Airtable Omni arrived in 2025 and was immediately positioned as the feature that would democratize app building. Some of our consulting peers panicked a little when it launched — if clients can describe what they want and Omni builds it, why would they pay a consultant? A year in, we have a clear answer.
Omni doesn't replace consultants. It changes what consultants do all day. It collapses the most mechanical, least-differentiated part of the job — "click here, add this field, drag this column" — and hands consultants back the time to do the parts clients actually value.
This is the operating manual for how we use Omni at Business Automated in real client engagements: the workflow, the pricing, the conversations with clients, and the parts where human judgment still carries the whole project.
The Old Workflow vs. the New Workflow
Before Omni, a typical small-to-medium Airtable engagement looked like this:
- Discovery (1-2 days) — Understand the client's process, data model needs, and goals.
- Architecture design (0.5-1 day) — Sketch the schema, views, and automations on paper.
- Initial build (1-2 days) — Create tables, fields, linked records, initial views, initial automations. Repetitive, not particularly creative.
- Refinement (1-2 days) — Iterate with the client, handle edge cases, add polish.
- Data migration (0.5-1 day) — Import existing data into the new structure.
- Training and handoff (0.5-1 day) — Walk the client team through the system.
Total: 4.5-9 days. The initial build phase was the most time-consuming part and the least valuable per hour — it's work a well-documented checklist could teach anyone to do.
With Omni, the same engagement looks like this:
- Discovery (1-2 days) — Unchanged. This is where expertise earns its keep.
- Architecture design (0.5-1 day) — Slightly faster because you can sketch in Omni and see the result live.
- Omni-assisted scaffolding (2-4 hours) — The tables, fields, initial interfaces, and basic automations come together in hours, not days.
- Refinement (1-2 days) — Same as before, maybe slightly longer because you have more time for polish.
- Data migration (0.5-1 day) — Unchanged. This is still very much a human job.
- Training and handoff (0.5-1 day) — Unchanged.
Total: 3.5-7 days. The savings — roughly a day of billable time per mid-sized project — come entirely from phase 3.
Where Omni Does the Work Well
Specifically, these are the parts of a client engagement where Omni has become our default tool:
1. Initial schema scaffolding. Feeding Omni the notes from the discovery meeting and letting it build the first draft of tables, fields, and linked records. The result is 70% of the way there. Naming conventions and field types usually need adjustment, but the skeleton is solid.
2. Dashboard prototyping. Clients respond much better to "here are three dashboard concepts, pick one" than to a written description. Omni lets us build three concepts in twenty minutes before the next client call. See our guide to building interfaces with Omni for the prompting specifics.
3. Basic automation wiring. Describing a straightforward workflow ("when a record enters the 'Ready to Invoice' stage, create an invoice record and send a notification") and letting Omni wire it up. Complex automations still need manual work, but the routine ones are Omni's sweet spot.
4. First-draft field agents. When a client wants AI classification or extraction on their records, Omni can set up the field agent with a reasonable first prompt. We then refine the prompt and add the review loop manually.
5. Data analysis during discovery. Pointing Omni at a dump of the client's existing data and asking it to identify patterns, outliers, or gaps. Great for finding surprises before you design a new schema on top of bad assumptions.
Where Omni Absolutely Doesn't Help
The parts of the job where Omni is irrelevant or harmful to lean on:
1. Discovery. No AI tool can sit in a conference room, ask the fifth follow-up question, notice the thing the client didn't say out loud, and connect it to an architectural decision. This is the core of consulting, and it's entirely human.
2. Schema judgment calls. Omni will happily build a schema that works. It won't always build the right schema for this specific business's specific constraints. Decisions like "should we split these into two tables or keep them in one?" or "should this be a linked record or a rollup?" depend on context Omni doesn't have.
3. Legacy data migration. Pointing Omni at a pile of inconsistent CSVs and asking it to "figure out the structure" produces messy results. We still do migration manually — it's the part of the job where paying attention catches the most bugs.
4. Business logic in formulas. Omni writes formulas, but it doesn't always understand the business meaning of those formulas. An SLA calculation for a law firm is different from an SLA calculation for a support team, and Omni doesn't know the difference unless you tell it. Review every formula Omni generates.
5. Change management and training. Building the system is half the job. The other half is getting the client team to actually use it, which involves training, documentation, and running point on the first few weeks of adoption. None of that is Omni's department.
6. Opinionated design. Good Airtable systems are opinionated — they make specific decisions about how things should work, which fields matter, and which views are the "one true view." Omni defaults to generic. Consultants default to opinionated. The opinionated version is almost always better.
How We Price Omni-Assisted Engagements
Pricing is where we see the biggest mistakes. Three common traps to avoid:
Trap 1: "Omni makes us faster, so we'll charge less." No. The time Omni saves is the boring work. The value you deliver — discovery, judgment, training, edge cases — hasn't changed. Cutting your price because a tool made a small part of the job faster is giving away the consulting margin for free.
Trap 2: "We bill hourly, and Omni reduces hours, so we'll just bill less." If you bill hourly and Omni cuts your hours by 30%, you just took a 30% pay cut. This is the biggest argument for shifting to fixed-price or value-based pricing. Clients pay for the outcome; Omni is a tool you use to deliver that outcome more efficiently.
Trap 3: "We won't tell clients we used Omni." Also wrong, and usually transparent. Clients notice. Be open about using Omni — frame it as "we use AI-assisted building to ship your project faster, so we can spend more time on the parts that actually need our expertise." That framing converts the AI story from a threat into a value-add.
What we do instead: fixed-price engagements based on scope and outcome, priced from what the client system is worth, not from the hours it takes us. Omni lets us deliver more polish in less calendar time, which increases client satisfaction without cutting our margin.
How We Talk to Clients About Omni
Clients have anxiety about AI-built systems. The most common concerns we hear:
- "Will it be robust enough to actually use?"
- "Will we be locked into whatever the AI generated?"
- "Is this going to be a 'no one thought carefully about our business' kind of build?"
Our standard answer has three parts:
1. Omni is our tool, not our decision-maker. We use it to speed up the mechanical parts of setup. Every decision that affects how the system behaves in production is made by a human consultant with your business in mind.
2. Everything is editable. The result of an Omni session is a normal Airtable base. There's no special AI container, no locked-in structure, nothing that can't be refactored manually. You're not buying a black box.
3. We review everything before you see it. Omni's output never goes to the client unreviewed. We catch the formulas that don't quite match the business logic, the field names that aren't in your language, and the automations that would misfire at the edges.
This conversation takes thirty seconds and defuses almost every AI anxiety we've encountered.
A Real Engagement Walkthrough
Anonymized but otherwise accurate. A professional services firm came to us wanting a project tracking, invoicing, and client communication system. Estimated scope: 5-7 days.
Day 1 — Discovery. Three hours of interviews with the managing partner and two project managers. Notes on process, pain points, team structure, and existing tools.
Day 2, morning — Omni scaffolding. Fed Omni a structured summary of the discovery notes. In about 90 minutes of prompting and iterating, we had: six tables (Projects, Tasks, Clients, Time Entries, Invoices, Payments), linked records tying them together, two initial interfaces (project dashboard, client overview), and a draft invoicing automation.
Day 2, afternoon — Refinement. Schema review and adjustments. Pulled apart a junction table Omni had built incorrectly. Added custom formulas for utilization and billability. Renamed fields to match the client's language.
Day 3 — Automation and polish. Built the real invoicing automation (with more logic than Omni's first pass could handle). Set up Slack notifications. Added conditional formatting and emoji flags to interface views.
Day 4 — Data migration. Imported three years of historical project data from spreadsheets and an old CRM. Mostly manual. Omni helped with initial field mapping but didn't handle the cleanup.
Day 5 — Training and handoff. Walkthrough sessions with the team. Documentation delivered. First-week check-in scheduled.
Total: 5 days. Before Omni this would have been 6-7 days with the same scope. The saved time went into extra polish on the interfaces and an extra training session the team appreciated.
The Mindset Shift
The biggest change Omni has made to our practice isn't technical. It's psychological. Before, we felt like we had to justify the time it took to build a new base — "here's why it's 5 days, not 3." Now we feel like we have to justify why it still takes any time at all. The answer is that the time has moved from the mechanical parts of the build to the parts that need judgment, and clients benefit from that shift.
If your consulting firm is still resisting Omni, you're competing against firms that have already adopted it — and losing. The adoption cost is essentially zero (it's built into Airtable and free to use), the learning curve is a few days of experimentation, and the productivity gains are immediate. The only real risk is delaying the adoption.
The Work That Still Only Humans Can Do
We've written this whole piece about what Omni does well, but it's worth ending on the other side of the line. The work that still can't be automated, even in an AI-native Airtable:
- Understanding what the client actually needs (not what they said they want in the first meeting).
- Making schema decisions that hold up as the business grows.
- Writing formulas and automations that match business logic nobody wrote down.
- Migrating data that was never clean to begin with.
- Training a team that's resistant, tired, or skeptical.
- Deciding when to say no to a feature request.
These are the things clients pay consultants for. Omni hasn't touched any of them. If anything, it's made them more visible as the work that matters — now that the scaffolding stopped taking all of our attention.
Work With Us
At Business Automated, we build Airtable systems with the new workflow: Omni for the parts where it shines, human expertise for everything else. Our clients get polished, production-ready systems in less calendar time, with the decisions that matter made by people who understand their business.
If your firm is evaluating an Airtable engagement and wondering how AI-assisted building changes the equation, get in touch. We'll walk you through our process — including where Omni fits in and where we still roll up our sleeves.
