John Glasgow

The argument comes from three seats he says he occupied: corporate finance, the accounts receivable side at Invoice2go, and the accounts payable and partnership side at Bill.com after its acquisition of Invoice2go.

• Looking back from the finance team's chair, the deciding factor was not feature comparison but the prospect of a long migration to software everyone already disliked.
• That experience is the stated basis for applying to Y Combinator to build the general ledger he wished his own team had been on.

At Invoice2go, Glasgow writes, the company ran one accounting system in one country and a different one in another, with consolidation, FX, department-level spend, and revenue accounting all living in spreadsheets, while the ERP was little more than a bank feed.

• The team knew it needed a mid-market system.
• It never switched, because everyone dreaded the migration.

The later posts shift from explaining why teams stay put to describing what an AI-native system does day to day, such as agents drafting entries and reconciling payroll costs, so the origin argument becomes the setup rather than the whole message.

Read original

Glasgow frames this as how he wants AI in finance to work, not as a limitation.

• Entries prepared by an agent are approved by a person and logged with a full audit trail, and the Slack post says the accrual feature was built to be audit- and SOX-ready.
• The workflow stays inside the thread where the team already works so the context stays attached to the entry.

In Slack, a team member can ask what engineering spent on AI tokens last month, or have the agent draft entries for accruals and allocations.

• The agent does the account lookup, the reversal date, and the tie-out; a person still owns the judgment call.
• Glasgow shows an accrual the team booked as an example.

Read original · Original 2

The Campfire-ADP announcement positions the fix as more than a native payroll integration: the company describes an MCP connection between the two products so agents can work across reporting and accounting.

• The stated goal is finance teams spending less time reporting numbers and more time affecting them.
• Glasgow credits Joseph Algerio and the ADP team as partners on the work.

Agents prepare the payroll accrual, reconcile headcount costs, and explain a headcount-driven variance while the period is still open. Each entry is prepared by an agent, approved by a person, and logged with an audit trail inside Campfire, and the data also feeds Campfire's Forecast product.

Read original

Glasgow describes Forecast as the first ERP that calls the month before the close, and the partnership post describes a continuous and autonomous close and forecast. The commercial argument attached to it is that finance teams can stop spending their time reporting the numbers and start spending it influencing them.

The September recap lists Forecast alongside an autonomous Migration Agent and an accounting foundation model among 22 product updates, and the ADP post says the payroll data flows into Forecast to support the continuous close.

Read original · Original 2

Basecamp, which Glasgow describes as the largest AI conference for corporate finance teams, is the vehicle for that argument, mixing hands-on labs with sessions on running AI at enterprise scale and going agentic in an audit environment. Guest speakers and their employers are named in the posts, and their views should be read as their own rather than Glasgow's.

Basecamp ran October 21 to 22 in San Francisco with a reported 500 or more finance leaders, labs where teams built a flux analysis that drafts its own variance commentary, and speakers including Stacie Faggioli of OpenAI in a fireside chat, Elena Gomez of Toast, and, in a separate post, Meta's chief accounting officer Aaron Anderson on how Meta uses AI in accounting.

Read original · Original 2 · Original 3 · Original 4 · Original 5

The September recap counts 22 product updates and a record month of new ARR, then folds in stage appearances, a webinar with Ramp, a conversation with a customer's accounting chief, and European customer wins. Legacy ERP displacement is stated directly: London customers moving off NetSuite, and a founding post about a migration that never happened.

The recap lists an accounting foundation model, an autonomous Migration Agent, Forecast, and Ember Projects as the month's highlights, and the billboard post asks customers who spot the sign between SFO and San Francisco to photograph and tag it.

Read original · Original 2

Glasgow traces the idea to three perspectives: corporate finance, where spreadsheets filled gaps in an inadequate ERP; accounts receivable and payable at Invoice2go and Bill.com; and partnerships work, where he saw ERP pain across business customers. He says teams avoided switching because migration to poor software seemed daunting.

For brands planning a campaign

For a founder-led interview, ask how finance teams can evaluate migration risk and identify spreadsheet work that signals their current systems are failing.

He gives concrete Slack examples: asking what engineering spent on AI tokens, or having Campfire’s agent draft accrual entries. He says the agent handles account lookup, reversal dates, and tie-outs, while a person owns the judgment call; the work stays in the team’s existing thread with context attached.

For brands planning a campaign

A collaboration could demonstrate an AI-assisted accounting task end to end, then make the human approval and audit controls part of the discussion rather than treating automation as a substitute for judgment.

The post describes an ADP Workforce Now integration and an MCP connection.

• Glasgow says agents can prepare payroll accruals, reconcile headcount costs, and explain headcount-driven variances while a period is open; a person approves each entry, with an audit trail.
• He also connects the data to Campfire Forecast.

For brands planning a campaign

For a partnership or podcast brief, explore how finance teams can connect payroll and ledger data, and ask what controls make agent-prepared entries usable in an audited process.

Glasgow lists 22 updates, including an accounting foundation model, a Migration Agent, Forecast, and Ember Projects, a persistent workspace for AI agents.

• He also mentions a webinar with Ramp, a conversation with Meta’s chief accounting officer, and customers ranging from startups to enterprise finance teams.
• The post is a founder’s product-and-growth roundup, not independent evidence of product results.

For brands planning a campaign

A campaign could focus on one of these specific workflows or on the challenge of moving from legacy ERP systems, while checking product claims and customer outcomes independently before publication.

The recurring argument is that ERP migrations stall not because finance teams lack reasons to switch, but because they dread a long move to software they already dislike. That history is the stated basis for the Y Combinator application behind Campfire.

First-person origin story built from named employers, roles, and one concrete failure to migrate.

For brands planning a campaign

Works as podcast or interview material because it is specific and personal rather than a product pitch, and it lets a partner brand align with a founder explaining why the category needed rebuilding.

Examples: Source 1

The ADP announcement describes agents preparing payroll accruals, reconciling headcount costs, and explaining a headcount-driven variance, with every entry prepared by an agent, approved by a person, and logged in an audit trail. The Slack post shows the same pattern at smaller scale: the agent handles account lookup, reversal date, and tie-out, while a person keeps the judgment call, and the workflow is described as audit- and SOX-ready.

Explains controls through worked examples rather than a feature list.

For brands planning a campaign

This is the most transferable talking point for finance leaders who want AI but are accountable for controls. A partner brand can build co-hosted sessions around audit trails, approval boundaries, and where human judgment has to stay.

Examples: Source 1 · Source 2

The same post covers stage appearances, a Ramp webinar, a conversation with Meta's chief accounting officer, and London customer wins described as moves off NetSuite.

List-style progress recap mixing product, revenue, events, and competitive displacement.

For brands planning a campaign

Useful for a partner timing a launch-adjacent collaboration, because it shows what John Glasgow treats as newsworthy and how often the company publishes milestones.

Examples: Source 1

Coverage includes speaker and sponsor lists, hands-on vibe coding labs, and a reported 500+ finance leaders, with a separate post citing 600+ at a prior summit. Session subjects reported include agentic workflows in audit environments and enterprise-scale AI adoption.

Event promotion with named speakers, sponsors, and agenda detail.

For brands planning a campaign

An event is the clearest partnership surface: sponsorship, a session, or a lab. The published sponsor list shows which companies already sit adjacent to this audience.

Examples: Source 1 · Source 2 · Source 3 · Source 4

Other posts point to a Ramp webinar and a Slack integration described as the deepest of any ERP.

Positions integrations as evidence of product depth, naming the partner and the technical mechanism.

For brands planning a campaign

Shows a pattern of publicly crediting partners with named, mechanism-level mentions rather than generic endorsement, which is a reasonable expectation to set before agreeing to co-marketing.

Examples: Source 1 · Source 2 · Source 3

The billboard post frames accounting as shifting from transactional to strategic work.

Contrast framing that names the incumbent category rather than a specific rival product.

This is a competitive position, not neutral commentary. A brand selling competing ERP or accounting software is a direct competitor, and the natural collaboration sits with adjacent categories such as payroll, spend management, or audit.

For brands planning a campaign

For an adjacent fintech the displacement message is useful framing; for a competing ERP the fit is poor and any engagement risks endorsing a rival.

Examples: Source 1 · Source 2 · Source 3

Length varies from a few paragraphs to a longer narrative. The corpus covers product, events, partnerships, and marketing moments, but not research essays, long-form analysis, or third-party media appearances.

Format-level observation across the whole reviewed set.

For brands planning a campaign

A partner should brief for LinkedIn-native, milestone-driven posts rather than long-form analysis, and supply the milestone or launch hook that fits that pattern.

Examples: Source 1 · Source 2 · Source 3 · Source 4 · Source 5 · Source 6 · Source 7 · Source 8

Posts cover Campfire's accounting foundation model, a Migration Agent, the Forecast product, Ember Projects, and a Slack integration for finance workflows.

View example

The integration builds an MCP connection between the two products for agentic payroll accruals, headcount cost reconciliation, and variance explanations, live for ADP Workforce Now and Campfire customers. ADP also appears as a Basecamp sponsor.

View example

The posts describe these as joint activities rather than Glasgow promoting Ramp to an audience on its behalf.

View example

The post treats this as a past role, not a current one.

View example

The post describes the role in past tense and notes the company was later acquired by Bill.com.

View example

No role title or dates are given.

View example

The posts do not otherwise describe the relationship status with Y Combinator.

View example

Like Glasgow, Choi explains finance tooling from hands-on operating experience rather than vendor marketing.

Choi writes as a fractional CFO surveying several vendors and is willing to leave a verdict open while an implementation plays out. Glasgow's posts argue the case for the ERP he founded, so the same operating detail points at one product.

Read original · Original 2

That is the same ground Glasgow covers when describing agents that prepare entries a person then approves.

Howe concentrates on the approval and coding layer around an existing ledger and is openly skeptical that autonomy should touch core systems. Glasgow's position is the reverse emphasis: an AI-native ledger is the fix, with audit trails and human sign-off built into the design.

Read original · Original 2

Both treat the questions finance teams actually ask as the material for product direction.

Chess writes from inside an established ERP vendor about meeting finance teams in the AI tools they already use, and the posts read as reflective analysis of observed patterns. Glasgow's posts are founder updates on one company's shipping pace, partnerships, and conference.

Read original · Original 2

Sun Choi advises multiple startups from seed to Series B and writes comparative tool roundups from the buyer's seat, placing AP and card tools (Ramp, Brex), accounts receivable automation (Tabs), and ERP candidates including Campfire, Rillet, Everest, and DualEntry side by side with the tradeoffs between them, such as Ramp's operating-account yield versus Brex's FDIC coverage, and calling out where legacy players like Recurly and Zuora have fallen behind.

• A separate post walks through board-meeting preparation.
• For a brief that asks finance leaders to evaluate AI-native accounting tools, this is the closest read to the reader's own decision: a practitioner testing software at early-stage companies and reporting what moves the needle, which suits an interview or a category-comparison panel.

Sun Choi is not a company founder, so collaboration is an independent review or advisory placement rather than a joint partnership with shared engineering and a co-announced integration.

• The audience skews to seed-to-Series-B startups, the exact stage of the companies in the roundup, which is a different reader from the enterprise finance teams John Glasgow reaches through the Basecamp programme and the ADP work.
• No follower, engagement, or sponsor data was supplied for Sun Choi, so reach and commercial availability are unverified and the tradeoff is a narrower, less provable distribution than John Glasgow's.

Examples: Source 1 · Source 2

The two reviewed posts each develop a single thesis with data and a worked example: one frames AI in finance as an architecture gap, contrasting the near-universal claims of agentic adoption with the share of AP teams still keying invoices by hand, then walks through a property operator's utility bills being read, split by service address, coded, and held for human approval when the line items do not sum.

• The other shows that real approval chains depend on where a cost is going, using a foundation contractor that routes bills by a letter inside the job number, a construction company with vendor-and-amount exceptions, and a property manager with per-building approvers.
• That style fits a jointly authored explainer on approval boundaries, auditability, and where a person stays in the loop, which is the same human-sign-off theme the John Glasgow brief is built around.

The platform behind the work is narrower.

• MakersHub is an AP and approval-automation product, not a full ERP with payroll, forecasting, and a partner ecosystem, so the collaboration carries less enterprise infrastructure to build on and no documented conference or partner programme to slot into, in contrast to John Glasgow's Basecamp event and ADP announcement.
• Only two posts were reviewed and no audience, event, or sponsor data was supplied, so co-marketing reach is unproven and the brand would be trading a conference-and-partnership play for a writing-led one.

Examples: Source 1 · Source 2

One reviewed post reports what customers actually ask an ERP, drawn from thousands of Ask Oracle conversations, and shows the questions escalating from overdue invoices to what is driving a cash shortage and which costs are pressuring margins.

• The other announces a Finance Analyst skill delivered inside Claude and ChatGPT, covering close management, reconciliations, variance analysis, AR and AP aging, and board reporting, with controls built in.
• For a brief about AI-native accounting, this is the closest match on real usage patterns and on the bring-your-own-AI platform angle, and the ecosystem-level framing, meeting finance teams inside third-party LLM tools rather than only in a proprietary product, is a neutral topic a brand can co-host without promoting a single challenger.

This is an incumbent platform voice, not a challenger.

• NetSuite is the legacy ERP that John Glasgow and Campfire frame as the system finance teams should migrate off, so a brand running a founder-led challenger story would be funding the incumbent's positioning by choosing this alternative, and would need to be comfortable alongside a large ERP vendor's agenda.
• The evidence covers two posts and no event, partner, or audience data, so the conference and partner-programme reach that makes John Glasgow a multi-surface collaboration is not documented here.

Examples: Source 1 · Source 2