Ketelaar leads. Agree in advance where a product mention will sit and whether the partner accepts it.
Publication dates vary across the available material. Confirm current posting cadence before scheduling a multi-part series.
Reasoning
The campaign's subject is the argument Ketelaar returns to most often in the posts: agents fail because of the data underneath them, not the model. Nearly four in five companies have adopted AI agents, while only one in nine run them in production; fewer than 20 percent say their data is ready for an agent, and Gartner is cited for 40 percent of agent projects failing by 2027. A data-readiness vendor fits that argument rather than competing with it.
The 60 percent rule offers a second angle. Ketelaar's published test asks whether the buyer is paying for a feature suite or an outcome, and whether the workflow can run without an IT committee. That's the same question the partner's customers are asking. A joint asset can apply the test across the partner's own category.
The hands-on evidence is specific. Ketelaar reports building 50 agents alone in 10 days and says workflow mapping took about 70 percent of the time, agent construction 30 percent. That supports a session on how to sequence the work, not just why it matters.
The format matches Ketelaar's writing: posts anchor a claim in a named source or figure, turn it into one action, then close with a question. That structure translates directly into a webinar with a live diagnostic and a downloadable worksheet.
Ketelaar's posts consistently address founders, investors and operators on AI agents, go-to-market and fundraising—the same ground as the proposed series. They support topic alignment, but not a measured profile of the audience.
When to choose a specialist
Poor fit. Decline or reframe the brief for a B2B buyer.
Read more
Brief
A mass-market consumer app wants a four-week short-form video push: twelve clips for TikTok and Reels, with entertainment-led creative and a contracted minimum number of views, clicks and installs.
If reframed
The brief would move closer to the good-fit case above if the company shifted to B2B buyers, dropped guaranteed performance and accepted a practitioner-led webinar or article series instead of short-form video.
Reasoning
The posts show no video, on-camera work or transcripts. There is no evidence Ketelaar produces entertainment-led short-form content.
The reviewed style is the opposite of short-form entertainment: long, dense analytical posts built on named third-party evidence and figures, with numbered lists and a closing question. The available work offers no evident basis for a 30-second lifestyle clip.
The posts provide no performance numbers: no impressions, reactions, comments, clicks, shares or past campaign results. The available evidence cannot support a contracted view or install minimum; follower count is an upper bound on reach, not proof of delivery.
The posts indicate topic alignment, not audience demographics. They show no age, geography, income or consumer behaviour to support a consumer install target.
The posts show no sponsorship policy or rate card. There is no basis for per-view or performance-based pricing.
A brief requiring Ketelaar to argue that bolting AI features onto an existing product is sufficient would also contradict the published 60 percent rule. Don't commission an asset that readers can see is off-brand.
A worked campaign example
Angle
“Before you buy the agent, fix the ground.” A practitioner session for founders and operations leaders, built on the claim that agents fail because of the data underneath them.
Assets
A co-branded, one-page pre-deployment audit with two diagnostic questions, a scoring column and the 70/30 sequencing rule. It must work as a standalone download, not only inside the webinar.
Two articles drawn from the recording: the first on the gap and the diagnostic, the second on sequencing and the hiring test. Each follows the recurring structure of the reviewed work: named evidence, one tactical move and a closing question to the reader.
Each figure carries the named source and date shown in the posts; none is invented or rounded up.
Brief
Co-hosted webinar and two-part article series on getting AI agents into production, from the enterprise data-readiness platform.
Measurement
Request the results pack first: impressions and engagement on relevant posts from the past six months, past webinar registration and attendance figures, and an audience survey, if one exists.
Define success by registrations, completed audit worksheets returned and follow-up meetings booked—not views.
Treat the first session as a format test: the posts show no live or on-camera record.
Session plan
Framing, 5 minutes: the adoption-versus-production gap, with each figure and its named source shown on screen. Present them as third-party figures Ketelaar cites, not as new research.
Live diagnostic, 10 minutes: ask the audience the two closing questions from the data-readiness post—how many places would an agent have to dig through to answer one real question about the business, and is the data ready to be reasoned with or just stored? Collect answers in a poll; use it as the partner's lead signal.
Build walkthrough, 15 minutes: work through the sequencing from the 50-agent build using one concrete function, such as deal triage or portfolio monitoring. Map the workflow first (reported as about 70 percent of the effort), then build the agent (about 30 percent).
Positioning test, 10 minutes: apply the 60 percent rule to two real offers—one from Ketelaar's material (the specialized procurement agent example), and one volunteered by the partner. Ask whether the buyer is paying for a feature suite or an outcome, and whether the workflow runs without an IT committee.
Close, 5 minutes: apply the build, wait or hire test to one open role each attendee is about to fill.
Terms
No package is available to select in advance. Negotiate deliverables, usage rights and compensation directly; agree in advance where Capwave mentions appear and whether the partner's name sits in the articles' closing lines.
Choose Spence when the campaign goal is to fix sales execution and revenue operations, not get AI agents into production.
Read more
Two posts argue that hiring more salespeople on top of a broken model only adds payroll, meetings and noise.
They lay out the alternative: GTM strategy before job descriptions; account-management standards for the largest customers; enablement before ramp pressure; forecasting and KPI governance so "pipeline" means the same thing to everyone; and a CRM people can actually use.
Spence reports building and scaling national sales organisations four times, including taking C-Suite Network from zero to 11 BDRs with playbooks, compensation design and data analysis already in place.
Partners selling to sales and revenue leaders, or pitching a rebuilt revenue engine, get a more exact practitioner voice here than in Ketelaar's agent-deployment frame.
Spence's argument sits in sales-organisation design, not agent deployment or data readiness.
Charlotte's posts don't address the adoption-versus-production gap, the 50-agent build, the 60 percent rule or the argument that agents fail because of the data underneath them.
A data-readiness vendor choosing Spence would reach a sales-operations audience with a sales-execution argument, not an agent-deployment one.
The proposed webinar and article topics would need to focus on hiring, enablement and forecasting rather than deployment.
Choose Harrington when the buyer is a law firm or professional-services practice and the campaign is about local organic growth and web infrastructure.
Read more
One post argues that law firms are adopting AI for contracts and documentation while still relying on slow agency retainers for client acquisition.
It describes restructuring legal-sector search work into a tiered lifecycle — foundations, growth, domination — and launching BRONSON around custom web frameworks and technical link routing.
A second post analyses nearshore hiring economics, breaking down a LATAM listing that bundled SEO strategy, video production, outbound prospecting, CRM administration and automation into one contract at roughly $1,200 a month.
Partners selling marketing infrastructure or managed search into legal services get directly usable material here.
The audience and argument both differ.
Harrington's posts address law firms and marketing operations, not software founders and operations leaders evaluating agents.
They don't address agent deployment, data readiness or GTM strategy.
The nearest overlap is AI as a delivery tool inside marketing, a different claim from AI as the thing being deployed.
A two-part article series would need a different subject and buyer.
Choose Timcke when the partner sells AI governance, controls, audit or risk tooling, and the audience is risk, audit, finance and IT leaders rather than founders and operations staff.
Read more
One post argues that internal audit is about to meet a population it cannot sample; controls in an agentic environment act jointly like a circuit rather than a checklist; and a monitoring agent needs its own identity, credentials the watched agent cannot see, and a different model family to avoid correlated failures.
A second argues that AI will dissolve deterministic work in Finance and IT and that the surviving function is a risk-first architecture.
Partners campaigning to govern agents that are already running get a more exact match here.
Timcke's frame starts after deployment: controlling a live agent population, testing constraints and rewiring organisational structure away from IT and Finance silos.
Assessing whether the data underneath an agent is ready, and why most purchased agents never reach production, comes earlier.
A vendor selling data readiness or agent onboarding would reach a governance audience focused on post-deployment problems.
The proposed deliverables would need different topics.
Charlotte Ketelaar's eight LinkedIn posts cover AI agents and go-to-market, enterprise positioning, data readiness, compute economics, hiring, venture valuations, and firm operations.
The posts have no publication dates, so the profile treats them as undated and makes no claims about how the arguments changed over time.
The posts here include no long-form articles, transcripts, video, webinar records, engagement metrics, audience demographics, sponsorship policy, or rate card.
Several posts cite podcasts, banks, and analyst forecasts; the posts do not provide those sources, so figures appear as claims in the posts rather than independently verified findings.
Similar-creator comparisons draw on posts by Greg Spence, Bronson Harrington, and Lindsay Timcke, dated between June and August 2026.
The follower count was observed 11 October 2026.
The posts here include no earlier career material, books, shows, or publisher pages; that gap is left open rather than filled.
