GmailJobberGuard · SheetFormulaBridge

Lightweight offline-first desktop connector that reliably watches a Gmail inbox, uses local LLM to extract work order details from forwarded threads, and creates complete Jobber Requests via GraphQL without polling or data loss on crashes.

Who pays

Small HVAC, plumbing, and landscaping companies already paying $30-250 one-time or recurring for VAs to manually process inbound emails into Jobber; the brief states budget ceiling of 30-250 USD for the automation, proving money is moving from manual workflow.

The evidence

Exact workflow in paid freelance brief [2]: 'monitors an incoming Gmail inbox, reliably extracts structured work-order data using an LLM, and creates fully populated Jobber Requests via Jobber's GraphQL API' with requirements for real-time Pub/Sub, forwarded thread handling, and robust error recovery. Also echoed in multiple paying-a-human briefs showing service businesses already spending on manual email-to-job entry.

Why now

Jobber's own API has improved but still leaves Gmail parsing fragile for non-dev owners; local LLMs (Ollama, LM Studio) now run well enough on modest hardware for accurate extraction without sending sensitive customer emails to cloud providers.

Smallest shippable thing

Electron app that connects to one Gmail account via Pub/Sub, uses local LLM (Phi-3 or similar) with a fixed prompt to extract fields, calls Jobber GraphQL to create Requests, with local SQLite queue for offline resilience. Ship in 3 weeks.

Competition & the wedge

Zapier/Make.com exist but require cloud LLM calls (privacy issue) and polling/webhooks that break on nested threads. Incumbents won't build a fully local, offline-first version that stores nothing externally because it cannibalizes their usage-based revenue and crosses email vs field-service boundaries.

Best case against it

Jobber already has native email-to-request features and many users just use Zapier; market of small service businesses needing this exact offline robustness is too small for a standalone product.

Why it survives that

The paid brief [2] proves the native integration and Zapier both fail at forwarded-thread extraction and reliability, which is why someone is spending real money on a custom solution right now. Offline-first local execution is a structural moat no cloud incumbent will match.

Cheapest test this week

Post in 5 active Jobber Facebook groups and DM 15 owner-operators with a 45-second Loom demo of the exact Gmail-to-Request flow from the brief + $49/mo Stripe pre-order link; green-light if 4+ pay the deposit this week.

Dead-simple Google Sheets add-on that adds true 3D formulas (e.g. =SUM(Sheet1:SheetLast!B2)) plus automatic relational list views that survive sheet renames and scaling, without forcing users into Apps Script or Airtable.

Who pays

Property management companies and villa/STR cleaning operations running complex monthly turnover reports across 20–50 property tabs in one workbook; they already pay for VA time or custom scripts ($75k+ salaries in related briefs) because scheduling and billing break when sheets are added/renamed.

The evidence

Direct quote from [8]: 'I'm trying to script an equivalent to Excel 3D formula into Google sheets... function sum3D... But this would sum the sheets only defined by their names' – shows the exact moment the built-in workaround breaks. Reinforced by multiple HN outgrew-the-spreadsheet posts complaining about Google Sheets failing at relational work that Excel handles natively.

Why now

Google has repeatedly refused to add native 3D references despite years of requests; with rising volume of property/portfolio businesses moving from Excel to Sheets, the exact gap is now painful enough for small operators to pay a narrow tool.

Smallest shippable thing

Apps Script add-on with two functions: SUM3D(range, startSheetIndex, endSheetIndex) that dynamically scans sheets, plus a sidebar that creates a synchronized relational view of tagged rows. Under 3 weeks for solo dev.

Competition & the wedge

Existing add-ons are either full BI tools or require heavy scripting. Google itself structurally will not ship this because it would reduce pressure to upgrade users to Workspace Enterprise or their paid database products; the wedge is staying inside the free/cheap Sheets ecosystem the buyer refuses to leave.

Best case against it

This is already solved by writing a better Apps Script function or switching to Excel Online; buyers will not pay for such a narrow feature.

Why it survives that

The unanswered softwarerecs-style post [8] shows even technically capable users cannot get the dynamic sheet-range version working reliably, and the target cleaning/property businesses explicitly do not want to hire devs or leave Google Sheets, matching the 'could not build it themselves' rule.

Cheapest test this week

Find 12 property managers via Google Maps in vacation rental hotspots, send them a Google Sheet with the exact 3D formula broken vs working mockup from signal [8] plus $19/mo annual pre-order link; green-light if 3+ purchase within 5 days.

Lightweight browser extension that records and replays safe bulk delete sequences for Facebook Business page photos, 'Your Photos', and 'Photos of you' folders without triggering blocks, saving auto repair and salon owners from manual clicking hundreds of obsolete images.

Who pays

Auto repair shops and salons who took hundreds of photos on their old Facebook business page years ago and now pay VAs or spend owner overtime doing repetitive clicking; money is already moving via freelance briefs for similar admin cleanup work.

The evidence

Exact steps described in [11]: 'I need to delete hundreds of images from my facebook business page... Left-click where the mouse is positioned (where the image menu is) Move the cursor to where the "Delete photo" option is...'. User wants a macro because Facebook offers no bulk delete and the page is for an old store now obsolete – pure manual-workflow pain with no answers.

Why now

Facebook's anti-automation measures have gotten stricter, but browser extension APIs plus short recorded sequences with random delays can still work reliably; small local service businesses are finally cleaning up old social assets as they adopt proper tools like Jobber.

Smallest shippable thing

Chrome extension using Puppeteer-like recording (or simple content-script click sequencer with configurable delays) that targets the exact menu paths described. Fully working for one vertical in under 3 weeks.

Competition & the wedge

General macro tools (iMacros, UI.Vision) exist but get instantly banned on Facebook. No incumbent will build a purpose-built safe delete tool for old business photos because it directly reduces engagement on the platform and sits outside their revenue model.

Best case against it

Market too small – only businesses with very old cluttered Facebook pages need this, and most will just ignore the old content or hire a cheap VA instead of buying software.

Why it survives that

The specific SE post [11] proves the workflow is painful enough for someone to ask publicly for a macro, and the buyer profile (local service businesses) matches the target who already pays humans for repetitive admin; the moat is the continuously updated safe click-sequence logic that casual macro tools cannot maintain.

Cheapest test this week

Search Google Maps for 15 auto repair shops and salons with clearly outdated Facebook pages (pre-2020 photos), DM them a 30-second video of the extension deleting 10 test images safely + $29 one-time Stripe link; green-light if 4+ buy immediately.